Tollymore Letters To Partners Dec 2021

Letters to partners
A letter to partners, December 2018

Contents
Tollymore investment results ................................................................................................................................ 2

A letter to partners, June 2017: an introduction ............................................................................................... 3
A letter to partners, December 2017: execution > philosophy ................................................................... 11
A letter to partners, June 2018: an error of judgment ................................................................................... 19
A letter to partners, December 2018: symbiosis ............................................................................................. 23
A letter to partners, March 2019: The six deadly sins of institutional money management ............. 38
A letter to partners, June 2019: the rise of the platform and implications for owners ........................ 49
A letter to partners, September 2019: incentives, wealth, and happiness ............................................. 60
A letter to partners, December 2019: three decisions ..................................................................................69
A letter to partners, March 2020: Tollymore’s process in a crisis ............................................................. 79
A letter to partners, June 2020: four unpopular opinions ........................................................................ 109
A letter to partners, September 2020: profiting from change without predicting it ........................ 116
A letter to partners, December 2020: operating systems .......................................................................... 122
A letter to partners, March 2021: collecting red flags ................................................................................. 156
A letter to partners, June 2021: Tollymore - the next chapter ..................................................................160
A letter to partners, September 2021: shrinking supply chains ............................................................... 163
A letter to partners, December 2021: shrinking supply chains part II .................................................... 177
Tollymore Investment Partners LLP is an appointed representative of Eschler Asset Management LLP which is authorised
and regulated by the FCA
Tollymore investment results
Dear partners,
$1 invested in Tollymore at inception was worth $3.35 on 31 December 2021, before fees paid to
Tollymore. Over the same period $1 invested in the MSCI All Country World Index would have
generated a return of $1.09. Around three quarters of this benchmark outperformance would
have accrued to partners invested at inception. Tollymore has generated returns of 21% per
annum net of all fees and expenses1, vs. 14% for the MSCI ACWI.
USD Tollymore (gross) Tollymore (net) MSCI ACWI

2016 17% 15% 8%
2017 25% 22% 24%
2018 (2%) (3%) (10%)
2019 21% 20% 26%
2020 92% 81% 16%
2021 1% (0%) 18%
Cumulative 235% 194% 109%

Annualised 24% 21% 14%
1Portfolio inception 12 May 2016. Managed account results are reported until 30 June 2021, after which results of
Tollymore Segregated Portfolio Share Class A are reported. Results of MSCI All Country World Index (ACWI) are total
return including dividends. Results are presented in USD, unaudited, net of all expenses, 1% management fee and
10% incentive fee in excess of a 6% compounding hurdle.
2
A letter to partners, June 2017: an introduction
Dear partners,
On 12 May 2016 I began managing your capital in separate accounts. The aim of this letter is to
communicate the performance of that capital one year later, and to explain the investment
philosophy and process that has delivered it, and that which I expect to produce satisfactory
investment results long into the future. Later in the letter I outline my objective to launch a
money management business, and expand upon the investment proposition for potential
investors who may wish to partner with me in establishing such a venture.
Philosophy
One of the cornerstones of my investment philosophy is a belief in the advantage that a long-
term mind-set affords to patient investors with patient capital. While the informational source of
edge for professional investors has decreased over time, the shortening of security holding
periods has increased the advantage of time arbitrage for long-term investors. Compound
interest, according to Einstein, is the eighth wonder of the world. As a long-term investor my
efforts are focused on finding companies capable of increasing their intrinsic values through
their own efforts. This compounding of value creation takes time, and the compounding can be
lumpy, but it can lead to pretty astonishing long-term results. However, to maximise the intrinsic
value of a portfolio of equities over the long-term it isn’t sufficient to identify these long-term
compounders, but to employ a strict price discipline, which will involve saying “No” to many,
many investment ‘opportunities’ along the way.
In assessing business quality, I distinguish ‘legacy moats’ from ‘reinvestment moats’. Legacy
moats are those businesses with sustainable competitive advantages but limited opportunities to
allocate capital to high returns projects. Reinvestment moats are businesses which generate
supernormal profits and have high reinvestment opportunities. The distinction is important
because it affects how one should think about optimal capital allocation. It also informs an
investor’s willingness to pay for a business as a multiple of normal earnings power. Efforts are
focused on estimating the returns that can be earned on existing and future capital, and the
reinvestment rate of the business. The keys to successful reinvestment moat investing are a long-
term horizon, being right about incremental returns, and being right about reinvestment
opportunities. If both reinvestment rates and incremental returns are high, over the long-term
the entry multiple will pay an increasingly small role in determining the IRR of an equity
investment. Of course, we may be wrong about returns and reinvestment rates, so we also want a
purchase price that is low relative to current earnings.
One of the consequences of this long-term focus is that catalysts are not a prerequisite to
investing. The timing of the market’s recognition of the true intrinsic value of a business is not
predictable. The key is therefore to only invest when there is a very large gap between the
market price and a conservative estimate of private business value. If the prospects for a
particular business warrant a price that is double the market quotation, then even if the market
takes five years to reflect this, we would still earn a very satisfactory aggregate return of 15% per
year.
It is also not sufficient to target investments that can generate high annual returns over the long-
term; this would reflect an impudent focus on the upside only. I want to invest in the equity of
companies for which the odds of permanent capital loss are low. Efforts are therefore focused on
finding companies for which the risk of structural intrinsic value impairment is low, whose
capital structure is conservative and appropriate for the business model, and for which the risk
of a material de-rating of the company’s multiple of earnings power is low.
So the hurdle is high. I want exceptional business quality. I want exceptional reinvestment
opportunities. I want exceptional management, properly aligned with business owners. And I
want all this for a price that will produce a very high probability of an acceptable annual
investment return over the long-term, with minimal risk of permanent capital loss. Despite a
global opportunity set, there will be a very small number of stocks which satisfy these high
demands. And when they can be found with conviction, it makes sense to put meaningful capital
to work. This means that portfolios are concentrated; by this I mean 10-15 holdings.
Diversification can be, in my view, an overused and poorly considered method to lower ‘risk’;
rather it makes mediocre investment returns considerably more likely. Excessive and
inadequately resourced diversification can prohibit the ability to carry out the due diligence
required to think like business owners, and to know how to act when the quoted market price
changes. Concentrated portfolios may be more volatile than the market, but short-term
performance is irrelevant in the long-term accumulation of wealth.
The core tenets of this investment philosophy are not unique. But this is a hugely competitive
industry, so what is the edge that will generate superior long-term investment results? It is
behavioural. It is the ability to act in a way that is faithful to the tenets of the investment
philosophy described above. I believe behavioural edge is the most obvious opportunity to make
better decisions and earn better returns over the long-term. Behavioural edge is about creating a
culture and environment that create the best chance of making optimal decisions. Optimal
decisions are those that allow us to satisfy our goals of creating, rather than transferring, value,
and of solving our investors’ problems. It is predicated on a belief that we can tip the odds in our
favour by repeatedly making good choices day in, day out.
Process
So the cornerstones of my investment philosophy are thinking long-term, maximising the

intrinsic value of a concentrated portfolio, and downside risk mitigation. I have developed an
investment process that I believe will give me the best chance to staying faithful to this
investment philosophy. A process governs the decisions we make day to day. These decisions
can relate to how we organise our time, where we focus our research, portfolio management,
operational and marketing efforts. We make so many decisions every day that creating an
4
investment process that gives us a better chance of making sound decisions can cumulatively
add up to substantially influence our prospects for success. There are several aspects to my
investment process discussed below:
Idea generation: Ideas are not generated according to a systematic magic formula. Ideas are
typically found by reading investment journals, company filings, transcripts, fund manager
letters and industry reports, attending investment conferences and having conversations with
likeminded investors. There is a healthy dose of serendipity involved. This works well if it is
accompanied by passion, high personal energy, and competence. I also use some basic screens to
potentially identify high quality businesses facing short term, but surmountable, problems, but
whose attractive long-term prospects are intact.
Due diligence: My goal is to identify easy to understand, high quality companies, run by honest
and competent management, and buy those companies when they are trading at significant
discounts to conservative appraisals of intrinsic value. In order to do that I will spend time
considering and appraising the business model, business and management quality, and forming
a view of the potential long-term returns that could be enjoyed at current prices. The research
process is bottom up, one company at a time. I will study factual information relating to the
company, its industry and its management in order to form an opinion on business simplicity,
capital structure, growth prospects, competitive advantage and stewardship. Valuation work will
typically involve estimating the long-term, through-cycle owner earnings’ potential of the
business, the opportunity to profitably reinvest those earnings, and the capital allocation
decisions management is likely to make. I am trying to find stocks which can compound at 15-
20% sustainably with limited risk of permanent capital loss. The due diligence process also
involves the use of fundamental and behavioural checklists. These are designed to lower the
risks of behavioural short cuts leading to investment errors. Research is documented, filed and
updated to retain intellectual honesty.
Portfolio management: The portfolio is concentrated, typically holding 10-15 securities. If I

successfully identify companies that can increase their own intrinsic values over the long-term,
holding periods will be many years. There are only three reasons to sell a stock: (1) the price no
longer offers an adequate margin of safety (the stock appreciates or intrinsic value declines); (2)
there is a more lucrative investment opportunity; (3) we have made an investment error. The goal
of avoiding permanent capital loss is facilitated by: avoiding leverage; employing a long only
strategy; investing in high quality businesses run by talented and appropriately incentivised
management; only investing in businesses I understand; and patiently waiting for unduly
distressed prices.
Environment: If we are going to make good decisions consistently, we need a physical working
environment that is conducive to independent thinking. We want to avoid devoting brain power
to battling distraction. Having space and time to work alone, actively disconnecting from always-
on devices and social media, closing Outlook and Bloomberg, and clearing schedules are useful
tools to help cultivate a focused work practice, and uncover the truth by enabling independent
thought.
Hedging: The portfolio is not hedged for currency exposure 2. There are several reasons why an
unhedged portfolio is consistent with the investment strategy:
• FX is not a real asset. It has no intrinsic value and therefore cannot increase or decrease its
value over time. ‘Investing’ in currencies is therefore not consistent with the stated
investment programme.
• I have no ability to intelligently predict currency movements 3.
• There is a cost that will reduce long-term performance.
• Over the long-term currencies have had little impact on the return of real assets. Even the
prolonged strengthening of the USD over the last 25 years would have added only one
percentage point per year to a UK portfolio of US assets (ignoring the indirect effects on
companies with global operations).
• It is a global portfolio with global companies. There are therefore indirect as well as direct
consequences of currency movements. E.g. consider a UK portfolio holding a US company
with a USD share price, exposed to foreign competition. USD strengthening will
operationally harm this company by affecting its ability to compete with cheaper imports.
And foreign sales of the US company are worth less in USD. Yet the translational impact on
the USD security is positive.
• Hedging is a price to smooth returns. Maximising returns is an objective; smoothing returns
is not.
• Investors can hedge their individual exposure to their holdings if they wish.
Investment case study
The largest holding in the portfolio is TripAdvisor Inc. TripAdvisor is also by a distance the worst
performing investment over the last 12 months, declining by 25% since I initially acquired shares
for $61 in September 2016. Prior to acquiring shares, the stock had fallen by 40% over the
previous two years due to a business model augmentation that, in my view, is temporarily
depressing profits but improving the long-term prospects for the business. Since the initial
purchase of shares the company has reported two sets of quarterly earnings which disappointed
analysts’ short-term forecasts, causing share price declines of c.20% on both occasions. I
acquired more shares each time, lowering the average purchase price to under $50, a price which
I believe considerably undervalues this exceptional reinvestment moat business.
TripAdvisor is the largest travel research company in the world. Its 500mn reviews and 390mn
unique monthly visitors are significantly higher than online travel agent (OTA) and travel
research peers. It generates revenue through targeting advertising sales to OTAs and other travel
providers. Revenues are driven by the number of monthly hotel shoppers, conversion rates, and
2I estimate that GBP weakness contributed c.9 percentage points of the 51% 12 month investment result.
3Or GDP growth rates, credit cycles or other macro variables; it is difficult enough assessing the long-term prospects
of individual companies.
6
price per click. Despite TripAdvisor’s material influence in travel commerce, the company earns
revenue on only half of its visitors. Monetisation of its customer base is impaired by travellers
conducting research on the TripAdvisor website or app but then clicking off to complete their
booking elsewhere. OTAs make 6-7x revenue per visitor vs. TripAdvisor.
To stem this economic leakage TripAdvisor has rolled out a feature called Instant Book. This
feature allows users to complete their travel booking directly on the TripAdvisor website or app.
Instant Book positions TripAdvisor as the most compelling one-stop shop travellers can use to
research, plan, price compare and book trips. TripAdvisor is also increasingly a compelling in-
destination travel companion, offering the ability to research and book attractions and
restaurants.
There is evidence of the presence of a strong moat. Average returns on capital over the last eight
years of reported financials have been 30%. Even over the last couple of years TripAdvisor has
out-earned its cost of capital despite the significant investments the company is making in
transitioning the business model to be better positioned to take a slice of the global bookings
pie. Amazon pulled the plug on its hotel booking service after just six months in 2015, perhaps
testimony to the scale of entry barriers, even for an operator with a huge global user base and
trusted brand. Entry barriers emanate from several factors: a globally recognised and trusted
brand; one sided network effects (more reviews = more users = more reviews), and two-sided
network effects (more travellers = more travel suppliers = more travellers). TripAdvisor is
investing in widening its moat by investing in adjacencies such as attractions and restaurants.
The CEO expects Attractions to be TripAdvisor’s next $1bn revenue business (total revenues
$1.5bn, attractions industry = $80bn in N. America and Europe).
So the moat seems strong. Are there substantial reinvestment opportunities? The addressable
market is large and expanding. 40% of the $1.3trn global travel market (growing at c5% pa) is
generated online. TripAdvisor’s market share of this is just $1.5bn/0.3%. Of the $50bn travel
advertising market, one quarter is generated online. Therefore, TripAdvisor should benefit from
the following trends:
• Expansion of the global travel industry.

• Online consuming a larger share of the global travel industry.
• Online travel advertising share (25%) ‘catching up’ with online travel spend share (40%).
• TripAdvisor generating revenues from travel bookings ($1.3trn market) as well as travel
advertising ($50bn market).
Management is long-term focused; Steve Kaufer, the CEO, has the capacity to suffer short term in
the pursuit of long-term shareholder wealth creation4. He has led the company though a
business model augmentation5 in the recent past, a transition he executed quickly and profitably.
4He refers to this as “optimising the multiyear financial outcome”

5TripAdvisor changed its advertising model from ‘pop-ups’ to ‘meta’, the result of which was to deliver fewer, but
higher quality/more easily convertible, leads to OTAs, in the process delivering a measurably improved consumer
experience.
7
Management’s focus on its strategic imperative has been unwavering as it seeks to better
monetise the 40-50% of global online hotel reservations the company influences.
With much of the reported margin declines reflecting investment in revenue growth vs. the
servicing of current business activity, it could conceptually be capitalised. It could be argued that
a business with very high maintenance operating margins and opportunities to reinvest capital
into value accretive projects should do so. Yet when that investment is reported as income
statement expenses, therefore driving operating margin declines, it is poorly received by short
term investors. TripAdvisor’s EBITDA margin in FY16 was 24%, and its ROIC was 21%6.
TripAdvisor adds back stock based compensation to its EBITDA measure; removing this results in
ROIC of 14%. This from a company heavily in investment mode.
The business is soundly financed; net cash held on the balance sheet is c.10% of the market cap.
The stock trades at 19x trailing FCF7. While the market reacted impatiently to analysts’ inability to
successfully forecast the company’s earnings, the business KPIs in my view are consistent with
management’s belief that IB will help to allow TripAdvisor to more adequately monetise the
influence it currently generates in the global travel booking market, and that this influence is
well protected by significant competitive advantages relating to the network and customer
proximity TripAdvisor has built over time. Without assuming a re-rating of TripAdvisor’s current
FCF, I believe the intrinsic value of the business could exceed $30bn over the next decade 8, with
the risk of permanent capital loss mitigated by a strong balance sheet, enduring product utility
and competitive advantages, and competent long-term oriented management.
My objective
It is my sincerest hope to one day launch an investment partnership with partners for whom my
investment philosophy and approach resonates with their own long-term wealth maximising
objectives. There are two aspects to this goal. The first is to create value in an industry that in
aggregate transfers value from investors to managers. The second is more selfish. It is to create a
flexible lifestyle in which I can organise and prioritise my time in a way in which there is no
distinction between work and play. In a way in which business and investment success are
determined by productivity, relationships and personal energy (and serendipity). This flexible
lifestyle, I hope, will allow me to commit unreservedly to the happiness of my family and the
delight of my investment partners on a multi-decade basis.
The opportunity
As mentioned, this investment strategy is not original. In fact, it is its decades-long efficacy,
evidenced by the track records of a number of superinvestors9, and its suitability to my
temperament, that toughens my conviction in its merit. If achieving a better long-term
6 Using the reported effective tax rate of 21% (ROIC = 17% using a normalised 35% tax rate)
7 Adjusted for net cash.
8 Implying annual compounding on an equity investment at the current share price > 20%.
9 Buffett, Munger, Akre, Gaynor, Rochon, Fisher, Pabrai.
investment result requires a competitive advantage (frequently referred to in the industry as a
“source of edge”), I believe it is behavioural. The cornerstones of this behavioural advantage are
independent thought, epistemological humility, and patience. The ingredients required to be
able to make decisions in accordance with these attributes are a long time horizon, the right
environment, the right temperament, and the right investment partners. To someone without a
broad overview of the institutional money management industry, it might seem like the adoption
of these unoriginal principles is an easily copied recipe for success. However, I believe that there
are major obstacles to achieving these qualities for the vast majority of asset managers, including
short termism, incentive structures, poor alignment and overconfidence.
Alignment
The core investment strategy is not flexible. I’m not trying to be all things to all people. The terms
of a potential future venture, however, are flexible. The goal of achieving genuine alignment of
interests between the investment manager and investment partners is, I hope, easily acceptable.
The mechanisms to best achieve this are more nuanced. Here are some proposed features of an
incentive structure that I would like to discuss with potential partners: operational costs that are
recharged to the fund in full, profits delivered above a high water mark that are shared between
the investor and the manager in a ratio of 4:1, and a minimum hurdle rate of 5%. The
performance-based incentive structure is consistent with a returns rather than asset-gathering
objective. The hurdle is designed to avoid the undesirable outcome that the investment manager
gets rich by delivering mid-single digit returns to investors. A 5% hurdle would mean that the
investment manager and the investor both profit from double digit returns and the manager
does poorly in the case of single digit returns 10. Due to budget-based management fees that are
recharged on an absolute basis rather than a percentage of assets, higher assets lower the
expense ratio; at $50mn AUM the cost of operations would be a 20-25bps drag on performance.
The commitment that the vast majority of my liquid capital will be invested alongside partners’
capital will provide the skin in the game that will help to facilitate optimal decision making in the
pursuit of long-term capital compounding. The goal is returns, not assets. However, the ideal size
depends on the investor base and the problem they need to solve. An AUM cap may at one
extreme help to maximise returns, but may prohibit my ability to be a problem solver for larger
investors. Most asset managers optimise for maximising AUM rather than returns, resulting in
value transfer rather than value creation. One further way to encourage long duration capital
might be to implement fee reductions the longer the assets are invested.
10 Performance fees are zero for returns < 5% or are below a high water mark.
9
Reasons not to invest
I’m fanatical about partnering with compatible investors. I believe that the right investment
partners are an important ingredient to sustaining a behavioural edge. I, very consciously, am not
trying to create something to suit everyone. In an effort to discourage incompatible investment
partners, here is a list of reasons why you should not consider becoming an investment partner
in a potential future venture:
• You want smooth returns.

• You will need the capital short term.
• You cannot take at least a five-year view of performance.
• You believe the investment performance is the most important information in this letter.
• You believe the investment performance since inception is remotely sustainable.
• You would redeem your capital if the market had a 40-50% correction.
• You measure risk in terms of volatility.
• You believe there is a positive linear relationship between portfolio concentration and risk.
I have had some wonderful conversations with potential future partners and am thoroughly
enjoying this process, regardless of where it might lead. To everyone to whom I have spoken,
thank you for your time, insights, and advice. I look forward to staying in touch and updating you
on my journey.
To those for whom I currently manage money, thank you for entrusting me with your capital.
Yours sincerely.
Mark
10
A letter to partners, December 2017: execution > philosophy
Dear partners,
On 12 May 2016 I began managing your capital in separate accounts. The aim of this letter is to
communicate the performance of that capital and to explain the investment philosophy and
process that delivered it, and that which I expect to produce satisfactory investment results long
into the future.
Philosophy: why quality matters
As a long-term investor my efforts are focused on finding companies capable of increasing their
intrinsic values through their own efforts. Ideally, I am trying to identify high quality businesses
facing short term, but surmountable, problems, but whose attractive long-term prospects are
intact. And I am trying to buy them at times when the market’s extrapolation of the business’s
near-term challenges into the future makes little sense in the context of the quality of the
company and its management. In the short term there can often be little correlation between
price and value, or stock success and business success. But over the long-term this correlation is
high; this disparity is the key to making money, but it can persist for long periods of time.
Discipline and patience are required to exploit these price/value discrepancies. As long-term
investors we can exploit the time arbitrage afforded to patient investors with patient capital.
While the informational source of edge for professional investors has decreased over time, the
ongoing shortening of security holding periods has increased the advantage of time arbitrage for
long-term investors. By focusing on quality companies, we can benefit from those companies’
intrinsic value appreciation.
In my experience alternative approaches that necessitate the prediction of sharp asset re-ratings,
and the identification of catalysts that will cause these, are meaningfully more difficult to
execute. By owning a portfolio of businesses steadily compounding the value of their economic
earnings, we are tipping the odds of earning a satisfactory return on our equity ownership in our
favour. Meanwhile we can exploit stock market volatility to manage the portfolio in a way that
may augment this underlying business value compounding. The average difference between the
52-week high and the 52 week low stock price for US large caps is 50%! This vastly overstates the
likely difference in private business valuation from one year to the next. I expect the change in
underlying per share value of portfolio holding companies to do most of the work for me in
delivering acceptable annual investment results, but sensible and occasional portfolio
management decisions should enhance these returns over the long-term.
Process: competitive advantage
So we are looking for businesses that can sustainably create value through their own efforts. This
means finding companies that can earn economic profits in excess of the cost of the capital
employed to generate those profits. We want companies that can do this sustainably. In the
11
absence of lasting unfair advantages, the entry of new capital and intelligent effort will drive
returns towards to the cost of capital. This capital cycle economic theory is academically and
empirically broadly accepted.
There are two ways to potentially profit from this capital cycle. The first is to invest in the mean
reversion of returns to cost of capital levels. This requires the ability to time the entry and exit of
capital within an industry, and the catalysts that might give rise to a change in returns on capital
e.g., the closure of factories, industry consolidation or bankruptcies. In addition, this requires
‘renting’ the stock for a period of the capital cycle when the market is extrapolating forward
financial performance that is likely to mean revert. For these two reasons, this approach is
inconsistent with an investment programme which has long-term business ownership as the
cornerstone of its investment philosophy.
The second approach is to identify companies whose supernormal profit potential is larger and
more sustainable than the market believes. This is consistent with the recognition that patient
temperament and capital are important sources of edge in executing a long-term investment
strategy.
Assessment of business quality involves gathering evidence that supports or refutes the
existence of an economic moat, developing an understanding of the factors that have created
this moat and whether the moat is likely to narrow or widen in the future. This involves a close
examination of management’s incentives and capital allocation ability. Management must be
able to sensibly compare the value of various capital allocation choices. A rigid cash use ranking
will not do. Dividends are value destructive if they are made in lieu of available positive NPV
investments11. Share repurchases are value destructive if shares are acquired at market
quotations materially above intrinsic value. Asset growth shrinks per share business value if
incremental cash returns are below the opportunity cost of capital.
While an appreciation of frameworks such as Porter’s Five Forces is useful in understanding

competitive dynamics within various industries, it is important to safeguard against an overly
prescriptive employment of these frameworks. Sometimes investors seem very keen to apply a
ready-made label to the source of a company’s moat e.g., switching costs, network effects,
intangible assets, cost advantages and so forth. Again, an understanding of these models and
how they can give rise to unfair advantages is helpful, but that understanding is accompanied by
the risk that investors thoughtlessly reach for the label that fits without really understanding
from first principles how the company generates value for its owners. In addition, disruption
across various industries is calling into question the relevance of previously accepted moat
sources. See for example the current debate about brand value in the consumer staples space.
Changing distribution models (from supermarket shelves to Amazon/Alexa ordering
11One of the most undesirable defences of rigid dividend policies I hear from (predominantly UK based) management
teams is that “shareholders expect it”.
12
mechanisms) are calling into question the equity of brands which serve to lower search costs 12.
In addition, the de-linearisation of TV viewing has impaired consumer staples companies’ scale
advantage in purchasing globally consumed TV advertising to reinforce brand advantage. This
has resulted in the strong emergence of challenger brands such as Dollar Shave Club and
BrewDog. In retail the emergence of platform business models such as dropshipping and virtual
inventory solutions are connecting vendors directly to consumers, obviating the requirements
for retailer warehousing, lowering the barriers to entry in online retail by eroding the SKU
variety advantages of incumbents.
Having said this, I tend to be attracted to companies that enjoy genuinely interactive 13 network
effects. As the number of network participants increases arithmetically, the number of
interactions between them increases geometrically, creating switching costs through the higher
per-user utility of the platform. This can lead to winner-take-most outcomes, creating assets
which are extremely difficult to replicate and, with appropriate managerial nurturing, can lead to
tremendous economic value creation.
Execution: making better decisions
Over my thirteen years of financial analysis and investment research experience I have gained an
insight into how the actors in the fund management industry behave according to the incentive
frameworks that are in place at large institutional money managers. If there is one thing in which
I have gained conviction over the years, it’s that incentives matter.
We, fund managers, should be in the business of creating, not transferring, value. In the pursuit
of this goal investors’ capacity to make good decisions is taken for granted. Making good
decisions is not commonly considered a source of edge. Yet my vantage point gained from my
experience on the sell side has given me an overview of the behaviours and incentives of actors
representing the majority of the institutional money management industry. The experience
serves as a reminder that the industry at large is not concerned with behavioural edge, it is
therefore not consciously focused on making good decisions.
As investors we are frequently encouraged to consider the edge that will allow us to outperform
markets. The debate surrounds the trio of informational vs. analytical vs. behavioural source of
competitive advantage. In my experience most investment firms are focused on the first two.
This is partly due to a mandate to gather assets vs. delivering superior investment returns. Undue
weight is given to what looks good in a pitch book vs. what works. In a world in which data has
been rapidly democratised, a focus on informational edge can encourage excessive spend on
travel/corporate access/over the top or misdirected primary research.
12 The value of brands which elicit higher consumer willingness to pay or create positional value, such as in the
luxury goods sector, or those that confer legitimacy, as in the case of Moody’s, JLL or PwC for example, seem more
intact.
13 i.e. those in which the good or service becomes more valuable as more people use it.
13
Analytical edge may be possible, but a belief that we can more intelligently analyse information
than our very smart and experienced peers can open the door to overconfidence and hubris.
When we become overconfident, we become unreceptive to evidence that contradicts our
opinions. It is important to develop strategies and systems to ensure we remain sceptical of our
own views and attentive to the facts and evidence that may support or refute them.
One such strategy is the recognition that behavioural edge is the most obvious opportunity to
make better decisions and earn better returns over the long-term. Behavioural edge is about
creating a culture and environment that gives us the best chance of making optimal decisions.
Optimal decisions are those that allow us to satisfy our goals of creating rather than transferring
value, and of solving our partners’ problems. It is predicated on a belief that we can tip the odds
in our favour over time by repeatedly making good choices day in, day out. These decisions can
relate to how we organise our day, where we focus our efforts, when to start and stop research,
whether to invest, how much to invest, when and how much to sell, forming views of
management, staffing, office location, investment partners, and designing fee structures.
Not terribly contentious so far. Yet there are constraints which make actually making good
decisions quite challenging. Most prominent amongst them the preponderance of short-term
investment strategies driven by short term capital and career risk. Fund managers feel they need
to justify high fees with exotic strategies, proprietary and complex idea generation engines and
creative investment theses. It isn’t clear to me that these lead to better performance.
The organisation of roles and responsibilities within large asset managers can lead to
overconfidence. Sector focused analysts (focused on analytical rather than behavioural edge)
have a small investment universe but are forced to pick winners and losers. This narrow focus
can encourage excessive data collection and detailed financial forecasting, leading to
overconfidence. The segregation of roles can impede rational decision making too. Sector
analysts may have done extensive research but portfolio managers have not. A portfolio manager
operating within such an investment programme never develops the conviction required to act
optimally in the face of share price volatility because his set up encourages him to take the
quoted price as a signal of value.
The design and implementation of thoughtful incentive structures is crucial to being able to
faithfully execute a genuine long-term investment strategy. Fee structures should avoid the
undesirable outcome that investment managers can become rich by delivering below cost of
capital returns to investment partners. Hurdles or investment rebates can help to achieve this
aim. Investment managers without skin in the game face a major impediment to good decision
making, because it makes it harder to fight the asset gathering imperative in pursuit of a returns-
focused investment objective. Entrepreneurs often say that execution is more important than
ideas14. The same is true of investment strategies. The ability to execute a strategy is more
important than the uniqueness or marketability of that strategy (for performance, not necessarily
14 Jason Nazar, the technology entrepreneur and investor, speaks and writes eloquently about this.
14
for gathering assets). So, it’s important to think about how we can best execute our strategy.
There are no points for difficulty or originality in this business; there are points for making
decisions that lead to satisfactory long-term returns for our partners and clients.
Diligent and thoughtful fundamental research on the handful of factors that matter is a pre-
requisite rather than a source of edge. But one should acknowledge the diminishing or even
negative returns of more information. Industry incentives are inconsistent with epistemological
humility; pressure to sound smart is highlighted by the widespread practice of detailed financial
forecasting. Phil Tetlock and James Montier have presented some pretty damning studies
denouncing experts’ ability to predict the future. Building detailed long-term financial forecasts
is a low ROI exercise. If you need a detailed three step DCF to find enough upside, pass. I focus
on multiples of owner earnings, returns on capital, reinvestment rates and capital allocation.
This approach is facilitated by a large opportunity set, which obviates the requirement to know
everything. It allows me to focus on the very obvious gaps between price and value, and pass on
the vast majority of mediocre investment opportunities.
Case study: Gym Group plc
In July 2017 I acquired shares of Gym Group, currently one of the largest holdings in the
portfolio. Gym Group was founded by the current CEO in 2007 and is today the second largest
low cost gym operator in the UK. The company has a coherent mission: to help people improve
their wellbeing, whatever their fitness of financial starting point or location. The fact that
consistently 30% of new members have never been a gym member before suggests this mission is
being accomplished. The low cost gym model has grown rapidly by addressing the historic
barriers to gym membership: high membership cost and being tied into contracts. The
proposition of ‘high quality, low cost’ appears to be well received: GYM’s net promoter score is
>60.
GYM can profitably offer average memberships 60% below private sector averages due to higher
utilisation of site space15. Materially lower labour intensity and customer acquisition costs are
facilitated by online customer signup and management and PIN entry systems. Market leaders
Gym Group and Pure Gym enjoy scale advantages, evidenced by falling new site investment
costs. Gym fit-out contractors are awarded contracts through more competitive tender
processes, better terms are agreed with equipment suppliers and service providers such as
cleaning. Marketing spend has better economics with a higher number of sites and members.
The health and fitness club industry is fragmented, but there have been signs of retrenchment
among midmarket private chains e.g. Fitness First has scaled back its portfolio in recent years
and LA Fitness was acquired by Pure Gym. Mid-tier and premium competitors have lower
margins than GYM; this lowers their headroom for profitable price cuts.
Low-cost gyms have been growing 50% pa but low cost gym membership in the UK is still only
3%. In the US and Germany half of gyms are low cost (and rising), accounting for the higher gym
15 170 stations per gym vs. 63 private market average, limited wet facilities and 24-7 opening hours.
15
penetration rates in those markets. The low-cost segment has both taken share from the
traditional segment and also grown the market (it’s worth repeating that in every year since 2008
>30% of new GYM members have not been a gym member before).
Management has been opening 15-20 gyms per year (15-20% of the current estate). These high
reinvestment rates mean GYM’s reported financials screen poorly: with a P/E ratio of 40x and a
FCF yield of <2%, many value investors may prematurely write off GYM’s investment merits.
However, owner earnings are c.£25mn16, a 10% yield to the market cap when we purchased
shares. Yet the estimated cash on cash returns on new gym openings are over 20%. So, for every
pound of owner earnings invested, GYM can create £2 of value 17. This investment ticks all the
boxes of high incremental returns on capital, high capital reinvestment rates, available to own at
a modest multiple of current owner earnings.
TripAdvisor Inc: an update
Since my last letter in June 2017 the shares of our biggest position, TripAdvisor, continued to
decline, ending the year at around $34 per share. In November the company lost almost a quarter
of its quoted value after lowering FY17 revenue growth guidance, stating that reigniting near-
term hotel revenue growth has been more difficult than expected due to partner bid downs
(Priceline redirecting spend away from performance and towards brand advertising). I acquired
more shares in December for the following reasons:
(1) TRIP’s customer proposition is continuing to strengthen. TRIP has almost four times the
traffic of Trivago (455mn monthly unique visitors vs. 120mn) but visitors are still increasing
17% per year. And these visitors are increasingly engaged, with reviews increasing 32% yoy.
Bid-downs will not harm TRIP’s customer proposition, but it will hurt its profitability.
(2) The big picture remains the market is worth $1.3trn pa (900x TRIP’s revenue) and TRIP’s
influence continues to grow in this market. TRIP has the largest selection of travel listings in
the world (2mn hotels and vacation rentals, 5.3mn restaurants and attractions). PCLN and
EXPE together have < 15% of the global travel market. In addition, the broader onus on brand
marketing that is evident in the third quarter results of all the travel companies may serve to
accelerate the overall shift online.
(3) The reported margin decline overstated the financial impact of these bid downs on the core
hotel business. Reported Hotel EBITDA margins fell from 31% to 16% yoy in 3Q17. The main
drivers were the lower cost per clicks and TRIP’s $42mn expenditure on brand advertising.
Assuming that incremental margins are 75% and return on advertising spend of 75% we get to
an adjusted EBITDA margin in the mid-twenties, i.e. 5-6 pp of dilution from PCLN’s bid
downs.
16 Owner earnings are higher than reported earnings due to (1) mature site EBITDA margins higher than the
immature estate average, (2) maintenance capex lower than depreciation due to new site fit outs, and (3) pre-
opening costs incurred while in expansion mode. The working capital benefit from rent free periods on newly
acquired leases is not included in my calculation of owner earnings as this is a benefit of asset growth.
17 If the >20p of returns are capitalised at an ex growth rate of 10%.
16
(4) I believe the most obvious opportunity for TRIP to improve monetisation of traffic is not
through higher cost per click but rather through higher conversion levels i.e. getting visitors
to click on links rather than getting a higher price for those links. There are three drivers of
revenue – traffic, conversion and CPC. Trivago’s unit economics reflect a customer with
much higher price comparison/purchasing intent vs. TRIP. TRVG has revenues roughly the
same as TRIP’s hotel business, yet its traffic is four times smaller than TRIP. This 4x
monetisation gap is the target of TRIP’s website redesign and advertising campaign; the goal
is to educate visitors that TRIP is a place to price compare. If we assume that TRIP has the
same revenue per click and consider that TRIP is earning $0.40 revenue per hotel shopper, it
implies that 70% of TRIP’s visitors are there to review and post content, and do so without
viewing a hotel page or a price comparison list for a particular hotel. TRIP is clearly trying to
lower this number by raising awareness of TRIP’s price comparison capability and doing so
would be very positive for revenue per hotel shopper.
(5) Hotels are also attempting to build direct relationships with customers – TRIP’s Instant Book
offering allows them to do this. They may therefore bid more for CPCs or pay more for IB
partnerships.
(6) As TRIP continues to build its non-hotel offering (80% of TRIP’s content, bookable products
+30%, revenues +26%, 35% EBITDA margins, up from 12% a year ago), its dependence on
OTAs will decline (currently c.40% of revenues from EXPE and PCLN). For the same reason
TRIP’s traffic and user engagement should continue to grow, improving its viability as a lead
generation partner vs. Kayak, TRVG and Google.
(7) At this point I think we are paying close to zero for the Hotels business. If we value the non-
hotels business on a similar EBITDA multiple as that implied by the post-write down Priceline
acquisition price paid for Opentable (similar bookable content and revenue growth, TRIP
lower, quickly growing margin vs. Opentable higher, steady margin), non-hotels is worth
$3.5bn. That would imply an EV of the hotels business of <$200mn. Given TRIP’s quickly
growing non-hotels margin, the same sales multiple/higher EBITDA multiple may be more
appropriate, implying a non-hotels valuation of $4.4bn. That would imply an EV of TRIP’s
hotels business of negative $700mn.
(8) Takeout potential. It seems to me that one way that PCLN can satisfy its objective to own the
customer relationship is to acquire TRIP. This is probably easier that trying to encourage
TRIP’s 500mn monthly unique visitors to go directly to bookings.com, which can offer them
no price comparison. I don’t think any brand campaign can engender such a level of
consumer trust that customers stop comparing prices (Amazon has potentially achieved this
but not through an advertising campaign). It could be the case that PCLN is flexing its
muscles to hurt TRIP, driving down the stock price in order to more cheaply acquire the
asset. This is potentially a risky strategy if the lower price attracts more bidders and the
tactics make Malone/Maffei less likely to accept an offer from PCLN. The change of control
amendments in the filings and the absence of the buyback renewal are potential signals that
an offer for TRIP could be made at some point.
17
I think M&A potential and the implied negligible value for hotels, which increasingly is not the
‘core’ business, mitigate substantially the risk of permanent capital loss for owners of TripAdvisor
shares. This is not the same as near-term further quoted price declines. Should PCLN
reaccelerate pressure driving CPCs another leg down this could cause the implied value of TRIP’s
hotel business to become negative. The likelihood that the private business value of the hotels
business is actually negative is remote. Maffei/Malone are smart enough to take appropriate
action if the hotels’ business subsidy is too large to justify the synergy enjoyed by the non-hotels
business. As such, TRIP continues to be the largest position in the portfolio.
Thank you for entrusting me with your capital. Yours sincerely.
Mark
18
A letter to partners, June 2018: an error of judgment
Dear partners,
In June I sold our equity interest in McCarthy & Stone, the UK retirement homebuilder. You
should expect the disposal of portfolio holdings to be a relatively infrequent occurrence, given
the long-term investment horizon of the investment strategy and my conviction in the merits of
time arbitrage as a source of investment edge. Specifically, there are three reasons to sell shares:
1. The margin of safety has declined, either through significant intrinsic value deterioration or
share price appreciation to an extent that lowers a conservative appraisal of the investment’s
IRR to a level that is not materially higher than the opportunity cost of investing.
2. There is a more lucrative investment opportunity, either within the current portfolio of
afforded by a new investment candidate.
3. I have made a mistake. I will make lots of these. Indeed, my investment process welcomes
them as opportunities to learn. The documentation and filing of investment research provide
a chronological journaling of decision making over time that can be held to account as the
facts unfold to support or refute the original hypothesis. Treading the fine line between
dogma and conviction is one of the most difficult challenges in investment management. But
exercising sound judgement is one of most important determinants of long-term investment
outperformance. Material share price underperformance or new negative fundamental
information will therefore trigger a reappraisal of investment theses. After which comes an
important decision. Average down, exploiting the market’s volatility to lower the average
cost of our ownership of a great business? Or recognise the flaws in our original analysis,
own up to our misjudgement and use the experience as a case study to improve future
decision making?
I think that McCarthy & Stone falls into the third category.
MCS is the largest retirement housebuilder in the UK. The company buys land, secures planning
consent, builds, sells and manages housing developments specifically designed for retirees. My
investment thesis was that MCS is a dominant niche homebuilder that was well positioned to
benefit from an ageing population and structural UK housing undersupply. I expected high
earnings’ reinvestment rates to compound MCS’s intrinsic value over a long period of time. At
the same time, I viewed the risk of intrinsic value impairment as modest due to my assessment
that MCS had low business risk, was conservatively capitalised and traded at a valuation that did
not reflect this positive risk reward skew. Taking these in turn:
19
Business risk
In MCS’s 40-year operating history, it has enjoyed a dominant position within a largely
uncontested niche. Major homebuilders have tried and failed to enter this market. MCS is the
dominant provider of owner-occupied retirement housing in the UK, with 70% market share.
Private competitors are much smaller, regional operators. A significant portion of retirement
housing is publicly owned and takes the form of ‘sheltered housing’, owned by local authorities,
the majority of which was built 50-60 years ago. Apartment design is standardised; MCS’s scale
and ability to utilise repeatable processes is a cost advantage. MCS also enjoys planning
advantages due to the social value of its products.
MCS acquires sites through conditional purchase contracts, mainly conditional on the granting
of planning consent. In many cases contracts will include commercial viability clauses which
gives MCS the flexibility to cancel purchases for projects which become uneconomic. In market
downturns, MCS can exit conditional contracts and sell land to non-residential interests such as
supermarkets and other commercial interests, lowering development risk.
Management incentives are long-term and aligned with equity owners; return on capital
employed is a feature of the company’s three-year vesting LTIP.
Balance sheet risk
MCS finances its operations from operating cashflow; shareholder equity represents 90% of
long-term capital. Financial gearing has dramatically reduced since the financial crisis and a low
level of operating leverage complements the prudent capital structure of the business.
Valuation risk
When I acquired shares MCS’s quoted enterprise value was barely higher than the company’s
invested capital. In my view this made little sense for a business generating 20% return on
existing capital employed and substantial opportunity to reinvest operating cash flows into the
development of more homes for a long time. MCS estimates that almost one million households
are in the optimal bracket of being 75+, living alone/with a spouse, and having high housing
wealth. Yet only 140k units have been built to date in the UK. This c.800k shortfall is large in the
context of the market leader’s targeted production of 3k pa.
In June 2018 management issued a press release stating that there had been a “noticeable decline
in reservation rates” due to “a slower secondary market and a softening in pricing, particularly in
the South East”. Due to a slower than expected spring selling season operating profit for FY18 is
expected to be £65-80mn. Previously management stated it was comfortable with analysts’ PBT
range of £ 91-108mn. This is a significant adjustment to make over a two-month period. The
forward order book is still growing +10%, but this has slowed from +13% two months ago.
Essentially, customers have responded to the declining values of their existing homes by
tempering their interest in purchasing retirement housing. This is likely to affect MCS more than
20
mainstream homebuilders as MCS’s customers are typically trading down, so the cash value they
can extract for their existing home diminishes.
My thesis has been that in a worst-case scenario (one in which the company earns no freehold
reversionary income due to the government’s ground rent ban), MCS could earn 11% after tax
ROTC. Being able to purchase the business for its invested capital therefore would be an
important safeguard against permeant capital loss.
Several developments challenged this thesis:
• The speed with which the business has deteriorated may mean that historic returns are not
an appropriate starting point. I want to avoid at all costs buying an average business in
favourable business conditions, mistaking it for a high-quality business. The June release
highlighted that the company’s profitability is influenced by external factors to a greater
degree than previously assumed.
• The chairman stated that he expects the business to generate mid-teens ROCE over the next
year, and the CFO confirmed this included FRI income. Removing c.£25mn of FRI income on
£70mn of operating profits (35%) would lower this pre-tax ROCE to <10%, or <8% after tax.
This would imply the market is expecting no material deterioration nor improvement from
here.
• There is yet another major personnel change as the CEO follows the previous CFO and
Chairman out the door.
• The chairman hinted on the call that a potential future business model change (US/Australia
type retirement village options) will be part of the company’s long-term considerations.
So where did I go wrong?18 In a nutshell, I believe I underestimated the business risk associated
with McCarthy & Stone. I believe that (1) the interrogation of MCS’s financial track record over a
favourable business environment, and (2) my underestimation of the capacity for populist
government policies to materially impair MCS’s economic prospects 19, caused me to
overestimate the quality of this business. As such, my estimate of the incremental returns that
MCS could sustainably enjoy was too high.
In addition to making mistakes relating to disaggregating business quality from favourable

economic conditions, I feel I gave too much credence to management expectations given their
strong track record of performance within this historic industry environment. This is especially
frustrating, as I have written previously about the role of corporate access in a thoughtful
investment process and the need to safeguard against management sales pitches.
18 It may of course transpire in the fullness of time that selling was the investment mistake. In which case it will serve
investors well for me to revisit this analysis.
19 The UK government banned leaseholds on all new residential properties. With MCS no longer able to sell Freehold
Reversionary Interests, this removed c. 5% of the company’s revenues but c. 30% of operating profits.
21
I will make more investment mistakes over the years. You can depend on that. You can also
depend on me to write about these errors of analysis or judgement transparently, and, I hope, to
learn from them to satisfactorily compound your capital over the long-term.
Thank you for entrusting me with your capital.
Yours sincerely.
Mark
22
A letter to partners, December 2018: symbiosis
Dear partners,
Over the past few years, I have discussed Tollymore’s investment philosophy and process in
detail in investment letters, interviews, articles and presentations. I have used investment
decision making examples as a mechanism for current investors and prospective allocators to
understand how I make decisions, and allow them to judge the consistency of these decisions
with the key principles of the investment strategy.
As an emerging investment management firm Tollymore has the advantage of organically

building a set of aligned investment partners, deeply appreciative and practically aware of the
significant behavioural edge that a long-term mindset affords in public market investing;
investors who think like business owners rather than traders.
Key elements of such a symbiotic relationship between manager and client are transparency,
integrity and authenticity. As such, this letter delves deeper into the investment merits of the
portfolio’s largest holdings, as well as the most recent investment. Understanding the intrinsic
value potential of a portfolio allows successful money managers to stay the course in periods of
self-doubt. By being transparent about the companies we own and why we own them I hope to
give investors the tools to share my fundamental appreciation of the portfolio’s potential,
allowing them to invest countercyclically to the ultimate benefit of the investment partnership’s
capacity to sustainably create value for all partners.
The partnership’s four largest holdings, in alphabetical order, are Amerco, Gym Group, ITE
Group and TripAdvisor. All four companies are founder-led20, provide enduring products and
services, and have demonstrative capacity to generate sustained supernormal profits. They differ
in their scope for organic capital reinvestment into valuable projects, and therefore the
importance of the stock’s multiple of owner earnings in determining the potential IRR available
to equity owners.
One thing they all also have in common is that they screen terribly. Reasons are various and are
discussed below. Screens miss one of the most compelling sources of mispriced securities: those
companies whose reported financials poorly reflect the underlying cash earnings’ power of the
business, due to for example high reinvestment rates, inappropriate accounting policies or
business model transitions. All such examples are described below.
20 Gym Group founder John Treharne resigned as CEO and became Founder Director in September 2018.
23
Amerco (UHAL.US)
Amerco ticks all the boxes of a reinvestment moat, stewarded by excellent long-term business
owners and capital allocators, and trading at a modest multiple of normal earnings power. The
DIY moving market was essentially created with the founding by Leonard and Anna Shoen of
Amerco in 1945. Today 40% of movers in North America are DIY vs. professional moving
companies. Shoen spent decades signing up land-owner franchisees across North America,
developing a network across the country which allowed users to collect vehicles in one location
and drop off at another, facilitating ‘inter-city’ house moves.
U-Haul, the brand behind Amerco, has made the most of its first mover advantage by
aggressively expanding its network of franchised moving service locations. Each additional
location increases the value of existing locations by enabling more convenient pick up and drop
off logistics for an increasing number of moving routes. This should increase the steady state
asset utilisation of the business. Franchisees are attracted by the number of U-Haul customers;
customers are attracted by the convenience of franchise locations. This virtuous circle drives
market share and revenue growth, which has been reinvested in more locations and vehicles,
making U-Haul’s dominance increasingly difficult to replicate. Today U-Haul as eight times as
many locations as its nearest competitor, and four times as many trucks. Management has clearly
identified the source of the moat for this business and has continually reinvested capital in
widening it over time. This network advantage limits the ability to add supply in the ‘one-way’,
inter-city moving market.
Trucks are necessary for moving but people move infrequently; these characteristics lend
themselves to a rental business model. Moving depends on idiosyncratic factors such as the
need to upsize, downsize, get divorced, get married, move jobs, attend and leave college etc. The
need for more affordable housing is a significant reason for moving, likely a countercyclical
driver of volumes. If humans live in houses, regardless of how those houses are financed, and if
changing life circumstances cause them to move homes several times through their lifetimes,
none of these drivers of demand is set to change materially long into the future. It is very
possible, however, that Amerco captures an increasing share of the total costs incurred in
making these moves.
Another example of management directing capital to widening the business’s unfair advantage is
the development of MovingHelp.com. Movinghelp.com is a platform bringing together supply of
and demand for moving labour. As the number of providers of labour grows, the more users are
attracted by greater availability. The accumulation of independent reviews also creates stickier
supply and encourages better customer service, increasing demand.
The U-Haul brand is well-recognised and ubiquitous; there are 65% more U-Haul locations in the
North American network (20k) than there are Starbucks coffee shops (12k). The orange and
white livery and U-Haul logos adorn its 100k trucks, 80k trailers, 20k truck rental nodes and 1,100
storage locations, serving as a free mobile and ubiquitous advertising machine.
24
Since its inception U-Haul has grown its dominance in DIY moving. Several competitors have
tried to enter or grow in the space, but none has succeeded in denting U-Haul’s market share.
The industry has consolidated down to three main players today: U-Haul, Budget Truck Rental
and Penske Truck Rental. Budget generates 15% of the revenue of U-Haul with a truck fleet one
quarter the size and has been retrenching. highlighting the economies of scale in the industry
and U-Haul’s network advantage.
The other major use of the company’s operating cash flow has been the purchase and greenfield
development of self-storage facilities. Like truck rental, self-storage is a commodity offering. The
site acquisition and build costs are low, lowering barriers to entry. Hence there are many players
in a fragmented industry. As fixed costs are high there may also be the temptation to cut prices to
fill capacity. However, once tenants are secured, switching costs are high and so rent increases
can be imposed. This, coupled with low maintenance costs and high operating leverage, mean
that returns can ramp up quickly. Amerco can use its U-Haul brand and ability to hook truck
customers into the sales channel via its website to lower self-storage customer acquisition costs.
This potentially lowers Amerco’s self-storage breakeven price and occupancy vs. peers.
I think several factors suggest that the company’s capital can be deployed at high incremental
returns: a wider economic moat, industry consolidation, capital reinvestment into very high
margin self-storage services, operating leverage, and declining labour intensity as technology
lowers the cost of generating business.
Multiple avenues for earnings’ growth suggest that owner earnings’ reinvestment rates can
continue to remain high for a long time:
• Continued share take in the self-moving market (less than half of the total moving market).
Truck rental is subscale for existing players. Budget has been scaling back its fleet and
Penske, the nearest competitor, has 10% of the number of locations U-Haul has. The overall
moving market is c. $15bn (U-Haul is less than one quarter of this).
• Encroaching into the professional moving market (>50% of the market), which a fragmented,
and therefore largely local, industry. There are 7k companies represented by the American
Moving and Storage Association alone, representing 14k locations across the US. Half of
these businesses employ fewer than five people. The sustainability of the professional
moving segment seems threatened by the scale of U-Haul’s network, provision of moving
helpers through Moving Help, and the potential introduction of autonomous vehicles. The
investments into Moving Help could be a significant enabler of this.
• Finally, the self-storage industry is twice the size of the moving industry, c. $30bn. This is a
fragmented market in which there are synergies with the moving business. Given these two
factors, it is possible that certain storage assets are worth more to Amerco than other
bidders. Amerco can rebrand the acquired facility with the ubiquitous U-Haul logo and link it
to its online reservation system, immediately connecting the assets to Amerco’s large
customer base. None of Amerco’s competitors, either in the self-storage or the moving space,
can exploit this synergy to the same extent.
25
Since we initially acquired shares in February 2017 for $370 per share, the stock has fallen 9%,
materially lagging the market over the last two years. There have been several short-term
pressures on Amerco’s profitability over this period. There has been some pressure on operating
profit margins due to the acceleration of truck and self-storage asset expansion and the tweaks
to the company’s depreciation policies to front load tax-deductible depreciation ahead of the
corporation tax reductions. The fleet increased 10% yoy in 2018 but management expects this to
be roughly flat in FY19 as truck acquisition slows and used truck sales pick up. EBITDA margins
however continue to improve.
Asset utilisation has been constrained by low occupancy rates on self-storage assets and lower
truck fleet utilisation. However, self-storage square footage has increased 50% over the last three
years and the CFO on the call stated that revenue per square foot continues to rise 3-4% yoy and
the average occupancy on units older than three years is 84%, vs. 39% for units younger than
three years. Given lofty self-storage market prices, UHAL has been focused on a build vs. buy
growth strategy, with its associated larger drag on average occupancy. Finally, a collapse in used
truck disposal profits was due to an effort to slow down sales and maximise the depreciable asset
base ahead of the corporation tax rate change, and poor execution in underwriting the sale of
waiver insurance, incurring elevated repair costs in getting vehicles into a sellable condition. The
elevated disposal profits are something we highlighted in our original research and normalised
for in the owner earnings calculation.
Listening to Amerco’s investor conference calls is instructive. Some comments from participants
on recent calls reflect a level of frustration at the long-term horizon of management’s capital
allocation choices, pleading for the installation of a regular dividend and buyback. I disagree. A
dividend should be the last thing this company is doing, and despite the stock failing to remotely
keep pace with the market’s rally over the last few years, a buy back would shrink an already
small free float. My sense is that management is making capital allocation decisions that are
designed to maximise the total future economic value of the business. I don’t care whether this is
realised later vs. sooner, if the company’s investments are returning more than our opportunity
cost of investing. With the company’s widening moving and storage network advantage and a
CEO who is paid < $1mn per year but whose family owns most of the business, I think this is
likely to be the case.
Over the last 12 reported months of business operations Amerco generated operating cash flow
of almost $1bn. Amerco’s enterprise value is less than 9x this number. With capex of over $1.7bn
Amerco is currently generating FCF of c. -$700mn. Adjusting for c. $1.1bn of growth capex,
maintenance FCF = +$4-500mn, a 6-8% yield to the market cap for a business with evidence of
high cash reinvestment rates and high incremental returns on capital.
26
GYM Group (GYM.LN)
Gym Group enjoys a cost advantage facilitated by superior asset utilisation and a long runway for
value accretive asset growth. GYM was founded in 2007 and is today the second largest low-cost
gym operator in the UK. It runs 150 gyms and has 700k members. All revenues are generated
from membership fees and joining fees. The company has a sense of purpose/coherent mission,
which is to help people improve their wellbeing, whatever their fitness of financial starting point
or location. The fact that consistently 30% of new members have never been a gym member
before suggests this mission is being accomplished. This is a relatively capital-intensive business.
However, while 30-40% of revenues are spent on capex, c. 4-6% of revenues relate to
maintenance capex.
GYM is a market leader in a fragmented industry: There are over 2k private gym operators in the
UK, running 3.7k private gyms. The ten largest operators account for c. 650 gyms, 18% of the total
number of private gyms, and GYM accounts for c. 4% of all private gyms. The low-cost gym
model has grown rapidly by addressing the barriers to gym membership: (1) high membership
cost and (2) being tied into contracts. The proposition of “high quality, low cost” appears to be
well received: GYM’s net promoter score is very high: +60. 55% of new members come from
referrals.
GYM can profitably offer average memberships 60% below private sector averages due to higher
utilisation of site space (170 stations per gym vs. 60 private market average, and limited
wet/racquet/café facilities/24-7 opening hours) and the employment of technology rather than
employees for customer sign up and management. Customers enter the gym via a PIN entry
system. The joining process is online, or via on-site internet connected kiosks, lowering
customer acquisition and management costs.
In reviewing thousands of potential sites over the years GYM has developed relationships with
landlords and property agents. It is conceivable that the brand and GYM’s strong covenant rating
may be competitive advantages when it comes to securing new sites with landlords.
Low-cost gyms have been growing 50% pa but low-cost gym membership in the UK is still only
3%. Low-cost gyms in the UK are 8% of the total. In the US and Germany half of gyms are low cost
(and rising), accounting for the higher gym penetration rates in those markets. The low-cost
segment in the UK has both taken share from the traditional segment and grown the market (in
every year since 2008 >30% of new GYM members have not been a gym member before). This
has resulted in low-cost club CAGR of >50% since 2011. Initial site investment costs have
declined because of economies of scale. Gym fit-out contractors are awarded contracts through
more competitive tender processes, better terms are agreed with equipment suppliers and
service providers such as cleaning. The economic return on marketing spend has improved as
the number of sites and members has grown.
27
Estate maturation should improve margins: Average mature site EBITDA margin/ROCE is
47%/32%. Yet the average site EBITDA margin is 40%, and the EBITDA margin for the group is
30%.
In my view the principal risks that may cause the future to unfold in a less favourable way than
the above analysis anticipates are: (1) member churn/customer response to real disposable
income erosion; (2) input cost inflation, and (3) irrational and aggressive competitive reaction.
Churn and customer demand elasticity: members can cancel without charge at any time. To re-
join would incur a £20 fee. Annual membership attrition (cancellations net of re-joiners) is 100%;
30% of leavers re-join. Management has stated that it does not consider cancellation to be a KPI
for the business. This is for two reasons: (1) Cancellation improves membership yield as new
members join at higher prices than cancelled members. And (2) the cost of acquiring a new
member is less than the joining fee. This is due to marketing economies of scale and word of
mouth recommendations (more than half of new joiners are costless referrals).
Competitive response: Mid-tier gyms may cut their membership fees to compete with the
increased popularity of low-cost fitness clubs. However, despite prices 60-70% below mid-tier
gyms, GYM’s margin profile is vastly superior. This is also despite having a lower percentage of
the estate comprising mature gyms vs. established non-growing mid-tier competitors. Mid-tier
and premium competitors have lower margins than GYM. This lowers their headroom for
profitable price cuts. The competitive response from the mid-tier/premium segment seems to
have been benign. Mid-tier/premium operators have upgraded their service to warrant the
premium they charge and/or have increasingly targeted the top end of the market which they
argue is not addressed by the low-cost segment. Mid-tier and premium gyms have consistently
increased their fees each year. One mid-tier operator, Fitness First, did attempt to create a low-
cost arm, opening Klick Fitness in 2012 with 6 gyms. Just over 12 months later the group exited
the sub-sector. The mid-tier peer group has been restructuring, with consolidation on-going in
the market.
Leverage/recession risk: In the event of a recession prices and memberships may erode. GYM has
high financial gearing in the form of operating lease obligations. Given GYM’s operating leverage
mature site profitability would fall significantly should the business experience a marked
revenue decline. I estimate that a 20% drop in revenues would reduce the owner earnings to
£17mn from £40mn in a no-growth scenario. This is still a 5% yield to the current market cap.
Current adjusted net debt/EBITDAR = 3.5x; assuming the maintenance profitability of the estate
this is 2.5x. This would be 6x with a 20% revenue decline assuming zero capacity to cut 100%
fixed costs, or 3.8x based on maintenance profitability. Under these assumptions GYM’s EBITDA
margin would be 13%/ and maintenance EBITDA margin 35% vs. the current 30% rate (and 47%
mature gym rate). It is difficult to envisage a 20% revenue decline due to the immaturity and
increasing penetration of low-cost gyms, as well as the 60-70% price discounts vs. mid-tier
competitors. When US membership growth was negative in 2012, Planet Fitness still grew its
28
memberships by 28% yoy. This might suggest that the large discount lowers the price elasticity of
demand for budget gyms.
High reinvestment rates mean GYM screens poorly: With a P/E ratio of 47x and a FCF yield of
negative 4%, many investors will write off GYM’s investment merits. However, owner earnings 21
c. £40mn, a 12% yield to the current market cap.
These owner earnings are largely being directed to clearly above cost of capital projects. I
estimate the total capitalised costs per new site are c. £1.5mn, and the economic profit per gym is
c. £0.3mn, leading to after tax returns on incremental capital in excess of 20%. So, for every
pound of owner earnings invested, GYM can create £2 of value. Unfortunately, the board has a
progressive dividend policy with a 10-20% payout ratio target. With the company’s unit
economics and runway for growth I would prefer that 100% of earnings were reinvested in asset
growth. However, 10-20% of reported earnings will be much lower than 10-20% of owner
earnings. If 90% of owner earnings are reinvested an equity investment in GYM could yield
annual returns more than 20%.
ITE Group (ITE.LN)
What should a demanding investor be willing to pay for:
• A simple, founder-led business.

• A provider of an enduring service, reflecting a business practice which is hundreds of
years old with a high utility to cost ratio.
• Geographically and industrially diversified end markets.
• UK headquartered and listed enterprise.
• Capital intensity of 1-2% of annual revenues.
• Strong FCF conversion of earnings due to deferred income balances 50-70% of revenues.
• Average through cycle returns on capital of around 30%.
• Defensible moat facilitated by two-sided network effects barriers to entry.
• Revenue visibility, with two thirds of following year’s sales forward booked.
• Conservative capital structure.
• Organic revenue growth of 11%.
ITE is involved in one business activity: it organises exhibitions and conferences. ITE hires
venues and gathers a group of exhibitors and visitors, monetising the exhibitor side of the
network by charging companies for floor space. The company operates across several sectors
including construction, food, energy and travel and tourism and predominantly emerging
market geographies.
21Calculated using mature rather than reporting margins, deducting maintenance capex and excluding working
capital inflows (these are driven by rent free periods on new sites and are therefore a benefit of asset growth).
29
While multi-year agreements can be struck to secure venue capacity, these agreements have the
flexibility to modify capacity commitments ahead of changes in demand. There is good revenue
visibility thanks to forward bookings.
Operating margins have been reasonably stable due to moderate operating leverage, but current
levels of operating profitability of c. 22% are around 5ppts below mid-cycle levels. Some of the
margin run off has come as a result of deferred revenue increases, with exhibitors securing their
participation at future events at more competitive rates. What therefore may be surrendered in
terms of income statement profitability is recovered via working capital inflows and FCF
conversion. Despite this, as we will discuss, management has a plan to restore the business to
operating margins in the high twenties.
ITE is an appropriately capitalised business given high revenue visibility, strong repeat business
and multi-year agreements. Despite £50mn of debt funded M&A in FY18, net debt = £80mn, 1.4x
EBITDA, two years of FCF and around one quarter of ITE’s equity. Revenue is recognised at the
completion of an event; cash is received from exhibitors in advance and booked as deferred
income. This is a costless source of finance and represents a working capital inflow when
revenues are growing (but a drag on FCF when revenues decline, as they did FY15). This
deferred income float has typically been c. 40% of sales, but since the founder of the business
returned as CEO this has grown to two thirds, strengthening cash conversion.
A long track record of supernormal profits seems to emanate from two sources of competitive
advantage: (1) Network effects: visitors are needed in order to attract exhibitors and vice versa,
making it difficult for a non-established player to gain traction. And (2) intangible assets:
trademarks and licences to operate venues, databases of consumers and exhibitors, brand
reputation that makes participants reluctant to move away from a tried and tested event. The
value chain is symbiotic: customers do not want competing trade shows; they want to know that
their customers are going to be at the event they are attending that year. This is a form of
efficient scale; there is only space for a handful of profitable operators within each region and
industry niche.
I have classified this as a legacy moat business (rather than a reinvestment moat or a capital light
compounder) due to the lack of greenfield organic opportunities to build new exhibitions
around the world. However, ITE does have a strong foothold in several emerging markets
including China and India. ITE has established market leading positions through fully owned
subsidiaries and controlling interests in locally dominant exhibition brands. For example, it has a
controlling stake in the market leading conferences business ABEC in India which runs 20
exhibitions across the country. ITE has offices in Beijing and Shanghai, operated through a JV
with Hong Kong based Sinostar. ITE also runs conferences across Indonesia via a JV partnerships
and wholly owned subsidiaries.
The organic growth trajectory has inflected positively, partly due to the economic stabilisation in
a number of ITE’s end markets, and partly due to a business improvement programme put in
place as a result of a change of CEO in 2016.
30
A series of macroeconomic and geopolitical shocks have beset ITE’s end markets and impeded
business progress in the three or four years prior to 2017. These included the Russia-Crimea crisis
and resulting trade sanctions at a time when Russia was a much larger part of ITE’s business, and
a 70% collapse in the price of oil harming business in ITE’s energy-dependent central Asian
markets.
This improvement in top line growth should bode well for the capacity of the business to
meaningfully increase underlying cash generation, especially given the strides management is
making in collecting cash for conference pre-bookings. As we will discuss, the market implies
that the exact opposite will happen; that the sustainable cash the business can generate will
wither significantly, and the stock has continued to move in the opposite direction to the
business’s progress.
As legacy moat business with limited opportunities to build new shows organically, M&A will
always be part of management’s capital allocation agenda. It is therefore important to
understand the management team, their track record and incentives. The founder Mark
Shashoua has returned to the business with a plan to improve the organic growth potential and
profitability of the current portfolio. He is divesting the least profitable, sub-scale events. He has
executed this transformation playbook at i2i Events, which enjoyed a doubling of revenues and
profits before being acquired by Ascential Events under his five-year leadership prior to
returning to ITE, the company he founded in 1991.
In 2018 ITE announced the acquisition of Ascential Events Limited from Ascential plc, based on
an EV of £300mn. The target assets comprise two global exhibitions brands, Bett and CWIEME,
and several UK exhibitions brands such as the Spring and Autumn Fairs and Pure. The CEO and
COO of ITE Group are very familiar with these assets; they ran them together 2011 – 2016 as CEO
and CFO.
Management incentives have improved with the management change and are reasonable but
not standout. Base salaries for the CEO/CFO are £450k/£250k. Bonuses are a function of headline
PBT, organic revenue growth, cash conversion and qualitative strategic targets. LTIP awards are
based on recurring EPS and relative TSR versus a combination of the FTSE Small Cap and FTSE
250 Index constituents. What is interesting is that ITE’s earnings per share need to grow by 75%
over the next two years for any of the LTIP to vest. For 100% of the LTIP to pay out ITE needs to
generate EPS of 14.4p by September 2020, almost treble the FY18 EPS, and a 25% yield to the
current share price.
All these measures were only introduced in 2017. While cash conversion is a sensible target,
some measure of incremental returns on capital would be a welcome addition to the incentive
programme, given the high cash generation of the business and the limited obvious long-term
redeployment opportunities. I would say that the track record of the new leadership and some
incentive framework progression are improving stewardship of a demonstrably wide-moat
business that has suffered macro headwinds which have shown signs of dissipating. This is not a
turnaround thesis; this is a fantastic, highly profitable business already, and one that can benefit
31
from three tailwinds: (1) cyclical/macro mean reversion, (2) structural portfolio improvement
driven by positive management change, and/or (3) a potential re-rating of the shares to reflect
the former drivers.
The share price has collapsed since 2014 due to a series of macro and geopolitical events
including the Russia/Crimea crisis, oil price collapse and rouble depreciation.
The share price has fallen by 50% in the last year to a 52-week low. More recently the company
has suffered non-fundamental selling pressure as the portfolio has become much less emerging
market focused. In my view the company’s stock price has moved in the opposite direction of the
business’s performance and private business value.
ITE’s cash flow and owner earnings are materially higher than reported earnings per share due to
a large deferred revenue float, impairments of prior capital allocation decisions and amortisation
of quasi-permanent assets such as customer relationships and internally developed brands.
Trailing FCF of the business is c. £30mn (adjusting for restructuring costs and biennial events);
this is 25% of ITE’s equity and debt funding and 19% of core Group revenues, excluding the
contribution from Ascential. Organic revenue growth is running at double digit levels.
So that is what the business is generating. How is the market pricing that? The current £450mn
market cap accounts for a £265mn rights issue-funded acquisition completed in the summer. So,
the market is valuing ITE’s pre-acquisition cash flows at a 16% yield. This implies material FCF
erosion in the core business or zero or negative FCF margins in the newly acquired assets. Or
that ITE has massively overpaid for Ascential.
This despite the core business recording 11% organic revenue growth in 2018 and the acquired
assets, which ITE’s management believe have been poorly managed, generating 31% EBITDA
margins.
So, did ITE overpay for the deal? They believe they are paying 11.5x EV/EBITDA for an ungeared
business that can grow revenues double digits. That seems reasonable. In the case of zero
revenue or cost synergies they are paying 12.5x EV/EBITDA for a business with high barriers to
entry and capex intensity of 1%. I don’t think that is expensive. It may just be the case that ITE is a
better owner of this business because it is a core asset for them vs. non-core for Ascential. This is
consistent with the comments Ascential have made about the deal (and with Mr. Shashoua’s
prior comments to me about Ascential not being a strong competitor as the company focuses on
its data and analytics businesses).
The market is not paying attention; I have spoken with large institutional owners of the shares
appearing near the top of the register; for them ITE is a small part of a diversified portfolio and
their understanding of the deal and management’s prior history with the assets reflected that. A
conservative appraisal of private business value would imply 70% upside to the current price.
That is if we capitalise the trailing FCF generation of the core business at 6% and add the rights
32
issue capital we arrive at an owner’s value of £750-800mn. The global industry is projected to
continue to grow at 5% pa, with many of ITE’s markets expected to grow high single digits. This
would also imply that management miss their own goals and LTIP targets substantially, as
previously discussed. In fact, management is targeting (and is over-executing) high single digit
revenue growth, margin expansion and large working capital inflows; this would result in FCF
compounding 1.5x revenue growth.
TripAdvisor (TRIP.US)
One- and two-sided network effects and a globally trusted, top-of-funnel, brand strongly
position TripAdvisor to take an outsized share of the growing global online travel market. I have
discussed my views on the TripAdvisor investment case in prior letters, so I won’t repeat them
here. TRIP continues to strike me as a good example of investors’ inappropriate application of
linear projections to platform business models. TRIP’s potential for potentially explosive low-
cost growth emanates from its ability to tap into existing demand (travel and leisure visitors) and
connecting it to acquired or developed supply (Viator, La Fourchette). Therefore, TRIP is
strongly positioned to be the winner in two winner-take-all markets: restaurants and attractions.
For both restaurants and attractions, the number of reviewed items is multiples higher than the
number of bookable items. These low penetration rates of bookable inventory (1% for
restaurants, 8% for attractions) will provide a long volume growth runway.
Adjusting the company’s trailing FCF for stock-based compensation, TV advertising and non-
hotel investments (assuming OpenTable margins achievable on a normalised basis), I estimate
the business is generating c. $400mn of owner earnings, a 6% yield to the current cash-adjusted
market cap, and c. 60% of invested capital. It’s clear that the opportunity and capital allocation
priority for the business will remain reinvestment for a long time. If (1) for every $100 we invest
in TRIP it generates $6 of owner earnings, and (2) that $6 is all reinvested into projects which
generate a recurring $3.60 pa, then TRIP has generated $36 of value, capitalised at our
opportunity cost of 10%. Assuming average incremental returns on capital of 25% over the next
decade and that three quarters of earnings are reinvested yields an IRR of c. 20%, and TRIP would
be worth c. $40bn over the next decade.
I have written in the past about the importance of being able to execute a long-term investment
strategy. The portfolio management decisions that I have taken, and documented, relating to our
ownership of TRIP shares since September 2016 are an example of the freedom that a sound
investment process, appropriate working environment and philosophy aligned to manager
temperament can afford in making decisions that are consistent with our stated investment
philosophy. TRIP’s quoted share price at the end of 2018 was $54, c.11% lower than our initial
acquisition price in September 2016. Yet TRIP’s positive contribution is responsible for c. 8% of
the cumulative performance of the portfolio since inception, thanks to an environment and
investor base that allow us to make investment decisions which are consistent with a long-term
business owner investment philosophy; that is to average down in the face of short-term market
pessimism and reduce exposure in response to excessive exuberance.
33
New portfolio investment
In December 2018 we acquired shares of Trupanion (TRUP) at a price of $23.3 per share. TRUP is a
founder-led, simple one-product-one-geography business with single-minded focus on a niche
service. TRUP has a superior value proposition and strong competitive position afforded by
gestation of distributor relationships, a data flywheel and customer switching costs. It operates
in a large addressable market with a significant penetration opportunity affording a potentially
multi-decade runway for compounding owner earnings. It may come as a surprise that TRUP is
also an insurance company.
TRUP provides medical insurance for cats and dogs in the US and Canada. The problem that
Trupanion is trying to solve is that it is difficult for pet owners to budget for the magnitude and
timing of pet illness and injury. Pet owners do not know the average costs of pet healthcare for
the pets that they own. Even if they did, they would not know if their pets will be lucky or
unlucky for that breed and location.
Trupanion’s solution is a cost-plus insurance product which spreads the risk that the customer’s
pet is unlucky by subsidising unlucky pets with lucky pet premiums. Like any insurance product
Trupanion allows customers to budget for the unpredictable timing and magnitude of loss, in
this case pet healthcare costs.
The distribution of TRUP’s insurance products is primarily through vet and customer referrals.
TRUP uses a network of c. 100 independent contractors called ‘Territory Partners’ to build long-
term relationships with vets; they are responsible for making vets aware of the benefits of TRUP’s
products to the vet’s customers, with the goal of earning the trust of the vets.
Like any insurer, TRUP must estimate and hold reserves for vet invoices which have been
incurred but not yet submitted, a complex process requiring subjective judgement. Unlike most
insurers TRUP is a cash in – cash out business; it does not have a substantial float, nor does the
investment income from that float contribute meaningfully to the business’s discretionary
profits.
Given pet insurance penetration rates in the US of 1%, TRUP’s primary competitor is the pet
owner who chooses to self-fund pet healthcare costs with cash or debt. TRUP is therefore
focused on growing the addressable market vs. taking market share from existing players; the
primary challenge in achieving this is the education of pet owners about the merits of TRUP’s pet
insurance.
The nature of the insurance business model – risk is spread over a large membership for lucky
pets to subsidise unlucky pets – is a barrier to profitable entry for small insurance providers. For
example, given the fragmented nature of the vet industry, it would not be possible for individual
vets to offer their customers insurance products.
34
TRUP is the largest player in Canada and the second largest in the US, behind Nationwide.
Nationwide started in 1982 and was the first pet insurer in the US. It has c. 550k enrolled pets;
TRUP has rapidly grown to c. 500k pets and it is likely that TRUP will become the largest player in
the US soon.
High retention and net subscriber addition rates are evidence of a strong value proposition. The
sources of this superior value proposition stem from (1) a cost advantage that is shared with
customers, (2) a data advantage driving more accurate underwriting, and (3) switching costs and
symbiotic value chain.
TRUP is vertically integrated; it owns its insurance subsidiary and is responsible for acquiring
and servicing existing customers as well as underwriting their insurance. TRUP estimates this
vertical integration has eliminated frictional costs of c. 20% of revenues. These economic savings
have been donated to consumers in the form of higher claims payout ratios. TRUP’s strategy has
therefore been to sacrifice the near-term margin upside of this cost advantage in the pursuit of a
larger and stickier customer base and subscription revenue pool. This cost advantage does not
manifest itself in lower prices, but rather the highest sustainable expenditure on vet invoices per
dollar of premiums.
TRUP has built a database over 15 years using 7.5mn pet months of information and > 1mn claims;
it has segmented the market into 1.2mn price categories in order to more accurately underwrite
insurance costs for a given pet. Of course, determining the point at which the marginal returns
on incremental data diminish is difficult, but according to the CEO it would take a competitor 13
years to replicate this data asset. Although Nationwide is larger by number of pets enrolled, its
data are likely to be less comprehensive for two reasons: (1) a lack of data for conditions not
covered by policies, such as hereditary and congenital diseases, and (2) pricing categories by
state rather than zip code, even though the cost of vet care can vary widely within states. TRUP
considers its ability to accurately estimate the costs of pet healthcare costs by granular sub-
categories crucial to its leading value proposition. This allows for the provision of more relevant
products for the customer.
Trupanion Express is software that was developed by TRUP and integrates with vets’ practice
management systems. Through Trupanion Express, TRUP pays vets directly, within five minutes
of a vet invoice being submitted, disrupting the traditional insurance reimbursement model and
obviating the need for customers to pay out of pocket and then submit a claim for the expense.
This is clearly a superior solution to the reimbursement model in solving customers’ cash flow
problem associated with unexpected pet healthcare costs. In general, pet owners do not switch
insurance providers due to the non-coverage of pre-existing conditions. Trupanion Express is
installed in 10% of the 20k hospitals being visited by Territory Partners each year. The integration
of this software is likely to improve the loyalty of pet owners and vets.
Finally, TRUP insurance seems to be a win-win-win proposition for pet owners, vets and TRUP:
35
• Vets’ treatment decisions can be dictated by efficacy rather than cost. Pet owners visit the vet
more frequently and are more likely to agree to the vet’s Plan A treatment recommendation.
The goal for Territory Partners is to sign vets up to Trupanion Express, which is free, removes
bad debt issues and therefore fosters better relationships with customers. Trupanion Express
also eliminates credit card fees, which may constitute c.15% of a vet’s profits.
• Pet owners have peace of mind that they will not be hit with unexpected large vet bills and
are therefore also less likely to choose economic euthanasia or suboptimal treatment plans.
With Trupanion Express they do not need to settle vet invoices out of pocket and then
attempt to cover the claim through the traditional reimbursement model.
• Through Trupanion Express TRUP improves the retention of its customer base, freeing up
discretionary capital for accelerated pet acquisition.
The addressable market is large and underpenetrated relative to other developed markets. The
differences in these other markets are not demographic, social or economic, but rather (1) the
length of time comprehensive pet insurance has been available, (2) the value proposition in the
form of higher claims payments as a ratio to premiums (higher loss ratios) and (3) vet vs. direct-
to-consumer distribution models. Pet insurance companies in the US typically do not cover
hereditary and congenital conditions, which are the forms of illness most likely to be suffered by
cats and dogs, they increase rates when claims are made, they impose payout limits, and pay
claims according to an estimated cost schedule rather than actual vet invoices. TRUP is different
in all these respects and as such expects to grow the addressable market in North America to
greater than 1% penetration. In any case, it appears to be the case that TRUP’s value proposition
is driving adoption in North America.
The unit economics associated with the pursuit of this opportunity to grow the company’s assets
are attractive. The cost to acquire a pet is c. $150, around 3x the average monthly ARPU.
Assuming the current 10% discretionary margin and a six-year average pet life, the IRR on new
pets is 30-40%. At a 15% discretionary margin the IRR would be double this. I estimate that both
ARPUs and discretionary margins would need to decline by 20-25% to render reinvestment in
pet acquisition a capital destructive pursuit. This would contradict the economic reality of a
market in which pet healthcare costs are increasing mid-single digits as new technologies and
treatments are ported over from human healthcare, and the scalability of the business model.
The CEO owns 7% of TRUP equity/$60mn and in total the executive leadership team owns 10%.
This is c 100x the CEO’s annual compensation. He automatically sells 2% of his shares each year
until he has sold 25% of his interest in the company by 2025. This has been a source of criticism
from short sellers but given the large gap between stock ownership and annual remuneration,
and the zero-dividend policy, I don’t think this represents misalignment with other minority
owners of the business.
TRUP’s quoted market cap is c. $800mn, c.6x BV and 24x owner earnings of c. $33mn. A 4%
owner earnings’ yield is reasonable for a business with TRUP’s high reinvestment rates and
incremental returns on new capital investments if these can be successfully maintained. The
strength of TRUP’s competitive position and evidence that their value proposition is attractive to
36
pet owners suggest that they can. These owner earnings are one quarter of book value, and the
company is growing its assets (enrolled pets) 20-30% each year; 6x BV implies a 6% sustainable
growth rate.
Yet incremental returns on reinvested capital are higher than the return on existing net assets,
leading to growth rates many multiples of that implied by the market. Finally, management
expects scaling of the fixed cost base to drive margin expansion, leading to economic earnings
growth higher than revenue growth. By 2020 if management achieves its targets, it should be
generating c. $450mn of revenues and $65-70mn of discretionary profits. Given the company
expects to reinvest all discretionary profits into growing enrolled pets, retained losses are
unlikely to improve over that time, leading to a potential RoE of c. 50% in 2020 on an owner
earnings basis. At the current price the stock would then be trading at a multiple of its book
value that implied zero growth, despite the ample room for enrolled pet expansion afforded by
low market penetration and a leading value proposition.
Wishing everyone a healthy and happy year ahead,
Yours sincerely.
Mark
37
A letter to partners, March 2019: The six deadly sins of
institutional money management
Dear partners,
My professional experience has provided a broad overview of the institutional investment

management industry. Tollymore benefits from this experience. It provides a context in which to
judge our decision making and prospects for delivering acceptable investment results to our
partners. Specifically, Tollymore seeks to profit from several behavioural constraints impairing
institutional money managers’ execution of a sound long-term investment programme.
(1) The pursuit of informational edge: Overconfidence can have profound consequences,
inflating investors' valuation of their investments, leading physicians to gravitate too quickly
to diagnoses, and making people intolerant of dissenting views. Studies suggest that
confidence and accuracy are not highly related 22. The problem with this as it relates to
investing is that the extra confidence causes us to increase the size of our bets without a
corresponding increase in our capacity to predict outcomes, causing us to lose money.
We do not build complex financial models, designed to convey broad knowledge vs. deep
understanding. We seek to shield ourselves from faulty heuristics by placing corporate
access at the end of an investment process. We strive for simplicity and conduct work with
the goal of gaining conviction in a handful of value drivers.
(2) The pursuit of analytical edge: The need to justify fees creates pressure to conceal rather
than acknowledge ignorance, to build large teams and create the perception of deep
intellectual expertise. This expertise increases the perceived validity of one’s own opinions
and makes one less receptive to ‘non-experts’. Deep perceived expertise promotes trust in
intuition and lowers the inclination for hard System II work. This also geometrically increases
the complexity of organisations, introduces group think, authority bias, and loss aversion by
slowing down decision making. We believe teams should be small, preserving accountability
for decision marking and direct communication channels. We would suggest that diverse
teams, and independence of opinions, are likely pre-requisites for collectively wise decisions.
Mental flexibility, introspection, and the ability to properly calibrate evidence are at the core
of rational thinking and are largely absent on IQ tests. Typical decision makers allocate only a
quarter of their time to thinking about the problem properly and learning from experience 23.
Most spend their time gathering information, which feels like progress and appears diligent
to superiors. But information without context is falsely empowering.
22 E.g. College football experts were asked to predict game outcomes (Organizational Behavior and Human Decision
Processes), professional horse handicappers predicting horse races (Paul Slovic). In subsequent rounds of the game
the experts were given additional information and asked to rank their confidence. The additional information did not
improve the accuracy of their predictions but did increase their confidence in their bets.
23 Michael Mauboussin
38
Tollymore’s large investment universe and concentrated portfolio obviate the need for
valuation precision, which requires analytical edge.
(3) Marketing led investment strategies: Asset gathering business objectives and gold-plated
cost structures magnify the imperative to grow AuM. Investment firms led by marketers
rather than investment managers create strategies tailored to what will sell rather than what
works24. Pressures to justify high management fees create an action bias that may be
antithetical to good investment outcomes. Tollymore’s strategy is capacity constrained; our
focus is on returns vs. assets and aligning the investment philosophy with the manager’s
temperament and investment partners’ objectives.
(4) Short term capital: Asset gathering mandates can lead to unidirectional manager due
diligence processes, misaligned investor-manager relationships and procyclical capital flows.
Short term capital leads to short security holding periods, which directs investment
managers’ efforts away from understanding long run business prospects and toward
predicting share price movements. However, markets, unlike meteorology, are complex and
reflexive; participants are second-guessing one another and the bases on which decisions
are made are altered by the decisions themselves. The volatility of publicly traded securities
makes it very difficult to guess short term price movements.
We do not direct our efforts to understand and analyse the market’s thoughts. We conduct
independent research and focus on long-term, lower volatility outcomes. As an emerging
investment manager, we can develop an aligned and sympathetic investor cohort through
two-way relationship building. We expect these efforts to allow us to focus on building
substantial behavioural edge through an educated LP base capable of investing
countercyclically over the long-term. The long-term investor can purchase securities from
sellers selling for non-fundamental reasons (redemptions from short term capital) or because
they think news flow will be temporarily negative (they have an eight-month holding
period).
Our objectives are to: build a long-term aligned investor base with a business owner
mentality, capable of tolerating periods of benchmark underperformance; assemble the
capital and working environment that will allow us to act decisively when the odds are in our
favour; create an environment that allows us to both average down and acknowledge
mistakes; and surround ourselves with intellectually generous peers and investment
partners.
(5) Manager/investor misalignment: Managers’ and investors’ fortunes are typically not aligned.
This is a barrier to sound investment decision making. A simple way to weed out the
managers that back themselves is to consider the presence and power of incentives – insider
ownership and performance-weighted fee structures. Managers without insider ownership
are less incentivised to limit the size of their fund, therefore limiting their achievable time
24 Examples may include primary research and idea funnels.

39
weighted return. In our view, and in the case of most institutional money managers, the
components of stewardship reflect a product to be sold rather than a strong belief in the
strategy.
(6) Stories: Investment management professionals often have short attention spans, driving the
articulation of eloquent, memorable investment theses. Stories emanate from our
continuous attempt to make sense of the world. As such they serve a purpose. The problem
comes when we conflate explanatory power and predictive power. Story construction itself
is problematic due to self-serving bias: our predilection for favouring decisions that enhance
self-esteem. This results in attributing positive events to oneself and negative events as
situational. The problem is compelling stories have characteristics that are antithetical to
truth finding. They are simple, ascribe outcomes to talent and stupidity vs. luck, and focus on
things that happened vs. things that failed to happen.
Narrative fallacy is the backward-looking mental drive to attribute a cause-and-effect chain

to our knowledge of the past. Without searching for reasons, we would go around with
blinders on, one thing simply happening after another. This helps us make sense of the world
despite sensory overload. However, it can cause us to make poor decisions. The power of
narrative causes us to violate probabilities and logic. Tollymore seeks to profit from narrative
fallacy by applying logic to anomalies. That is, by specifically seeking out stocks without
good stories, or those with bad stories, and to write about the companies we own in
involved, long shelf-life letters to partners.
The forest and the trees: profiting from time arbitrage
Tollymore is in the business of applying logic to anomalies. Our goal is to identify mispricings
that are afforded to us as long-term investors, and to exercise sound judgment when we
encounter potential opportunities. The application of logic comes in the form of having the
temperament, working environment and investor base to see the bigger picture. And to use
noise to our advantage in lowering our cost of business ownership when appropriate, and
therefore the rates of return we can enjoy as owners of publicly listed businesses. We describe
below two recent examples of investment decision making under conditions of uncertainty. In
both cases we used our advantage as a long-term owner to acquire more shares of materially
mispriced businesses at cheaper prices. We do not direct efforts to distilling the market’s
‘thoughts’ into a concise narrative. The complexity and reflexivity of financial markets renders
this a low ROI pursuit. However, in the following examples it is our view that price action was at
odds with the fundamental development of the companies’ long-term economic prospects.
TripAdvisor (TRIP.US): We added modestly to our position in TRIP after 1Q19 results sent the
stock down 11% to $49/share. Some of the news headlines that accompanied these results from
the financial press included “Sales Declines Hit TripAdvisor”, “core hotel segment was flat,
missing analyst expectations” and “TripAdvisor wasn't able to deliver the top-line growth that
many were counting on seeing”. We are not trying to predict share price movements; we are
40
therefore not focused on analyst expectations nor trying to guess how our companies will do
relative to them. It is also conceivable that management commentary on the call about softer
than expected international demand, leading to a cautious Q2 revenue outlook may have also
caused investors to sell their interest in this business as they extrapolate these demand
fluctuations25. That was the trees.
How do we see the forest? We think the current level of profitability in the hotels business can
be taken as a proxy for owner earnings 26. TRIP’s reduction in performance marketing spend has
driven large increases in profitability; the hotels segment recorded 41% EBITDA margins vs. 30%
a year ago. Yet hotel revenues still increased slightly yoy.
Meanwhile restaurants and experiences revenue grew 35% yoy27, but EBITDA losses widened to -
$24mn from -$4mn as TRIP accelerated investments in adding bookable supply 28, which
continues to double yoy. This is the right move for a company concerned with enjoying the long-
term barriers to entry of two-sided network effects. But it clearly dilutes short term profitability.
Three quarters of c. $180bn29 global experiences market is booked offline. A 5% share and 20%
commission would imply sales of greater than TRIP’s current total revenues. Yet experiences and
dining today are just 20% of TRIP’s revenues. Long-term we think it is reasonable to assume that
the dining and experiences segment can enjoy profit levels at least in line with the hotels
business: the customer base is more fragmented; as such attractions earn higher commissions
than hotel OTAs (c. 25%).
What kind of forest is priced into the share price today? TRIP expects to deliver double digit
EBITDA growth in 2019. Let’s assume the business can generate $470mn of adjusted EBITDA 30.
This is 8% of the current $6bn enterprise value. TRIP generated c. $350mn of reported FCF over
the last 12 months despite significant and accelerated investments in growing bookable supply in
dining and experiences. If we assume that the hotels business is generating a proxy for owner
earnings and normalise the profitability of experiences and dining to 40% EBITDA margins, this
would add another c. $80mn to post-tax owner earnings. If we also capitalise the investments in
TV advertising this would add a further $90mn after tax = $520mn. These owner earnings are
being generated on invested capital of less than $700mn, a 75% after-tax return, and a 9% yield to
the cash-adjusted market cap. This implies an inability for TRIP to ever grow its earnings.
A high internal reinvestment rate is the right capital allocation priority for intrinsic value
compounding. TRIP has been able to solve the chicken and egg problem associated with the
25 The first question on the Q&A session delved straight into this commentary. We do not think that this is a structural
change in TRIP’s relevance to its customers; auction pricing was stable in the quarter.
26 The earnings a business can generate after investing what it needs to preserve unit output and competitive
position.
27 Constant FX.
28 For example, TRIP acquired Bokun, an online booking management SaaS company which has been providing
services to attractions owners for free.

29 Phocuswright 2020 projected total bookings.
30 Deducting stock-based compensation of c $120mn and capitalising the expected TV advertising costs of $115mn in
2019.
41
development of strong platform business models by tapping into to its existing demand in hotel
and connecting it to acquired or developed supply in non-hotel. For both restaurants and
attractions, the number of reviewed items is multiples higher than the number of bookable
items, and we expect low penetration rates of bookable inventory to provide a long volume
growth runway.
GYM Group (GYM.LN): In March 2019 we acquired more shares of GYM at 190p/share, the same
price at which we originally acquired shares in July 2017, and in the process increased our equity
ownership quite meaningfully. Over that time the stock appreciated to 334p/share by August
2018, and subsequently fell back to our original purchase price. Over this same period our
estimate of the owner earnings of the business has appreciated materially, and our
understanding of the long-term opportunity to grow the company’s assets has not changed. As
such, GYM represents a larger weighting in our portfolio today than it did back in July 2017.
What were the trees in this instance? Management commentary around price reductions in
several gyms across their estate, together with the capacity expansion plans of low-cost gym
operator Xercise4Less prompted the publication of cautious/negative sell-side research
suggesting that market saturation and price competition are likely to erode future profits. In
addition, management’s comments relating to back-end loaded gym opening plans caused
analysts to reduce their near-term financial projections for the business 31.
But the bigger picture remains that there is both scope for addressable market expansion (still
one third of new LCG members have never been a member) and market share gains from
independent and multi-site private and public gyms. GYM might represent a quarter of LCGs
but is still just 2% of the total number of gyms in the UK. There are more council gyms in the UK
than private independent gyms, which are typically priced materially higher than their public
counterparts. We think it logical that this proposition gap shrinks as low-cost gym chains
become increasingly mainstream. It is not our expectation that GYM becomes a monopoly
operator of gyms; the implausibility of this as a bear case improves GYM’s investment merits as
we see them.
In July 2017, we estimated GYM’s owner earnings to be c. £23mn, a 9% yield to the market cap at
that time. In March 2019, our estimated owner earnings were c. £35mn, a 14% yield to the market
cap. Over this time the incremental returns being enjoyed on new site investments (> 20% after
tax cash on cash returns), and the capacity to reinvest mature estate cash flow into estate
expansion remain largely undiminished.
31 Anecdotally, but specifically relevant to the pricing power of this business, GYM recently increased my own
monthly membership from £21 to £23/month. I have been a member for three years; over that time the price for new
joiners has increased from £21 to £25/month.
42
New portfolio investment: Sea Limited (SE.US)
In March 2019 we acquired shares of Sea Limited at a price of $24 per share. SE is substantially
owned by insiders: Forrest Li, the founder CEO, owns 31% of the business 32 and Tencent owns
33%. All directors as a group own 44% of the company. Management has shown a preference to
direct efforts and capital to projects that they believe will create long-term value. SE operates
three platform businesses in gaming, ecommerce and digital payments, primarily in seven
Southeast Asian markets.
Garena distributes mobile and PC online games in its markets. Most games that Garena
distributes are done so exclusively. Garena also recently had significant success with its first
internally developed game, Free Fire, which was the fourth most downloaded game in the world
in 2018 and was available outside of SE’s core SE Asian markets, including Europe, LatAm and
Africa.
Shopee is an ecommerce marketplace which has adopted a mobile-first approach since its
inception in 2015. Shopee is a platform for connecting buyers and sellers of long tail products
across fashion, health and beauty, home and living, and baby products. Shopee provides tools
such as payment, logistics and fulfilment.
AirPay is a digital payments provider launched in 2014. Consumers can use the AirPay app as an
e-wallet to pay for products and services. AirPay is integrated into the Garena and Shopee
platforms.
All three business models are platforms which require investment to drive scale and barriers to
entry but have potential winner-take-most economic outcomes.
Garena’s network effects emanate from the social, multi-player nature of the games distributed.
Each new gamer increases the value of the platform for existing users. This dynamic might
suppress the cost of acquiring new users as the network grows in scale, as current users will tend
to invite new users to the platform. Strong and long-tenured developer relationships turn the
flywheel: Garena’s success in distributing games for local game players has facilitated
relationships with international game developers such as Tencent, Riot Games, Electronic Arts
and PUBG Corporation. This has allowed Garena to source high quality games from world class
developers, many of whom work as exclusive partners in SE Asia. Management is focused on the
virtuous cycle dynamics of attracting more users with high quality games, which attracts more
high-quality developers. The more users they have and the more games they distribute the better
they become at localising games, increasing their appeal to gamers and developers.
Shopee also has platform dynamics: as the number of buyers increases, Shopee attracts an
increasing number of sellers, resulting in increases in SKU variety available on the platform,
which increases the purchasing opportunities, and therefore monetisation potential, or value,
32 44% of the voting rights due to his ownership of B shares, carrying three votes each.
43
for each of those buyers. Shopee was the largest ecommerce platform in SE’s region in 2018 by
GMV and total orders. Shopee was also the most downloaded app in the shopping category in
Southeast Asia in 2018.
The long tail products that are the focus of Shopee’s marketplace support margins due to lighter
price competition vs. top-selling products. Ecommerce lends itself to long tail selling due to the
capacity for predictive analytics and personalised recommendations to stimulate liquidity in
niche markets. Sellers are supported through a network of payment providers and logistics
partners, integrated into the platform, as well as local teams to help sellers make use of Shopee’s
business management tools. Shopee provides a one stop shop allowing sellers to streamline
store setup, inventory and revenue management, delivery and payment collection.
The capacity for platform businesses to create substantial barriers to profitable participation may
be quite broadly understood. As is the economic characteristic that additional users in a
platform business model add more value for existing users. However, in addition, platform scale
strengthens the ability of the platform to offer white labelled goods and original content. Netflix
uses customer preference data collected over time to deliver not just a marketplace of products,
but original shows. Like Netflix, Garena intends to develop more original content while
distributing third party content. Based on customer intelligence, Amazon has c. 70 private label
brands, which were started in niche categories like batteries. Amazon’s marketplace dominance
allows it to scrape the content of user reviews and return feedback of third-party products and
use that information to create superior products that are more valuable to consumers.
Garena’s user base growth and engagement are driven by the launch of new games, the
expansion of existing games into new markets, and the improvement and launch of new content
in existing games. Southeast Asia is the fastest growing games market in the world; despite >60%
internet penetration, the region has the most engaged mobile internet users on the planet.
Garena organises hundreds of esports events annually and operates the largest professional
league in the region33. In 2019, the global esports economy will grow to $900 mn, a yoy growth of
38%. Three quarters of this will come from sponsorships and advertising. Media rights, tickets
and merchandise make up the remainder. Global esports audiences, currently c. 380mn people,
or 5% of the planet, have been growing in the mid-teens. 20% of the global population is now
aware of esports. This growth has been driven by streaming platforms such as Amazon-owned
Twitch, which attracts 15mn viewers per day each spending 100 mins per day watching live
gaming.
There is evidence to support the assertion that esports is becoming increasingly mainstream:
33Garena World 2018, which was held in Thailand in April 2018, had an attendance of 240,000, attracted 11mn views
online and more than 11,000 teams’ participation. Garena was also one of the organisers of the Arena of Valor World
Cup held in LA in 2018. In Garena’s markets, the Arena of Valor World Cup competitions attracted over 33 mn views
online.
44
• Personnel have been recruited from mainstream sports media. A few years ago, Activision
announced that it was forming a dedicated esports division, and it hired Steve Bornstein,
former CEO of ESPN and the NFL Network, to lead it.
• Broadcasting rights deals are being struck with Twitch as well as mainstream broadcasters
such as Disney and ESPN.
• Sponsorship is becoming more mainstream. Esports teams have traditionally been able to
pull in sponsors that are already closely associated with gaming e.g. from PC gaming
companies like Razer, computer-makers HP and Intel, to Toyota and T-Mobile.
• Employment conditions of players are formalising. Guaranteed contracts with minimum
salaries are becoming more common, and teams are investing in state of the art training
facilities, including coaches, chefs, dieticians, and sports psychologists.
• Esports is big enough to fill an Olympic stadium. The finals of the League of Legends World
Championship were held at the Beijing National Stadium. Esports are increasingly included
in the thoughts of Olympic Games organisers around the world. Esports were featured at the
2018 Asian Games as a demonstration sport and will be a medal event at the 2022 Asian
Games. Paris 2024 Olympic organisers were "deep in talks" about including esports as a
demonstration sport at the games.
Shopee’s capital allocation priority is to build marketplace scale and liquidity, and increasingly
on monetisation as GMV and market share continue to rise. It is this higher scale and liquidity
that increases the rate at which the business can monetise its assets. In management’s words in
the 2018 10K:
“We have made a strategic decision to invest in the growth of our Shopee marketplace by
incurring sales and marketing expenses in advance of our monetisation efforts. We believe that
taking a thoughtful approach to monetisation by building our user base and increasing
engagement first will allow us to maximise our monetisation in the future.”
And on the FY17 conference call:
“It's very clear in our mind, and it becomes clearer with every passing day, that almost all of our
markets are consolidating very quickly and more quickly than we would have anticipated that
even six or nine months ago. Secondly, as a matter of principle, when given the choice to ease
our spend and maintain our share or invest more heavily to expand our share, we've chosen the
latter strategy. Reason being, we believe that investment is going to help us achieve dominance
in the categories that are so important to us, female long-tail categories. That kind of dominance
and the ability to be the go-to platform for these important and very profitable categories as
we've talked about in the past should bring us to higher monetisation levels going forward. So
really, just to conclude, at the end of the day, winning a merchant or a customer today in our
mind is much better than having to spend more to win them in the future.” The emphasis is
ours; it seems consistent with a capital allocation objective to maximise total long-term value for
owners and reflects a capacity to suffer that is a desirable quality for such long-term owners.
Ecommerce penetration is materially below global averages in almost all Shopee’s markets, but
ecommerce and m-commerce engagement in Indonesia, Shopee’s largest market representing
45
almost half of GMV, is the highest in the world. The GMV of the internet economy is 2.8% of SE
Asia’s GDP in 2018, up from 1.3% in 2015 — and is projected to exceed 8% by 2025. SE Asia is
almost 10 years behind the U.S., in which the GMV of the internet economy was 6.5% in 2016.
The monetisation of Shopee’s customer is improving, driven by (1) higher take rates increasing
the gross profitability of transactions on the platform, and (2) falling shipping subsidies driving
sales and marketing leverage. Shopee’s take rate (revenues/GMV) is currently suppressed by
efforts to build scale and market leadership, entrenching the network effects’ barriers to entry of
the business. But the take rate has been increasing and management expects this to continue
through a combination of commissions, advertising fees and value-added services. Outside of
Taiwan, Shopee charges zero commissions. In Shopee’s most competitive market, Singapore,
peers are charging between 3% and 30% commissions. Qoo10 charges 8-12% seller commissions
in Singapore. 11street charges between 3% and 12% commissions rates to sellers in Thailand and
Malaysia. In fashion, Shopee’s largest category, sellers are charged 12%. Lazada charges on
average 6.5% commissions plus a 2% payment fee in its markets. In fashion, sellers are charged
12% including the payment fee. Naturally the direct profitability of Shopee’s business is driven by
the take rate that Shopee can command on the GMV passing through its platform. In 2017
Shopee’s take rate was practically zero and its gross margin was -125%. In 2018 the take rate was
2.8% and the gross margin was -65%. In 4Q18 Shopee’s take rate was 3.7% and its gross margin was
-52%. Even if we can expect diminishing marginal returns to higher take rates, the capacity to
increase the take rate bodes well for the potential gross profitability of future transactions.
Shopee has heavily subsidised the cost of shipping for its sellers in order to build supply scale.
The extent of these subsidies has been declining without any noticeable impact of GMV growth,
resulting in sales and marketing expenses declining as a percentage of revenues. Shipping
subsidies declined qoq in absolute dollar terms in 4Q18 as free shipping is now only available on
baskets of a minimum size. This marketing efficiency improvement was achieved while
GMV/orders grew 27%/31% quarter-on-quarter. Sales and marketing as a percentage of GMV for
Indonesia, Shopee’s largest market, was lower than the ratio for Shopee as a whole, supporting
the contention that scale drives operational leverage with respect to the cost of acquiring new
customers. Management expects sales and marketing expenses to decline in absolute terms from
here, signalling a potential inflection point in ecommerce profitability.
What are the counterarguments to our conclusion that this is a high-quality business with
avenues for profitable redeployment of capital? Negative experiences spread quickly online, and
SE’s business success is driven by customers’ trust in the platform. Customers must believe they
will be protected in order to transact safely online. As the number of connections and
transactions grows exponentially there is a risk that this additional complexity renders SE’s risk
control measures inadequate. This might lead to negative network effects as one or both sides of
the platform are driven away by unpleasant experiences. Shopee verifies sellers, screens listings
and has teams dedicated to dispute resolution. Shopee also offers a ‘Shopee Guarantee’ under
which buyer payment is held by Shopee until delivery of the goods, reducing settlement risks
and encouraging buyers to purchase online. The huge increases in active user engagement
46
across a variety of online services suggests consumer comfort with transacting online is
increasing.
Ecommerce competition may inhibit user monetisation. Shopee may not be able to adequately
monetise the transactions taking place on its platform. So far KPIs relating to the company’s
monetisation progress are moving in the right direction. Take rates are increasing and subsidies
are reducing without harming the company’s asset growth.
Changing gaming tastes. Garena has three- to seven-year agreements in place with multiple
developers. It can use its experience as a distributor of games developed by others to improves
the prospects for success in its own internal development ambitions. Free Fire is a short, but
encouraging, piece of evidence that they may be able to do this successfully.
Assuming, for now, that AirPay is worth nothing, what is the implied valuation of Shopee for a
range of sustainable growth hypotheses applied to Garena’s profits?
If we assume that Garena can never grow its EBITDA, an investor with a 10% opportunity cost of
investing might be willing to pay 10x EBITDA, or c. $5.5bn to earn this opportunity cost. This
would imply a valuation of Shoppe of $3.3bn, one third of SE’s current enterprise value of $8.8bn
and 5x Shopee’s expected sales this year.
If we value Garena at 16x EBITDA, or less than 4% sustainable EBITDA growth, both Shopee and
AirPay are priced by the market as worthless. The implied value of less than $9bn for Garena
would seem conservative given the business’s demonstrated profit growth and potential.
If Shopee and AirPay burn through SE’s $2bn cash pile but fail to make progress in
demonstrating the ultimate path to sustainable profitability, we might be willing to pay $9bn to
own SE’s equity, c. 18% downside from today’s market cap of $10.8bn.
What assumptions does an owner of this business need to make to render his interest worthless?
We could assume that Free Fire is a one hit wonder, and therefore Garena’s profitability shrinks
by some 30%, and never grows again. We might also need to assume that the company’s cash
balance of $2bn is used to fund investment in AirPay and Shopee which earns zero return, and
that in addition SE continues to burn cash at the current rate of $750mn per year for the next five
years (undiscounted). There is evidence to suggest that this is not an appropriate set of
assumptions: Shopee’s take rates are improving; Shopee’s sales and marketing leverage is
improving; and Shopee’s user, order and transaction growth remain strong.
If Garena can sustainably grow nominal EBITDA in line with real GDP growth in the region of c.
5%, AirPay is worth nothing and Shopee is valued at 5x 2019 revenues (making no adjustments to
revenue to reflect the potentially supressed take rate), SE’s equity is worth c. $16bn, 50% higher
than the current quoted market cap.
If Garena can sustainably grow nominal EBITDA modestly more than real GDP growth in the
region of c. 5%, AirPay is worth nothing and Shopee is valued at 10x normal revenues, adjusting
47
the estimated take rate from 4% to 10% in line with competitors in the region, SE’s equity is
worth c. $37bn, 240% higher than the current quoted market cap.
Private market transactions also suggest Shopee could be materially undervalued. Tokopedia, a
C2C business, like Taobao or eBay, but operating only in Indonesia, raised capital in 2018 which
valued the enterprise at $7bn, c. 1.5x the estimated GMV. Flipkart, the dominant Indian
ecommerce business, was acquired by Walmart at a $21bn valuation in 2018, implying an
EV/GMV of 2.8x. If we apply a 4% take rate to management’s adjusted revenue guidance of $630-
660mn, Shopee could be processing c. $16bn of GMV this year. At 0.5x EV/GMV Shopee is worth
the entire enterprise value of Sea Limited.
Thank you for entrusting me with your capital.
With my best wishes,
Mark
48
A letter to partners, June 2019: the rise of the platform and
implications for owners
Dear partners,
Many of our holdings are platform businesses or linear businesses with underappreciated
emerging platform elements. We discuss below the rise of platform businesses, the challenges
and benefits that are special to platforms, and the implications for equity owners.
A history of business models
Since the industrial revolution, when the invention of machine tools and the development of
steam power marked the transition from hand production to mechanised manufacturing
processes, the industrial landscape has been characterised by linear business models. Linear
businesses take inputs such as physical production assets and labour to make goods and services
that are sold to customers, according to a linear, unidirectional value chain. Think car
manufacturers, steel makers and consumer goods companies. Distributors such as Walmart are
also linear business models.
In many cases technological advancement has not changed the linear aspect of these companies’
value chains. But in several industries the internet has dramatically lowered distribution costs,
permanently impairing capital in those industries that enjoyed economies of scale in physical
distribution. Examples include travel agents, newspapers, encyclopaedias, bookstores, video
rental and shopping malls. Businesses with high fixed costs of incremental production faced
competition from companies enjoying negligible marginal costs of internet distribution.
The transition that many software companies have made to recurring revenue SaaS models has
been driven by the opportunity to lower the marginal cost of distribution 34 and increase
incremental returns on capital. But SaaS companies still operate a linear supply chain. Linear
businesses create products and services using their internal resources and means of production.
These internal means of production are a constraint on the sustainable growth of the business.
What is a platform?
A platform is a business model that facilitates interactions and the exchange of value between
producers and consumers of something that is not produced by the platform itself. Rather than
creating value by making and selling things, platforms provide a mechanism for exchange. As
such, platforms are much less capital- and labour-intensive than linear businesses. Platforms
generate revenues by capturing a share of the transaction value their network enables.
Producers in a platform accept an additional cost of unit production in return for the larger
34 Software is distributed over the internet, obviating the requirement to build and maintain physical servers.
49
volume opportunity that a platform affords35. As such platforms’ breakeven points tend to be at a
larger scale than linear businesses.
The economic characteristics of platforms
While the internet has lowered the costs of distribution for many businesses, platforms also have
very low marginal costs of production, driving superior unit economics relative to linear
businesses. Access to these superior economics is clearly desirable, but they are elusive. Building
a successful platform necessitates solving a chicken-and-egg problem of having enough
producers to attract consumers and vice versa. Without adequate scale and liquidity, the value to
users will be lower than the cost of participation. It is this dynamic that makes platform
businesses more difficult to build from a standing start vs. their linear counterparts. Once past
the tipping point at which user value is greater than the cost to participate, network effects can
take hold to rapidly increase the per-user value of the network36.
In addition to network effects, successful platforms benefit from producer switching costs.
Consumers provide non-monetary benefits to producers in the form of reviews, ratings and
other feedback such as likes and shares. This is an important mechanism encouraging good
behaviour and preserving required levels of customer service37. As producers build up this
goodwill asset, the costs of leaving the network and establishing a reputation from scratch
increase.
The challenges in building and protecting valuable platforms
One of the principal risks associated with linear

business oligopolies is antitrust. The objective
of antitrust is to promote fair competition for
the benefit of consumers. Traditional
monopolists control the means of production.
They earn supernormal profits by restricting
output, charging anticompetitive prices and
minimising consumer surplus. Platform
businesses do not own the means of
production, but simply the means of
connection. Platforms do not control the prices
and outputs of producers. But by facilitating
more touchpoints between consumers and producers they can increase competition, to the
35 In addition to tools of production and distribution a platform may provide, such as Opentable’s reservation
management software, or Amazon’s fulfilment services, effectively swapping upfront capital commitments for
variable costs of doing business.
36 A network effect is present when the behaviour of one user impacts the value of a platform to other users
(positively or negatively).
37 Uber drivers are required to maintain at least a 4.6 star rating on their most recent 100 trips to stay on the app.
Persistently low scoring passengers are now also being banned:

https://www.theguardian.com/technology/2019/may/31/uber-to-ban-riders-with-low-ratings
50
ultimate benefit of the consumer. The goal to increase platform scale benefits both its owners
and users. Platforms are therefore characterised by an alignment of interest between
shareholders and customers that is not the case in traditional linear monopolists. A regulatory
intention to curb a platform’s dominance could have the undesirable effect of harming
consumers’ interests.
Platforms make fragmented supply available to consumers in one place. Consider consumers
wishing to find the best price for a hotel room. There is an opportunity cost associated with
manually finding the owners of that inventory and comparing availability and pricing.
Conducting that price comparison exercise on TripAdvisor is a more efficient process. And price
transparency leads to a more competitive outcome for the consumer. The platform increases
choice and saves the consumer time and money. And the more scale the platform has the greater
the choice, and the more competitive the price. In this way platforms grow by delivering more
value to their users, rather than by controlling the supply chain. However, while these academic
considerations may be logical, until the regulatory regime evolves to be suitable for 21st century
business models, incumbents will use arguably unfit-for-purpose antitrust frameworks to make
life difficult for dominant platforms such as Airbnb and Uber.
Implications for owners
Winner-take-most economic outcomes 38 imply a race for scale, leading to potential battles for
dominance. If access to financing is inadequate, or efforts to monetise the network are
premature or clumsily executed, the platform may not acquire the requisite scale to enjoy the
fruits of market dominance. With supportive owners emerging platforms can avoid the pressure
to generate short term profits to maximise total long-term value creation. eBay entered China in
the early 2000s and made the mistake of charging commissions to sellers immediately. By
keeping Taobao free and by allowing buyers and sellers to communicate with each other
(unencumbered by eBay’s desire to prevent transactions outside of the network), Alibaba was
able to grow its relevance and benefit from stronger network effects. Alibaba also introduced
Alipay, which held customer cash in escrow accounts until buyers verified product receipt. We
see parallels with the way in which Sea Limited is growing its Shopee ecommerce platform in
Southeast Asia. It is adopting a patient approach to monetising its network 39 in order to build
non-replicable platform scale and liquidity. It is allowing suppliers and consumers to
38Examples: smartphone operating systems, travel OTAs, web search, ride sharing, social media.
39Shopee’s take rate (revenues/GMV) is currently suppressed by efforts to build scale and market leadership,
entrenching the network effects’ barriers to entry of the business. As the number of buyers increases, Shopee
attracts an increasing number of sellers, resulting in increases in SKU variety available on the platform, which
increases the purchasing opportunities, and therefore monetisation potential, or value, for each of those buyers.
51
communicate with each other40, and is overcoming trust hurdles with the Shopee Guarantee 41.
Shopee is also lowering the friction of ecommerce by integrating its AirPay business42.
Appraising addressable markets when the future ≠ the past
Aswath Damodaran, a respected finance professor at the Stern School of Business and
considered an academic authority on company valuation, wrote a paper43 in 2014 in which he
dismissed the private business value of Uber at the time of $17bn. By making assumptions about
Uber’s potential market share and addressable market, which he defined as the global taxi
industry, Professor Damodaran estimated Uber’s value to be $5.9bn. Bill Gurley, a partner at
venture firm Benchmark, and Uber investor, published an article 44 in response in which he
contested that Damodaran had vastly underestimated Uber’s addressable market. Gurley argued
that Uber’s potential market is greater than the taxi market due to a superior experience vs.
traditional taxis (cheaper, shorter pick-up times, more convenient payment methods, greater
safety and trust). As such the relevant addressable market was not just the taxi industry, but
included public transport, walking, rental cars and driving owned vehicles. Today Uber’s market
cap is $75bn, more than four times the company’s private business value five years ago. The
outsized return on investment for Uber’s owners was made available in part because of the
market’s failure to look beyond the incumbent industry in determining the prospects of those
disrupting it. Airbnb is another prominent example of a platform which has expanded the
addressable market for lodgings; it has opened an entirely new source of supply (spare rooms
and empty homes), therefore growing largely through addressable market expansion rather than
market share take45.
40 Shopee employs a seller rating system and has a live chat function which allows buyers to ask questions of sellers.
41 Buyer payment is held by Shopee until delivery of the goods, reducing settlement risks and encouraging buyers to
purchase online. Shopee also verifies sellers, screens listings and has teams dedicated to dispute resolution.
42 AirPay allows customers, including those without bank accounts, to process payments on Shopee as well as third-
party merchants for products and services such as food, transport, telecoms and paying utility bills. Consumers do
not need a credit card or a bank account as the AirPay App accepts account top-up payments in cash through physical
AirPay counters located across the region (a ‘reverse ATM’). For consumers with a bank account, AirPay is connected
to almost all major banks.
43 https://fivethirtyeight.com/features/uber-isnt-worth-17-billion/
44 http://abovethecrowd.com/2014/07/11/how-to-miss-by-a-mile-an-alternative-look-at-ubers-potential-market-
size/
45 Studies have shown that Airbnb volume has not been correlated with trends in hotel bookings or revenues per
available night.
52
So addressable markets can be
underappreciated due to a rigid appraisal
of incumbent demand. They can also be
underappreciated when there exists the
potential to leverage the strength of the
network to invest into adjacencies. Uber
is leveraging the supply side of its
network to expand into food delivery
with UberEATS. TripAdvisor is
leveraging the demand side of its
network to expand into attractions and
restaurants. TripAdvisor has a rare
opportunity to circumnavigate the chicken and egg problem associated with the development of
strong platform business models by tapping into to its existing demand in hotel and connecting
it to acquired or developed supply in dining and experiences. For both restaurants and
attractions, the number of reviewed items is multiples higher than the number of bookable
items, and we expect low penetration rates of bookable inventory to provide a long volume
growth runway. TripAdvisor is leveraging existing behaviours: people already book tickets to
museums or visit famous landmarks and make restaurant reservations. But TripAdvisor is
investing in an opportunity to reduce transaction friction by making it easier for customers to
find, review and book restaurants and attractions. In doing so it can bring vast untapped supply
into its existing travel ecosystem. Crucially, this supply is large 46 and fragmented, making its
aggregation within a platform valuable.
Valuing platforms
Perhaps the single biggest potential source of platform business undervaluation is the market’s
application of linear business prospects to a set of outcomes which could include exponential
value growth47. Linear businesses’ customer acquisition efforts result in singular relationships
per customer added. Recurring revenue business models may lower the cost of customer
acquisition and retention, but they still result in a singular profit stream associated with each
customer. When platforms acquire users, whether they are customers or producers, those users
form relationships and transact with the existing users in the platform. As the number of users
grows arithmetically, the number of potential relationships and interactions grows
exponentially. And due to the very low marginal costs of distribution and production, profit
margins can expand materially.
46 Phocuswright estimates the total attractions market to be c. $180bn, three quarters of which is booked offline.
47 Metcalfe’s Law states that the value of a network is proportional to the square of the number of connected users.
While this law typically applies to single-sided networks, the same principle applies to platforms, or two-sided
networks. There are also important shortcomings of Metcalfe’s Law, such as the assumption that all connections in a
network are valued equally. But network effects are local: Facebook’s utility is increased more by the addition of
friends than strangers. Additional Uber drivers within an existing user’s postcode are more valuable than those in
another country.
53
The application of linear growth expectations to platform businesses therefore implies a failure
of the platform to facilitate, nurture and monetise these potential interactions. Once we
understand the dramatically lower cost of growth for these businesses, we should question the
relevance of previously accepted valuation methods. Over the last decade Mastercard has
compounded its earnings per share at 20% pa, and its share price has grown 30% pa. Ten years
ago, Mastercard was available for sale at a mid-teen multiple of its per share earnings. An investor
could have paid 80 times the company’s after-tax profits and still achieved an opportunity cost of
investing of 10% per year for the next decade. The availability of this opportunity to public equity
investors implies a dramatic underappreciation of the exponential growth properties of
successful platform companies.
54
Sotheby’s: one of the world’s most enduring platforms
We initially acquired our ownership of Sotheby’s in May 2016

for $28.7 per share. In June 2019 Patrick Drahi announced the
takeover of the company for $57 per share and we sold our
entire interest for $55.5 per share. The annualised gain of our
exit price over our initial entry price was +24% per year,
consistent with the annualised returns of the portfolio over that
period. However, our ownership of Sotheby’s has been the
single largest contributor to Tollymore’s cumulative investment
results. Portfolio management is responsible for this outsized
contribution. This is a pleasing outcome given our efforts to
create an environment that allows us to execute a long-term
investment programme. Specifically, we have benefited from a
set-up that has allowed us to use share price volatility to lower
our average cost of ownership. In doing so we have shrunk the
risk of permanent capital loss. Assuming an appropriate
investment temperament and aligned investment partners, volatility is inversely correlated with
risk. The volatility of publicly quoted securities drives our strong preference for public vs. private
market investing in the pursuit of acceptable unleveraged investment results. Sotheby’s stock
volatility has been driven to a greater degree by large changes in its earnings’ multiple rather
than the variability of those earnings, potentially reflecting poor aggregate understanding of or
conviction over what this company is worth on a long-term, through-cycle basis. Sotheby’s
business model embodies a wonderful combination: the provision of enduring services with very
high barriers to profitable participation on the one hand, but the delivery of services which are
subject to seasonal, cyclical and otherwise unpredictable 48 fluctuations on the other. The former
allows us to benefit from the intrinsic value compounding of the underlying asset, while the
latter affords us the scope to earn an equity return superior to that rate.
Sotheby’s is a 275-year-old global art business whose operations are organised under two
segments: Agency and Finance. The Agency segment earns commissions by matching buyers
and sellers of authenticated fine art, decorative art, jewellery, wine and collectibles through an
auction or private sale process.
The Finance segment earns interest income from the provision of loans secured by works of art.
This segment offers advances secured on consigned property as well as general purpose loans
secured by property not presently intended for sale (<50% LTVs). Many traditional lending
sources offer conventional loans at a lower cost to borrowers than the average cost of Sotheby’s’
loans. Many traditional lenders also offer borrowers integrated financial services such as wealth
management, which are not offered by Sotheby’s. Few lenders, however, are willing to accept
48Consignments often become available as a result of the death or financial/marital difficulties of the owner. These
factors cause the supply and demand for works of art to be unpredictable.
55
works of art as sole collateral; they lack access to market information required to effectively
appraise underwriting risk.
There is evidence of the presence of lasting unfair business advantages. Excluding the finance
business, the core auction platform has generated historic returns on invested capital of 25-35%.
Sotheby’s and Christies have had high and stable market shares for extraordinarily long periods
of time. Sotheby’s has pricing power; it charges sellers a c. 10% commission and a buyer’s
premium of 14-25% depending on the hammer price. However, in order to win larger or more
prestigious consignments, Sotheby’s and Christie’s will often waive the seller commission (and
sometimes even share the buyer’s premium with the seller). This is effectively an investment in
brand prestige and can be considered part of the cost of acquiring and retaining buyers and
sellers. The finance segment is a form of captive financing which has earned mid-high teens
returns on equity.
What are the sources of this moat? Sotheby’s business model enjoys barriers to entry facilitated
by a duopoly competitive structure, two-sided network effects, a global brand that lowers search
costs and elicits a higher level of trust and willingness to pay, and non-replicable heritage. Brand
and reputation are arguably more valuable assets in the art market vs. other industries due to the
subjective, difficult-to-estimate value of the products being sold. Sotheby’s duopoly with
Christie’s should allow the business more freedom to set pricing to maximise volume*price =
revenues.
Consignors are often the executors of estates who are responsible for monetising illiquid assets
with opaque and objective valuations. They typically have not consigned before and therefore
their risk-aversion attracts them to a global brand with a multi-century history and access to the
highest quantity of relevant potential buyers. That is: unrivalled platform scale.
The finance segment is an important ancillary business for Sotheby’s. It helps to secure and grow
client relationships. Sotheby’s has a clear collateral liquidity and underwriting advantage vs.
traditional lenders unable to take artwork as loan security.
Most global art transactions are private sales, but Sotheby’s and Christie’s account for more than
a third of global auction sales, and 80% of auction sales priced higher than $1mn. Smaller
competitors such as Phillips and Bonhams have been unable to weaken the dominance of the big
two and penetrate the market for the most valuable works. Changes in the buyer’s premium
enacted by one auction house have typically been followed in both directions by the other.
Arguably Sotheby’s’ long-term value creation has been lower than one would expect from such a
demonstrably wide moat business. This could be due to value accruing to executives rather than
shareholders and a reluctance from prior management to make long-term investments into the
business, and/or excessive discounting in order to win high profile consignments. One could
argue that Sotheby’s’ stagnant profits over the years reflect a failure to exploit the considerable
increase in the number of individuals with the financial means to purchase expensive art.
56
Relatively flat global art market sales over the last decade seem at odds with the trebling of
wealth owned by the globe’s billionaires over the same period.
Since Tad Smith has taken the helm in 2015 the company has made good progress on capital
allocation (scrapping the dividend and making significant stock repurchases at depressed prices)
and on investing in digital capabilities and product adjacencies to improve overall SVA of this
business. However, little ground seems to be made on employee cost efficiencies and it is
disappointing that the company has altogether stopped reporting some of the metrics relevant
to these goals49.
When we initially invested, we identified some opportunities to improve return on operating

expenses: travel and entertainment costs were $17k per employee, ‘professional fees’ were more
than $50mn pa, the average employee salary was almost $190k, and management outlined its
expectation for mid-single digit declines in headcount. Unfortunately, there is limited financial
evidence to suggest that any headway has been made in lowering the fixed costs of doing
business. Headcount has continued to increase, average employee costs have risen by 2% per
year, and Sotheby’s stopped disclosing several of the above metrics from 2Q 2017.
Nonetheless, there is evidence to suggest that the future economic value that Sotheby’s will
generate for owners will be larger than historically has been the case due to (1) the development
of ancillary revenue streams outside of the core auction, and (2) the digitisation of the delivery of
Sotheby’s’ services.
Ancillary revenue development: Private sales are a low proportion of Sotheby’s revenues today.
In 2018 Sotheby’s generated $1bn of private sales, a 75% increase in two years, but still just 16% of
consolidated sales. Private sales may be preferred by consignors wishing to preserve anonymity,
seeking to minimise costs of the sale50, to test an aspirational price hidden from the public eye or
to avoid the publicity of a failed sale. The advantage of an auction is the competitive bidding
tension created, and the potential for auction dynamics to lead to irrational bidding strategies to
the ultimate benefit of the seller51. Endowment effect52, social proof53 and scarcity54 are all at play.
There may be an opportunity for Sotheby’s to leverage its relationships with consignors and
buyers, and its access to data and valuation expertise, to capture a higher share of private sales,
still the dominant mechanism for settling art transactions. CEO Tad Smith has outlined his goal
of building enough understanding of customers’ preferences to allow Sotheby’s to offer
unsuccessful auction bidders the opportunity to buy similar items via a private sale within 24
hours of the auction. Sotheby’s recruited David Schrader from JP Morgan in 2017 to lead a
49 E.g. incentive pool as a percentage of EBITDA.

50 Private sale competition such as galleries is more fragmented that the duopolistic auction industry structure.
51 Half of eBay auctions result in higher sale prices than the “buy it now” price.
52 The auction catalogue plays a crucial role here. This forms part of the buyer’s psychological journey as he imagines
owning the item in advance, in doing so placing a higher value on it.

53 Other bidders’ actions irrationally inform our view of the value of the object to us.
54 Auction items are unique (there can only be one buyer), and scarce in time (the process is not repeated).
57
dedicated team with the goal of growing private sales business. We would suggest that Sotheby’s
is well positioned vs. private dealers and art galleries in facilitating private art transactions
between buyers and sellers given the breadth of its customer relationships, collection of
transaction data and ability to offer advisory and financing services.
There are opportunities outside of art too. Sotheby’s already sells furniture, jewellery, wine, cars
and watches, and is expanding further. These items are often the gateway into the purchase of
works of art.
Service digitisation: The digital delivery of services could make the

art market more accessible and lower the costs of doing business.
The company is seeing record numbers of new bidders, two thirds of
whom are bidding online. Online only sales are generating just 4% of
total revenues but are responsible for one third of new customer
additions to the company’s database. Art ecommerce growth
materially lags retail ecommerce growth. Digitisation is
strengthening the network and is creating the potential for an
acceleration in revenue and profit growth by increasing the
accessibility of Sotheby’s’ services and the competitiveness of its auctions of unique objects
(fixed supply and growing demand). Sotheby’s’ capacity to access untapped international
demand also improves as it increasingly conducts its business online; three quarters of revenues
are generated in NYC and London, yet art has global appeal.
The slowness of art sales to migrate online over the years may be a function of the presence of
asymmetric information and value of physical inspection. These do not seem to be surmountable
challenges for a business with Sotheby’s’ credibility, valuation advisory resources and
increasingly technological solutions to verifying authenticity. The opportunity is reflected in a
world in which millennials represent one third of HNW collectors.
Digital service delivery lowers physical venue dependency, requires less travel and less paper
processing including catalogue production. Technology lowers the labour intensity of art
authenticity evaluation, sourcing consignments55, matching buyers and sellers, and valuation.
Capital allocation and incentive structures are improving. Tad Smith reduced the share count for
the first time in ten years. Management incentives are now linked to ROIC, a KPI that was not
even mentioned in prior proxy materials. Smith’s future stock awards were linked to high stock
price hurdles. The share price had to at least reach $85 by 2020 for him to receive the maximum
award.
Competitive discipline seems to be improving. Inventory sales fell by half in 2018, generating
most of the operating margin improvement from 16% to 19%, and perhaps suggesting that the
company has been more disciplined in its use of auction guarantees. When we initially acquired
55 Increasingly through an online consignment platform.

58
shares in 2016, margins had collapsed driven partly by a spike in auction guarantees, the poor
judgement of which led to a loss on inventory sales.
Management has shown a willingness to invest for the long-term. Management has
demonstrated an understanding of the company’s unfair business advantages and has deployed
capital to strengthen these. Acquisitions over the last three years include:
• The Mei Moses Art Indices, recognised as the preeminent measure of the state-of-the-art
market. Previously publicly available, this is now an internal asset.
• Orion Analytical, a scientific research firm which has developed technologies and
procedures to verify authenticity.
• Art, Agency Partners, a consulting firm advising art collectors on portfolio strategies and
philanthropic matters connected with the arts.
• Thread Genius, an AI start-up which has developed algorithms to identify objects and
recommend similar images to viewers. Sotheby’s has been developing an object database
which is designed to physically locate objects of interest and match buyers and sellers.
In 2018, Sotheby’s was able to spend $300mn repurchasing shares, almost triple capital
investments, and lower its leverage ratio. It did this while delivering a 24% return on
shareholders’ capital. This is a very good business.
We also thought it was a very cheap business. Sotheby’s trailing 12-month EBITDA of c. $200mn
is broadly in line with its 10- and 15-year averages, an 8% yield to the quoted enterprise value
prior to the acquisition. This was the price of a business with a wide, and widening, moat, for
which the costs of doing business could decline, for which corporate governance and incentive
structures have improved, and with several paths to profitable revenue growth, including the
digital democratisation of art increasing auction competitiveness, and taking market share in
private sales.
Final thoughts
Tollymore is a multi-decade learning effort. Striving for self-awareness and sound judgement
guides our decision making every day. We are grateful for our relationships with intellectually
generous peers and investment partners; these help to compound knowledge and improve our
chances of becoming just a little bit wiser together, day by day. Thank you for your partnership.
Mark
59
A letter to partners, September 2019: incentives, wealth, and
happiness
Dear partners,
Tollymore’s mission
Tollymore’s first investment letter to partners stated that its objectives were to (1) create value in
an industry that in aggregate transfers value from investors to managers, and (2) create a flexible
lifestyle in which there is no distinction between work and play. In a way in which business and
investment success are determined by productivity, relationships, serendipity, and personal
energy. It is this flexible way of living which facilitates a commitment to the happiness of
Tollymore’s principals, their family, and the delight of the firm’s investment partners on a multi-
decade basis.
The goal here is to secure sustainable freedom. Underlying such freedom is the permission to
make decisions in the interests of creating value for those whose lives Tollymore influences. That
is: the investment partners of the firm, and the friends and family of those individuals who are
the firm.
“The really important kind of freedom involves attention, and awareness, and
discipline, and effort, and being able truly to care about other people and to
sacrifice for them, over and over, in myriad petty little unsexy ways, every day.”
David Foster Wallace
Is the accumulation of personal wealth consistent with this goal? To the extent that personal
financial progress may allow one to free up more time to devote to the interests of the
meaningful people in our lives, it may seem like an important ingredient in achieving Tollymore’s
goals. Yet there is ample, hardly undiscovered, evidence questioning the correlation between
wealth and happiness56.
56Gallup/World Bank correlation data, Happiness, income satiation and turning points around the world, Princeton
study
60
Specifically relating to the investment
management endeavour, a money manager’s
personal objective to maximise her own personal
wealth may be harmful to the prospects of
creating the most aggregate dollars of value for the
investment firm’s owners and investors. This is
because investment management, in its most
widely practised form, is extremely scalable. In its
most widely practised form, there exist strong
incentives to gather assets, scaling revenues
without commensurate increases in the costs of dispensing fiduciary responsibilities and
exercising sensible investment and business judgement. These incentives include managers not
invested in the funds they manage, high management fees, no performance fees, and opacity
regarding the appropriate benchmark for complex strategies, and therefore an inability to
measure the manager’s capacity to create value. An asset growth imperative is compounded by
large and expensive teams and bloated, gold-plated cost structures designed to convey
institutional suitability.
More generally, wealth can be a distraction. This observation places no judgement on how
financially successful people choose to enjoy the fruits of their success. But such enjoyment may
create a conflict of interest with those who helped create it, and to whom there remains an
ongoing responsibility to deliver acceptable results. An investment manager may become
distracted by the enjoyment or redistribution of the prosperity he has acquired as the owner of
an investment management business. This may be through the obvious channel of consumerism
– yachts, private jets, property and luxury travel acquired in increasing doses via hedonic
adaptation – but also through more laudable, philanthropic avenues. This is not an opinion of
the devotion of capital to charitable projects. But the commitment of time to philanthropic
causes could be a distraction from the business of money management. This industry is hard
enough for those dedicating their entire working existence to the pursuit of investment
outperformance.
This is not to say that wealth is an undesirable outcome, nor that its pursuit is not worthwhile.
Indeed, the objective to create substantial personal wealth may well contribute to higher life
satisfaction. But (1) the benefit of wealth as an outcome is often overstated, and (2) the pursuit of
individual wealth may be at the expense of more likely contributors to happiness. Addressing
these in turn:
61
Process vs. outcomes in the wealth-happiness correlation
There is evidence to suggest that the method of wealth

acquisition in an important determinant in the happiness
associated with such wealth. People whose prosperity has been
earned, vs. won or inherited have report higher levels of life
satisfaction57. There are also studies to suggest that altruism is
linked to more enduring versions of reported happiness than,
for example, the purchase of experiences. That is, people who
are emotionally and behaviourally compassionate enjoy greater
well-being, happiness, health, and longevity58. One of the
reasons for this is the connection to others altruistic behaviour
empowers. This self-serving benefit of empathetic behaviour
highlights the question over whether true altruism exists. The
observation that altruistic acts are self-interested was most
profoundly underlined by one of the greatest philosophers of the 1990s: Joey from Friends. After
Phoebe deliberately allows herself to be stung by a bee “so it can look cool in front of its bee
friends”, Joey explains that the bee probably died as a result, and concludes that “there are no
selfless good deeds”.
An extension of these observations might suggest the accumulation of personal capital through
a working life dedicated to creating value for others vs. transferring value from others, is more
likely to lead to greater happiness. We seek “non-zero-sumness”, or symbiotic value chains, in
the small selection of companies in which we invest, and in the relationships with the small
number of investment partners whose capital we manage alongside our own.
The role of relationships in leading a good life
“A good life is a life that feels good to live, a worthwhile life of progress,
achievement, enjoyment and engagement – and above all, with good
relationships at its heart.”
AC Grayling
One of the most treasured virtues of Tollymore is the freedom it allows its principals to think
independently. Nonetheless, humans are social animals; we function better in communities
rather in isolation. Studies suggest our mental health and well-being are connected to the quality
57 This is known as the IKEA effect. In this context, it is a cognitive bias worth seeking out, rather than avoiding. HBS
IKEA effect
58 International Journal of Behavioural Medicine, 2005, Vol. 12, No. 2, 66–77
62
of our relationships59. This involves nourishing valuable connections and minimising or
removing negative associations.
Tollymore is a multi-decade effort to build and sustain trusted relationships with long-term
investment partners. Great investment partners are a competitive advantage: an evolving,
mutually appreciative and increasingly resilient relationship is an enabler of greater and more
sustainable value creation.
“An important dimension of successful investment management organisations

– true of great organisations in many fields – is having great clients. If you
have clients you do not enjoy or admire, or clients that do not expect much of
you, you should seriously consider terminating the relationship with them.
They will hold you back. If you have great clients, wonderful clients, reach out
to them and ask them to demand even more of you. The great role of the
client is to challenge you to be the very best that you can be.”
Charles Ellis, The Characteristics of Successful Investment Firms
There is a natural alignment between an investor with a long-term, possibly perpetual,

investment horizon, and a younger, emerging, investment management organisation. As the
manager ages and the fund matures, this alignment may be challenged if the manager’s personal
investment objectives become more influenced by preservation vs. growth. Assuming good
health and sound mind, Tollymore’s investment runway is 30-40 years. But a recognition of this
potential future misalignment is important. As is a willingness to work with investment partners
whose missions outlast the biological constraints of the investment firm’s owners, in order to
think about succession or other safeguards.
Several aspects of Tollymore’s investment and business process are designed with this goal to
engender meaningful and valuable relationships in mind:
• A balance sheet and a cost structure that enables us to adopt a patient approach to
building these relationships.
• Transparent and candid investor communication acts as a filter for investors who can
think unconventionally and act countercyclically.
• Incentive structures which make it more sensible to forgo additional management fees in
lieu of excess returns on insider capital.
• Frameworks for economic profit-sharing which reward acceptable performance and
coordinate manager and investor enrichment. These include the employment of hurdles
to make weak performance cheap and strong performance expensive.
59 Harvard Study of Adult Development

63
Tollymore’s new fee structure
Over the last few months, we have reappraised the consistency of Tollymore’s business
principles with the objective of non-zero-sumness. As a result, we are introducing a fee structure
for investors who invest a minimum of £1mn of net contributed capital, and who agree to commit
that capital for a minimum period of three years. For these investors Tollymore will offer a choice
of fee structures as follows:
• 1% management fee and 10% of any investment returns above a 5% hurdle
OR
• 1% management fee and a 20% performance fee above a benchmark return (the MSCI All
Country World Index (USD)).
The rationale: we want to ensure that investors receive most of any outperformance generated,
not just most of the absolute return. As such we want to offer a fee structure that allows us to
talk about integrity with some legitimacy. Too many managers charge egregious fees with the
sole purpose of enriching themselves at the expense of clients.
A benchmark hurdle has not been part of the original fee structure 60 due to Tollymore’s
benchmark agnosticism and absolute return remit. However, we recognise it is the opportunity
cost for many investors. A benchmark hurdle is an elegant solution to ensuring that investors
receive the majority of any alpha generated. The fixed hurdle is an option for those investors not
wishing to pay fees in a year in which Tollymore generates negative results which are superior to
the benchmark. The fixed hurdle is also lower than one might expect the very long run
performance of a global benchmark to be.
60 A choice between 1% management and 15% incentive, or 1% and 20% above a 5% hurdle.
64
Grubhub Inc: Profiting from lazy journalism
Narrative fallacy and the Murray Gell-Mann Amnesia effect
For part of my career I was a sell-side equity research analyst. I was a specialist covering pan-
European telecoms companies. And as a specialist I remember reading newspaper articles
relating to the telco sector in respected business broadsheets and feeling that the journalism was
often subjective, unsubstantiated and shallow. Often it was factually spurious. I would then turn
the page to read an article about some other sector outside of my specialism, and afford it much
higher trustworthiness than the telco piece. Michael Crichton gave this a name: The Murray Gell-
Mann Amnesia effect61.
Serious media publications invent stories to explain outcomes, without the resources or
inclination to determine causality. This often manifests itself in major descriptive U-turns as the
outcome changes with the wind. The matters about which financial and political journalists
opine are complex. This limits the mechanism to scrutinise these stories and hold their authors
to account. And there is value to their readers and listeners, who can paraphrase talking heads’
memorable soundbites at cocktail parties rather than acknowledging ignorance or retrieving the
relevant facts from their addled brains. Authority bias plays a role: media appearance confers
credibility, the belief in which is counter to independent thought and self-awareness.
Unsubstantiated conjecture is rife. As Mr. Crichton puts it: “one problem with speculation is that
it piggybacks on the Gell-Mann effect of unwarranted credibility, making the speculation look
more useful than it is”.
The goal of epistemic humility is consistent with maintaining a careful distance from today’s
media. To exercise good judgement, we should shield ourselves from the Gell-Mann effect.
Financial markets, political and economic systems, unlike meteorology, are reflexive;
participants are second guessing one another and the bases on which decisions are made are
altered by the decisions themselves. Speculation thrives because it is cheap and speculators are
not held to account, but forecasting is foolish when nobody knows the future.
Narrative Fallacy is the backward-looking mental drive to attribute a cause and effect chain to
our knowledge of the past. Without searching for reasons, we would go around with blinders on,
one thing simply happening after another. This helps us make sense of the world despite sensory
overload. However, it can cause us to make poor decisions. The power of narrative causes us to
violate probabilities and logic62. Tollymore seeks to profit from the narrative fallacy by
specifically seeking out stocks without good stories, or those with bad stories.
61 Why speculate?
62 The Linda Problem
65
GRUB: negative stories create opportunity
We purchased an equity interest in Grubhub (GRUB)

during the quarter. GRUB provides a marketing service to
restaurants to help them generate takeaway orders. The
company does this via an online and mobile platform
connecting 125k local restaurants with 20mn active diners
wishing to order takeaway food, either pick-up or delivery,
in 2.4k cities across the US. The company was founded in
2004 and listed in 2014.
The restaurant proposition is the generation of higher

margin takeaway orders at full menu prices. The diner
proposition is a more satisfactory and less error-prone
mechanism vs. paper menus and telephone ordering; more choice, more convenience, better
informed decisions.
Traditional offline takeaway ordering is the principal source of competition. That is, the dollars
that most restaurants direct to paper menus and local advertisements. The largest marketplace
and delivery competitors are DoorDash, Uber Eats and Postmates. The following quotes sum up
the prevailing press commentary on GRUB’s business progress in recent times:
“Grubub posted its best revenue growth in four years in the third-quarter; but it also reported
increased expenses, particularly in marketing, suggesting competition is finally heating up.”
“Grubhub is relying on partnerships to counter competition from the likes of Doordash and
UBER Eats.”
“Grubhub once owned close to 70% market share as recently as 2017, but extreme competition
has caused it to dwindle”.
“…battleground in the escalating competition for dominance…”
“As competition increases…Grubhub has had to boost its advertising and promotional spending
to compete.”
“Grubhub has responded to increased competition with marketing campaigns and technology
updates, which has cut into margins”
The implied causality between higher spending and higher competition is unsubstantiated.
There is little evidence for this in the form of higher customer acquisition costs, which have
remained stable. Rather, the company’s accelerated spending in recent quarters has resulted in
strong acceleration in customer growth, with net additions running at treble their historic rates.
To refute the ill-evidenced claims of lazy journalism requires no special insight, insider
66
knowledge nor proprietary analysis. But the acknowledgement of publicly available information
has led us to a different conclusion.
Which is that this seems to be far from a zero-sum situation. Despite the entrance of well
capitalised rivals such as DoorDash and Uber Eats, GRUB’s diner churn has not been affected.
GRUB’s quarterly food sales growth is accelerating. The attrition rate of restaurants leaving the
platform has also not changed in several years, and diner retention rates continue to improve.
Amazon closed its meal ordering and delivery business Amazon Restaurants in the summer of
2019, after four years of failing to build profitable scale.
Rationality is a superpower
Tollymore is charged with applying logic to anomalies. In our view the objective investment
merits of GRUB are compelling, and are at odds with the prevailing narrative concerning the
company’s prospects. GRUB enjoys lasting unfair business advantages which we expect to
protect the company’s discretionary profits.
Barriers to profitable participation in this industry emanate primarily from two-sided network
effects. That is, a marketplace connecting restaurants with diners. This is a platform that can
create considerable value due to fragmented supply. One sided network effects are also present
in the form of user reviews. GRUB has invested in the expansion of the two-sided network via
marketing dollars and business acquisitions. The merger with Seamless in 2013, for example,
allowed the company connect GRUB’s diners with more restaurants.
GRUB’s value proposition extends beyond diner acquisition and food delivery. The LevelUp
acquisition is an example of investment directed to deepening the level of service integration
between the restaurant and the platform. But GRUB also provides 24/7 access to support staff.
GRUB tracks restaurant performance on the platform and helps restaurants to manage capacity
and adequately resource demand fluctuations, and to price to maximise takeaway revenues.
As the platform grows, we could expect that brand and product awareness grows, lowering diner
and restaurant acquisition costs as word of mouth and reverse solicitation replace marketing
efforts in the generation of business leads.
GRUB enjoys an incumbent advantage: restaurants do not wish to partner with many digital food
delivery platforms. Dealing with five different partners and operating five different tablets and
systems becomes cumbersome and unappealing. They are likely to have partnerships with a
small number of platforms.
GRUB has an opportunity to redeploy discretionary profits into value-accretive activities:
The takeaway market in the US is c. $250bn, almost 50x GRUB’s gross food orders. If GRUB has
about a third of the market the total digital penetration is still in the single digits. Then there is
the very real possibility that the current takeaway industry underestimates the addressable
market if digital food ordering platforms take share from grocery stores and home cooking. In
67
2015, for the first time, Americans spent more money at restaurants than at grocery stores. Food-
service locations account for 40% of all new leases in Manhattan, more than clothing stores,
banks, and health clubs combined. In 2020, more than half of restaurant spending is projected to
be off-premise. This off-premise spending is projected to account for 80% of the industry’s
growth in the next five years.
While the annual cost of customer support and payment processing fees is unlikely to be
meaningfully scalable, the principal source of GRUB’s capital intensity is technology and
software, which is scalable. A quarter of operating expenditure is directed to sales and marketing
activities and primarily focused on growing existing assets (monetisable restaurant and diner
relationships).
Significant portions of cost are directed at growth investments: 95% of sales and marketing
according to management. And at the end of 2018 GRUB spent more on advertising that it ever
has in the past. On the FY17 conference call management stated that after stripping out overhead
and only including direct costs such as customer service and credit card processing fees the
company’s profit per order = $3.40, much higher than the current reported EBITDA per order of
$1.23. There will always be some maintenance tech and marketing requirements to service
existing business operations. But the gulf between these two numbers highlights the
reinvestment rates of the business. At $3.40 per order, GRUB’s 2018 EBITDA would be $540mn, a
10% yield to the current enterprise value.
The press speculation around margin compression and competition have contributed to an
erosion of two thirds of GRUB’s market quotation over the last year, creating a compelling
opportunity for long-term investors capable of making purchase decisions on the basis of very
positive and possible long-term outcomes, but in the face of near-term uncertainty and media
noise.
Thank you for your partnership.
Mark
68
A letter to partners, December 2019: three decisions
Dear partners,
It is our view that fund managers must have the capacity and freedom to repeatedly make the
choice between averaging down and acknowledging mistakes when confronted with quoted
price declines. We made three portfolio management decisions over the course of the year
which we believe will set the course of our long-term investment results on a higher trajectory.
The small number of investment choices worth discussing is consistent with our comfort with
inaction, and our conviction in the merits of making only occasional and sensible portfolio
decisions in the interests of maximising long-term investment results. We encourage partners to
use these decisions to judge the consistency of our actions and words, in order to interrogate the
faithfulness of our investment programme to the stated philosophy.
TripAdvisor
“When I’m wrong, I change my mind. What do you do?”
John Maynard Keynes
We initiated a position in TRIP in September 2016 for $61. We

sold our final shares in the business in November 2019 for $32.
Over the duration of our holding, TRIP has had the largest
weighting of any stock, averaging 12.8% of assets under
management. Despite the large difference between the initial
acquisition price and the final sale price, TRIP has detracted
only very slightly from Tollymore’s cumulative results; the
average purchase price was $44 and the average sale price $43.
While reasonable portfolio management mitigated the potential
for permanent dollar impairment of partners’ assets, we have, of
course, suffered a significant opportunity cost of investing in TRIP. And we have made
fundamental errors of judgment from which we seek to learn.
TRIP makes money by generating traffic, converting that traffic into a monetisable event (an
advertising lead or booking), and charging an auction-based price for such conversion. We saw a
significant monetisation opportunity given the huge global influence TRIP has on billions of
travel decisions globally, and the low customer acquisition costs TRIP enjoys as a result of its
trusted, top-of-funnel brand. The company saw this opportunity too. Management embarked on
multiple methods of stemming the company’s economic leakage – from Instant Book to media
advertising. While these initiatives depressed near-term operating results, we contended they
were sensible efforts to materially improve the company’s long run earning potential.
69
What really excited us about TRIP was the potential to connect existing demand (500mn unique
monthly visitors) to acquired or developed supply in adjacent verticals (experiences and dining).
In so doing TRIP could leapfrog the chicken and egg problem associated with nascent platform
business models. We remain excited about this potential.
However, our enthusiasm for the prospects of TRIP’s hotels business has waned. With hindsight,
we were the proverbial boiling frog in recognising the escalating challenges faced in the core
meta business. We underestimated (1) the declining utility of meta to consumers, and (2) the size
of the advantage afforded to Google as a result of being positioned just a few inches closer to the
consumer.
TRIP’s 3Q19 earnings report highlighted an accelerated decline in hotel meta revenues,
predominantly because of Google siphoning off high margin SEO traffic. That is, encouraging
customers to click on its own hotel meta links and pushing organic results from TRIP,
Booking.com and Expedia further down the page. EXPE doesn’t seem to feature on the first page
of Google at all nowadays. This is an issue that TRIP CEO Steve Kaufer has highlighted before;
Google has rebuilt the front end of its meta product from scratch and has strongly increased
review volumes and user engagement (although still far behind TRIP).
What are the competitive implications of this? We think Trivago, TRIP’s most like for like pure
play meta peer, will have a very hard time surviving Google’s tactics. As discussed, TRIP has other
business adjacencies and revenue opportunities such as media, experiences and dining that
continue to report encouraging progress. TRIP’s investment dollars are being increasingly
directed outside of the branded hotels business, into the E&D platforms and media advertising
opportunities. To this end they recruited Lindsay Nelson as ‘Chief Experience Officer’. This
seems to be a very critical hire in determining the company’s ability to turn shots on goal into
actual goals with regards to monetising TRIP’s global influence. The latest rendition of these
efforts seems to be a dawning that TRIP is a media company which is seriously under indexing vs.
its reach and engagement. Google’s actions may accelerate antitrust efforts. Only 40% of Google
searches now result in a click to a non-Google site. In congressional hearings Google has refused
to adhere to its original goal to “get you out of Google and to the right place as fast as possible”. A
move to a mission statement around “helping you get things done” has been underway for
around five years now.
But our appraisal of the quality of TRIP’s meta business was off-the-mark. We had always
assumed that it was not necessary for TRIP to take a monopolistic economic share of the digital
migration of travel decisions and advertising dollars for its equity to be a great investment. We
considered TRIP part of a global duopoly in hotel meta. And in our view the principal competitor
to TRIP’s product offering was the large portion of travel bookings and advertising still taking
place offline. But we were too dismissive of the value that Google commands by being right at
the top of the funnel for most hotel booking experiences. From this position Google has the
power to inflate OTA and meta companies’ customer acquisition costs by replacing their organic
results with ads or Google’s own inventory occupying the most valuable top-of-page real estate.
70
It was also our contention that OTAs’ efforts to spend their way into a closer relationship with
the traveller were unlikely to represent a permanent shunning of meta investments. This was
because consumers care about price, especially for homogeneous inventory. As the inventory
owner, OTAs and hotels can only offer one price. Thus, we perceived meta companies’ capacity
to scrape the internet for hundreds of prices as an enduring utility that was impossible to
replicate by those further down the funnel. We also like the fragmented supply in hotel
metasearch unlike, for example, flight metasearch. As such, we applauded management’s effort
to educate its users about TRIP’s price comparison capability.
However, over the last decade there has been steady OTA consolidation, putting meta
companies’ claims that they compare hundreds of prices on shaky ground. It also makes for a less
competitive auction, lowering achievable prices per click that have never been in TRIP’s direct
control.
If meta is a business in secular decline, TRIP’s prospects may be determined by good co./bad co.
dynamics in which the plateau of the core business causes revenue growth to stagnate, but then
accelerate as the good co. becomes an increasing part of the mix. The branded hotels business
generates c. $300mn of EBITDA. If this profit stream were to decline by 10% per year consistently,
it would be worth c. $1.5bn to an investor with a 10% opportunity cost. This is conservative as
only a (significant) minority of the hotels meta business is free SEO traffic. That would leave the
rest of the business valued at $2bn. Management expects Display revenues to double in the next
three to five years; this should be very high incremental margin revenue. And we think the
normal margin of the E&D business is c. 40% (more fragmented supply and less loyal demand
than hotels). If we apply 40% margin to the Display and E&D revenues, the maintenance EBITDA
of that business is c. $230mn, valuation 8-9x normal EBITDA. Even assuming a perpetual decline
in its meta revenues, the group can still grow at mid-single-digit rates with operating leverage
and a capital light business model.
But the investment proposition has changed. TRIP’s profitable future now lies in its ability to
monetise its traffic at the ‘research’ phase i.e. further down the funnel via media advertising, as
well as building its experiences and dining offering 63. As such, TRIP’s investment merits pale in
comparison to Tollymore’s other holdings, and so we exited our long-standing investment
dispassionately.
63 Although the future of this business may also be less attractive if core meta traffic continues to suffer.
71
Wonderful Sky Financial Group
Wonderful Sky (WSFG) stock continued

to strongly underperform global
markets in 2019. In each of the last two
years its quoted price has lagged the
MSCI ACWI by more than 40% in GBP.
Throughout our ownership of the
business we have purchased additional
shares at lower prices. WSFG is
comfortably the largest detractor from
Tollymore’s cumulative results. In our
two-year ownership our average
purchase price was 1.22 HKD, vs. the current share price of 0.65 HKD. We haven’t sold any shares.
Before we delve into the investment merits, a thought experiment: what should the enterprising
business investor be willing to pay for a simple owner-operated business which has a long
history of dominating its niche; has generated 19% annualised revenue growth over the last
decade to c.600mn HKD today (a majority of which is recurring in nature); and average net
income margins of c.35%?
Would you accept this business as a gift? What if the owner not only offered you this business
for free, but also offered you more than 100mn HKD in cash to own it? That is the price available
to investors today in the public equity markets.
WSFG is a Hong Kong based financial PR firm founded in 1996. WSFG manages corporate
disclosure responsibilities and shareholder and professional investment community relations for
corporate clients. Typically, after being engaged for a corporate finance activity (listing,
fundraising or merger/acquisition) WSFG will act as the vehicle for communications with
investors and media. Specifically, WSFG engages in media monitoring and media relations
management, press release and speech writing, design and production of corporate marketing
materials, shareholder identification and investor targeting, events coordination, international
roadshow logistics support, design of investor relations websites and financial printing.
WSFG’s value proposition is creating a positive perception of the company client, or overcoming
negative publicity, helping clients to improve their stock price and obtain favourable financing
terms. This is a labour intensive, capital light business. Operating leverage is moderate; perhaps
30% of costs can be considered fixed, although staff costs are stepped-fixed costs.
In 2019, global IPO deal numbers declined 19%, while proceeds fell by 4%, potentially as a result
of global trade tensions. The Hong Kong stock exchange has consistently ranked number one in
the world for funds raised through IPOs, and currently constitutes almost one fifth of global IPO
proceeds. Last year was no exception, with 154 IPOs raising a total of US$38b in 2019, Hong Kong
72
ranked first among global exchanges by deal numbers and proceeds 64. Notwithstanding local
political and social uncertainty, the pipeline also seems strong; more than 168 companies having
submitted their A1 filings to HKEX. Capital markets activity is cyclical but WSFG’s services for
both IPO and non-IPO clients help to mitigate the impact of capital markets downturns. The
proportion of revenues generated from recurring sources vs. IPO/project work has increased
from half to two thirds in recent years.
WSFG is the clear market leader in the world’s largest IPO market, consistently assisting 70-90%
of the largest firms listing. This dominant position is the result of a quarter of a century of
experience providing financial PR services in Hong Kong, over which time it has accumulated a
track record and reputation, as well as a network of clients and counterparty relationships. The
post IPO market is more fragmented, but WSFG is still the clear market leader, thanks to
extremely high post-IPO retention rates.
Customer price inelasticity suppresses customer acquisition costs. Financial PR and IR fees are a
small percentage of overall funds raised but are crucial to the successful execution of the IPO,
which only happens once. Deviating from the market leader is a risk seemingly not worth taking,
either by the corporates being listed or the other service providers who daren’t risk their own
reputations by recommending a PR firm other than the dominant market leader with the longest
and most successful delivery history. This lowers customer price sensitivity and allows IPOs to
serve as a starting point for building longer-term relationships. Accumulated expertise,
relationships, and WSFG’s integrated offering are all likely to contribute to switching costs which
have preserved the company’s excellent financial track record.
The growth potential of financial PR services in the region is highlighted by HKEX’s own
expansion strategy. According to its Strategic Plan 2019-2021 the HK Stock Exchange expects to
expand into the PRC: “China’s onshore capital markets continue to internationalise, driving
increasing flows of global liquidity into China…We will provide international investors with the
most comprehensive range of Chinese securities investments”. China’s market cap to GDP ratio
is 0.8x, vs. 1.2x for Japan, and 1.7x for the USA, despite enjoying considerably higher economic
growth. Hong Kong has several advantages over mainland China as a listing destination,
including less onerous listing requirements and a more trusted legal system. As a result, Hong
Kong has consistently raised more IPO funds than Shanghai/Shenzhen.
Liu Tianni is the CEO and founder of WSFG. He owns two thirds of the equity and has never sold
a share, including when the company listed in 2012. High dividend payout ratios of c.60% have
been a prominent feature of Mr. Liu’s capital allocation strategy but were recently cut to
purchase the company’s office building, and as a response to recent IPO market weakness. With
a net cash balance of c.900mn HKD and a quoted market cap of c.775mn HKD, the market now
values this business at minus 125mn HKD. While part of WSFG’s stock decline has been due to
macro uncertainty, social unrest and a volatile IPO market, we have also constructively engaged
64 Raising 40% more capital than the next highest exchange, NASDAQ.
73
with management to address the business’s overly conservative capital payout policies and a
suboptimal capital structure. There is conservative capital allocation, and there is irrational
capital allocation. If WSFG generated zero earnings forever it could still afford to pay 100mn
HKD pa in dividends for almost a decade. Assuming the business can sustain, but never grow, its
current (depressed) earnings, it could pay 100mn HKD per annum and cash balances would still
increase by 100mn HKD pa (>+10% pa). With a 50% payout ratio it would have 1bn HKD of cash
and investments on the balance sheet next year. And this would grow every year without any
earnings’ growth! This is an extraordinary valuation anomaly, the unwinding of which we expect
to be most valuable to our investment partners.
GrubHub
We discussed in detail the reasons behind our purchase of

GRUB shares in our September 2019 letter to partners. In
short, we believe the business scores very highly across the
core aspects of business quality we seek in holding
companies: a simple business; soundly financed; strong
track record; well positioned to defend and grow intrinsic
value over time; and thoughtfully stewarded. We are
conscious that GRUB is a heavily shorted stock that divides
investment opinion. This is neither something we seek nor
shun in our investments. Regardless of others’ thoughts of
the companies we own, our decisions to own them are
made independently, after conducting research calmly in a quiet room and, we hope, based on
objective information and insights. Emotionally bashing or defending stocks on investment
forums and social media is a waste of cognitive energy at best. In addition, it seems likely to
introduce hurdles such as commitment bias into decision making.
74
A few weeks after we initiated our position the company released its third quarter results,
accompanied by a letter from management which sent the stock down 43% in a single day. The
vitriol from short sellers and the financial press was deafening.
The letter detailed a slowdown in order growth as a result of diner promiscuity, leading to a reset
of long run industry growth expectations. Diner trends began to change in August, when orders
per customer trended lower than expected. This is particularly true for newer customers in
newer markets. As a result, the letter noted that “many of the supply innovations have played out
and that the industry is returning to a more normal long-term state which we believe will settle
in the low double digits, except that there are multiple players all competing for the same new
diners and order growth.” Diners also appear to be less loyal to individual platforms. The lower
diner loyalty that GRUB has seen since August seems to imply that the LTVs of new customers
deteriorated in the few months prior.
As a result of these observations management decided to accelerate investment into restaurant

inventory and diner loyalty. Management acknowledged that the single biggest determinant of
diners’ use of marketplaces is restaurant choice. Historically GRUB has chosen only to list
partnered restaurants. Due to competitors’ decisions to list non-partnered restaurants diners are
finding more restaurants on competing platforms. Given the recent decline in diner loyalty,
management are now aggressively moving to invest in eliminating this difference by adding non-
partnered restaurants. In the three months since the letter’s publication the company has
already doubled the number of restaurants on its platform. In addition, GRUB is ramping up
sales staff to quickly turn these restaurants into partners. Finally, GRUB will invest in restaurant-
funded diner loyalty programmes to improve diner stickiness and address the recent trend of
diners splitting wallet share. They do not intend to run ephemeral promotions and price
discounts.
Is this accelerated investment offensive or defensive? We think both. The decision to lean in at
this time rather than harvest maintenance profits was determined not only by these recent
trends, but also the rising pressure on GRUB’s competitors to demonstrate profitability to their
owners.
There is no change in views regarding the size of the addressable market nor the proposition gap
between platforms and single restaurant digital delivery mechanisms and offline ordering. But
the letter cautions against extrapolating the explosive pace of growth over the last few years, a
pace of growth enabled by supply side innovations which have largely played out. i.e. Uber
leveraging logistics assets and Doordash adding unpartnered restaurants. We suspect this dose
of reality was aimed as much at the owners of Uber and Doordash as GRUB’s own shareholders.
These events do not challenge our original observation that there was no evidence
demonstrating causality between profitability and competitive tension reported by short sellers
and the financial press. It was this confusion from which we believed owners of GRUB could
profit; reported customer acquisition costs and gross food sales per cohort were stable.
Accelerated investments since 4Q18 were offensive; they resulted in a trebling of net diner
75
additions and no inflation in CACs. However, management is now implying that indeed since
August the LTV is trending down due to lower diner loyalty, a trend which will cost money to
address.
Since the company’s letter it’s easy to get caught in the minutiae and try to triangulate pieces of
data to gain insights that don’t move the needle. Big picture we think GRUB still has a defensible
position. GRUB’s moat lies in barriers to entry rather than competitive differentiation. However,
we do believe that GRUB’s management has the most appropriate strategic priorities and the
richest insight into the economic dynamics of online food (demand generation: good, point to
point hot food delivery: bad). As such their investment efforts are directed to restaurant loyalty
tools vs. episodic and unsustainable discounting. GRUB also has the largest partnered network,
meaning a more durably positive experience for restaurants, diners and drivers. It has always
been our premise that a very small number of players will enjoy the economic fruits of an
inevitable trend, given the proposition difference of digital ordering vs. paper menus and
individual restaurant digital ordering solutions. There may be periods of accelerated
competition as each player does seem to believe that being the biggest is important. But the most
relevant competitor – the old way of doing things – is the least discussed.
Barriers to profitable participation are high. We don’t see any new evidence to contest that.
Switching costs are more modest and less tangible than a brand, marketing scale efficiencies and
two-sided network. But they include: (1) An inherently profitable, marketplace model. 80% of
GRUB’s restaurants are independent SMEs who value demand generation highly. GRUB adds
more than five partnered independent restaurants for every additional large-brand QSR location.
40% of orders are delivery, which is a cost-plus model and offered only to improve inventory and
therefore the strength of the marketplace profit centre. Contrast with the lack of profitable
scalability of competitors’ delivery-first, variable-cost, point to point, QSR-heavy, B2C business
model. This is not hub and spoke nor B2B delivery, both of which allow for the ‘stacking’ of
orders and delivery from one location to multiple customers. (2) The largest partnered network.
This enables the lowest prices to the diner, who is not covering the entire cost of the service
through menu mark-ups. (3) LevelUp/integration with restaurants and the ability to offer them
white label websites and run loyalty programmes.
What is the current reality? Industry commentators share an obsession with aggregate market
share in a three-player industry with local network effects and a huge, potentially
underestimated, addressable market. The digital migration may be less rapid, because there are a
no longer the innovation leaps to accelerate it 65. There are clear barriers to entry but unclear
switching costs. This was always the case, despite the evidence of diner loyalty. Consolidation
makes sense, has happened in the US and is happening elsewhere. There has been substantial
capitulation. Management is smart, thoughtful and candid. Post the sell-off the company could
65Although GRUB CEO Matt Maloney himself believes that pick-up, which accounts for the majority of the $250bn
TAM, is the next big potential innovation escalation. Over half od digital orders are pickup, yet only 10% of GRUB’s
orders are pickup orders. GrubHub Ultimate is part of the company’s efforts to pioneer and profit from this
innovation.
76
be purchased for ten times its maintenance earnings. Which it can grow at a minimum pace of
double digits for a long time. Trailing 12 months of revenues = $1.3bn, less non-discretionary
cash costs of ops/support, tech and G&A of c.$900mn = $430mn, $340mn after tax. A 13x owner
earnings multiple. There is now greater uncertainty over whether we can assume 40% as a
normal margin for the marketplace business. Despite the well-adjusted language management
uses to describe the challenges of this industry, they expect that the business could generate $2
EBITDA/order absent their investment initiatives. This supports a 40% marketplace margin
assumption. So does a comparison of mature international marketplaces. So does the income
statement normalisation exercise above. If so, and as delivery is a cost-plus undertaking, the
company traded on high single digits of normal EBITDA after the third quarter results. We think
that if diner loyalty is genuinely and sustainably lower, it seems more likely now that the
oligopoly will further consolidate into two players 66.
We responded to this extreme stock volatility by averaging down quite meaningfully at

$35/share; GRUB remains a core investment holding.
The value of independent thought
“Responsibility to yourself means refusing to let others do your thinking, talking, and naming for
you; it means learning to respect and use your own brains and instincts; hence, grappling with
hard work.”
Adrienne Rich
The booming digital distribution of opinions has increased the value of independent thinking to
investment decision makers. Those who read news and views on social media are consuming
tailored versions of events and stories to suit their specific tastes and political, economic and
social ideals. These messages are designed to hold our attention so that we may be sold
advertising messages; they are not designed to help us calibrate our attitude and improve our
judgment, for which there is no social media business case. Our world view becomes more
embedded by confirmation bias endemic to the social media echo chamber 67.
Forming opinions by examining facts, logic and data is old-fashioned and effortful compared to
listening to talking heads and internet commentators telling us where to stand on any issue we
66 Doordash’s 2019 funding rounds valued the business at c. $13bn for an annualised GMV of $7.5bn. A 1.7x EV/GMV
would value GRUB’s equity at more than $10bn. Takeaway’s merger with Just Eat values the latter at £6.2bn, on
revenues of £1bn, 6.2x EV/sales and 1.2x EV/GMV. Takeaway acquired Delivery Hero’s German business for €930mn,
on €100mn of annual sales. 9x EV/sales would value GRUB’s equity at $12bn, 2.6x the current market cap. Just Eat
acquired Hungry House from Delivery Hero in 2016 for £240mn for £29mn of profitable revenues, a sales multiple of
> 8x, over double that on which GRUB currently trades.
67 This is self-reinforcing. The echo chamber is enjoyable and addictive, causing us to come back for more
corroboration of existing views.

77
care to imagine. Market participants with higher confidence and lower accuracy are what fund
managers able to think independently should wish to compete against.
Regurgitation of others’ opinions or the capacity for such opinions to shake our own beliefs due
to inadequate independent research, are symptomatic of an unexamined life and a hindrance to
a successful career. Bad decisions are inevitably the consequence of ineffective thinking.
Our investment holdings may from time to time be controversial and divide opinion. A very
long-term horizon is most valuable when others are throwing in the towel. At any one time it is
likely that some of our companies are knocking it out of the park, and others are grappling with
hard business problems they are well placed to ultimately solve. For those who deserve it, we will
support them with patient capital.
Wishing you a healthy and happy year ahead.
Yours sincerely.
Mark
78
A letter to partners, March 2020: Tollymore’s process in a crisis
Dear partners,
The COVID-19 pandemic has caused the sharpest share price declines in history. In the first
quarter of 2020 the average US stock fell by 36% and declined 46% peak to trough 68. We do not
know the timing, duration or magnitude of further market declines nor subsequent recoveries.
To offer an opinion on this is market timing and anyone claiming ability in this field has self-
awareness issues. While many public market commentators and investment bank strategists
were calling for a correction in markets, no one stated a respiratory virus and a Saudi-Russia oil
price war would be the cause. Corrections are caused by things we have not anticipated. There is
clearly a lot of uncertainty today. But there is always uncertainty. There was uncertainty in 2007.
We just ignored it until 08/09 came along and the uncertainty was suddenly reflected in asset
prices.
We could conjure reasons to justify why markets have overreacted to the pandemic, not least
social media’s role in the more rapid and subjective dissemination of bad news. But to do so is
just marketing, and predicting the economic, political and social consequences of such events is
well beyond the scope of our abilities. To do so would be inconsistent with a focus on what is
important and what is knowable.
Future investment decisions should be made on data, not emotions. The median market
performance after a correction suggests investing into corrections is the mechanism through
which long-term investors outperform market speculators. It does not mean that if you invest
today the market will not go down further. And if the market declines further it does not mean
investing today was the wrong thing to do – process vs. outcomes. It means you didn’t pick the
bottom. But trying to pick the bottom is a tiny ROI game to play.
Investing is at times uncomfortable. If it doesn’t feel uncomfortable you are acting

conventionally, and you are destined for average investment results. The only way to navigate
these periods with sanity and resilience is to have a very long-term view. Those who have the
temperament, mindset and patience to invest countercyclically over the long-term do the best.
This letter reminds our investment partners of Tollymore’s risk framework; we explain how we
have used that framework to re-underwrite our investment holdings, with a special focus on
capital structure and liquidity. We discuss the investment merits of our four largest holdings,
now comprising more than 55% of assets under management. We conclude with a discussion of
decision making in bear markets and the importance of not being a bystander.
68 Source: Value Line Geometric Index.

79
Risk, capital structure and liquidity
Tollymore’s risk framework
We invest in companies with low business, balance sheet and valuation risk to mitigate
permanent capital erosion. Business risk is low when a company enjoys lasting unfair advantages
which protect intrinsic enterprise value. Balance sheet risk is low when the company’s capital
structure is appropriate for its business model and cost structure, resulting in enterprise value
growth accruing to owners rather than lenders. Valuation risk is low when we pay a purchase
multiple of current owner earnings which does not cause the IRR on our equity investment to
materially lag the intrinsic value growth of the business. In a nutshell, we own prudently
financed high quality assets at modest multiples to their normal earnings power. In assembling a
portfolio of investments, we do not use diversification as a tool to dampen volatility and tend
investment results to average, we do not employ leverage and we do not short securities.
Tollymore’s process in a crisis
In many ways our investment programme is unchanged in periods of economic distress and
market dislocation. We still believe in the utility of independent thinking, in the power of deep
work and the pitfalls of forced collaboration. We still seek to own companies that score highly
across our vectors of business quality. We still think of risk in terms of business quality, finances
and valuation. We still believe in exploiting a set of constraints facing most active money
managers in executing a long-term investment strategy. We still believe our investment partners
are a competitive advantage. We still believe stock market volatility is a reason to prefer public
market vs. private business ownership.
But admittedly this is no ordinary crisis. One quarter of the global population is under lockdown.
This means many businesses will need to navigate a period of zero revenues. This reality caused
us to re-underwrite our investments with a special focus on liquidity runway. In our ‘survive and
thrive’ analysis we estimated how long each business could survive without revenues;
protections mitigating revenue loss, such as contractual agreements; and what forms of cost
relief may be available to our companies, due to the natural variable cost structure of the
business, the curtailing of capital reinvestment, or government support. We compared the
estimated cash burn to liquid assets and ranked our businesses according to those most likely to
survive these extraordinary circumstances. While acknowledging the uniqueness of this crisis,
and despite espousing the merits of interrogating business prospects from first principles, we
have not worked hard enough at overcoming the faulty heuristic that low levels of net debt are
synonymous with balance sheet strength. A company with zero net debt may nonetheless
collapse if it has high fixed costs, sells discretionary services and no access to short term
financing. Even if this finance can be found, lender concessions may permanently destroy the
wealth of the company’s owners through equity dilution.
80
This capacity to survive massive revenue disruption formed one part of our analysis. The second
step involved a consideration of the opportunities to flourish at the other end of this period. Do
the restrictions placed on populations in quarantine aid or inhibit our companies’ ability to serve
their customers? Do our companies sell products or services which have replacement
characteristics, or for which demand is likely to build up over time, leading to lumpier rather
than measurably lower cash inflows? Assuming ample liquidity runway, does the operating
leverage in the business work in our favour when revenues pick up again? Does the business
benefit from structural trends that may be accelerated by the behaviour of customers confined
to their homes? Do the cost and capital structures of competitors increase the potential for
competitive rationality or consolidation?
The first two steps of this survive and thrive analysis are insufficient to help inform sensible
portfolio management decisions. We have changed the weightings of our holdings in the small
number of situations in which our expectations for the company’s staying power and ability to
flourish in more normal times are at odds with those implied by the quoted price change. We
have lowered our cost basis, sometimes very meaningfully, where we have determined this to be
in the overwhelming interests of our partners. This has included purchasing more shares of
smaller, less liquid companies where there may have been non-fundamental selling pressures
exaggerating stock price declines. Capacity constraints are an important safeguard against
permanent capital erosion; they facilitate the ability to exploit mispricings and safeguard against
being a bystander, a costly but common sin of many investment managers. The paucity of
appropriate capacity constraints is a function of widespread greed, something we hope to
exploit.
Tollymore’s own liquidity runway and aligned investment partners have allowed us to remain
calm throughout this period. This is part of being a public equity manager; stock price volatility,
which in March reached its highest level ever 69, is the price of the potential to earn superior
long-term results.
69 The Cboe Volatility Index, known as the VIX, hit a record high of 82.69 on 16 March 2020, surpassing the peak level
of 80.74 on 21 Nov 2008.
81
Portfolio holdings
The portfolio management decisions taken in the first quarter have increased portfolio
concentration, with the four largest holdings representing more than 55% of assets under
management. Our largest investments, in alphabetical order, are Aspen Group Inc, Franklin
Covey Co, Gym Group plc. and Sea Limited.
Aspen Group Inc (ASPU)
ASPU is an education technology company operating

two universities, Aspen University and United States
University. 80% of all students are degree-seeking
nursing students. ASPU is committed to the cause of
making education affordable. As such, it aims to offer
among the lowest tuition rates in the sector, which
means that only a quarter of students take federal
financial loans to fund tuition. Students are
incentivised to enrol and graduate by an incremental
salary payback of less than one year. The principal mode of instruction is online distance
learning. A web-based portal stores and delivers course content, provides interactive
communication between students and faculty, and supplies online evaluation tools. Programmes
are therefore targeted to self-directed learners.
ASPU is a cost leader in a defensive industry, enjoys outstanding unit economics and operates in
an addressable market 80x group revenues. There are almost five thousand US colleges and
universities serving traditional college age and adult students. Competition is fragmented and
varies by geography, programme offerings, delivery method, quality, and admission criteria. No
one institution has a significant share of the total postsecondary market. ASPU’s primary
competitors are online universities. Publicly quoted peers include American Public Education,
Adtalem Global Education, Grand Canyon Education and Strategic Education. ASPU also
competes with privately owned Apollo Education Group, which includes the University of
Phoenix and is considered the market leader based on total enrolments. These competitors have
degreed enrolments ranging from approximately 40-90k students, vs. ASPU’s student body of 9k.
Public institutions receive substantial government subsidies, and public and private institutions
have access to government and foundation grants, tax-deductible contributions that create large
endowments and other financial resources generally not available to for-profit schools. As such,
public and private institutions may have instructional and support resources that are superior to
those in the for-profit sector.
The for-profit education sector has shrunk by 40% over the last decade due to: (1) A poor value
proposition driven by reliance on financial aid. That is, admitting students who were not
academically ready, delivering poor outcomes and performance, and high pricing leading to
substantial student debt. This seems to have been the result of substantial reliance on federal
82
student aid and loans. In the absence of federal student loans, a normally functioning free
market would facilitate competition between educational institutions. Prices would tend to an
equilibrium, market clearing, point. A point beyond which students would not be willing to pay,
and private loan companies would not be willing to underwrite. The undisciplined provision of
financial aid seems to have created an imbalance that ultimately resulted in several lawsuits and
investigations into many colleges and increasing scrutiny on accreditation procedures. And (2)
The growth of the online program management (OPM) industry. OPMs have helped traditional
non-profits launch their degree programs online in direct competition with the online for-
profits. With similar tuition rates the name-brand university has a material customer acquisition
advantage. The realisation of these challenges has prompted some for-profits to convert to non-
profit status.
Despite this industry backdrop, ASPU is rapidly growing market share. ASPU enjoys several
barriers to profitable participation:
• Accreditation provides external recognition and status. Employers rely on the accredited
status of institutions when evaluating an employment candidate’s credentials.
• High student satisfaction scores and graduation rates.
• Cost advantage one: online distribution - Except for the recently launched pre-licensure BSN
hybrid (online/on-campus) nursing program and USU’s hybrid (online/on-campus) MSN-
FNP programs, courses have been designed entirely for online delivery, and faculty are
recruited and trained for online instruction.
• Cost advantage two: lower customer acquisition costs - When Michael Mathews became
ASPU’s CEO in 2011, he developed ASPU’s proprietary, internal, internet marketing program.
ASPU does not use third-party online lead generation companies to attract students70.
ASPU’s CRM system has also been designed to achieve lower CAC and higher student
persistence71. ASPU management contends that this CRM does not exist in the higher
education market and will drive industry-leading persistence rates and LTVs. There is
evidence to support such claims. Aspen pays $300-$1,000 to acquire a student, while
competitors typically have acquisition costs that are three times higher. Aspen converts
more than 10% of the students that have interests in enrolling on average because it is
70 For-profit online competitors utilise multiple third-party online lead generation companies to obtain most of their
student leads. Third-party leads are typically unbranded and non-exclusive. Lead generation firms sell student leads
to multiple universities; the conversion rate for those leads tends to be appreciably lower than internally generated,
Aspen Group university-specific branded, proprietary leads.
71 ASPU’s CRM includes an algorithm that recommends to enrolment advisers in priority order what follow-up cost
should be made in each day to complete the enrolment process for prospective students in that given EA's database.
The algorithm was created by studying the daily habits and activities of the three most productive EAs in Aspen's
history. This recommendation engine then automatically updates in real time after each follow-up or action is
conducted by an EA. These advanced features are not offered by any CRM software company in the industry, leading
to the highest lead conversion rates in the country. The biggest persistence challenge amongst the growing
population of fully online students in the US is the lack of timely student support. Students struggle with many
challenges including academia, finances, personal issues and time management. ASPU’s CRM provides real time data
on these aspects of a student's career. 30 at-risk events have been determined; those which, without intervention,
could lead to voluntary course withdrawal. The CRM is intended to turn the student services department into a
proactive student support group, by alerting an academic adviser when an at-risk event occurs in real time so the
adviser can contact the student to discuss ways to mitigate or solve the issue.
83
selective in its student selection process, whereas the traditional competitors have mid-
single-digit conversion rates.
Given that Aspen's cost of enrolment is significantly lower, the company can lower tuition costs
and offer a monthly payment plan. As a result, Aspen is reporting 40%+ enrolment growth, while
the industry's largest nursing school with over 30k students, Chamberlain, reports mid-single-
digit enrolment and revenue growth. Tuition costs are therefore 25-50% lower than industry
averages for nursing degree programmes. It is management’s contention that Aspen is “arguably
the most affordable nursing school in the country”, while maintaining well above industry
completion rates. The replication of these advantages by peers would require them to
aggressively lower tuition rates, slowing their already modest growth.
The for-profit education sector has proven itself unable to overcome the incumbent advantages
of not-for-profit education providers, or to provide a reasonable ROI to students. Since Michael
Mathews assumed the role of CEO in 2011 ASPU has grown enrolment almost 30% per year and
has navigated the business through a period in which its scale disallowed profitability.
But we expect that to change. ASPU operates in a niche industry with secular tailwinds and a
$4bn addressable market, via a business model with extraordinary unit economics. Half of nurses
are older than 40, and 20% of Americans will be older than 65 by 2030. The pursuit of the
Magnet Recognition Designation from the American Nurses Credentialing Centre (ANCC)
requires 80% of nursing staff have a BSN degree or higher; 40% of nurses only have a 2-year
associates degree. It is against this backdrop that ASPU’s target is to be the largest nursing school
in the US.
Modest student acquisition costs compared to the revenues generated by these enrolments,
together with the scalability of faculty costs could bode well for improving operating margins. A
single adjunct faculty member can work with as few as two students or as many as 30 at a time.
There is evidence of operating leverage in the company’s expanding gross margins, which have
been increasing c. 10ppts pa, and admin costs which are falling as a percentage of sales. Half of
cost of sales relate to instructional expenses, which are unlikely to be scalable, and the other half
relate to sales and marketing, which are growth costs. As the blended cost of acquiring students
has fallen due to the higher mix towards superior unit economics courses, sales and marketing as
a percentage of sales has fallen from a quarter to a mid-teens level. Below gross profit, admin
costs have also fallen strongly as a percentage of revenues, driving operating profit improvement
from $mid-single-digit losses a couple of years ago to a breakeven position today. These financial
results lend credence to management’s goals: “we expect to deliver increased value to our
shareholders by achieving three important financial targets: continued strong revenue growth,
rapidly improving profitability and a path to substantial positive free cash flow”. Big picture: in
the most recently reported results revenues increased by $4mn, and net losses declined by
almost $2mn; c.50% of revenue growth is falling through to the bottom line. On a sequential
basis the leverage is even more impressive: revenue rose sequentially by $1.7 mn while net losses
dropped sequentially by $1.4 mn – 83% leverage on the bottom line.
84
The cost per enrolment of these programmes ranges from $340 to $1.1k, and the LTV72 per
enrolment ranges from $7.4k to $30k. The contribution margin per marketing dollar spent across
these programmes ranges from 4x to 26x i.e. between 300% and 2,500% incremental return on
capital.
Nurses probably have the greatest job security in the world at this moment. Aspen University’s
pre-licensure BSN clinical degree program and United States University’s MSN-Family Nurse
Practitioner clinical degree program require in-person teaching as well as online instruction.
They are relatively unaffected by this crisis because the in-person simulation/immersion
activities have, with regulatory guidance, been developed to work in a virtual environment. The
demand for ASPU’s services accumulates. Course starts in Aspen’s traditional 100% online post-
licensure nursing degree programs are typically seasonally affected in the summer months as
these nurses tend to take time off in the summer. However, given these nurses are in an all-
hands on deck situation in response to COVID-19, ASPU is likely to experience more seasonality
than is typical.
ASPU has the liquidity to comfortably manage such cash flow lumpiness, thanks to a debt
refinancing and equity raise in January 2020. Meanwhile, we expect ASPU to thrive as we
approach more normal social conditions as the crisis may catalyse a structural acceleration
towards digitally delivered education and likely impending recession may accelerate a
preference for low-cost education options.
We increased our ownership in ASPU as its stock declined by 50% in just five weeks in February
and March. Today ASPU’s enterprise value is $135mn, c. 2.8xthe expected (mostly trailing)
revenues of the business in FY20 (Apr y/e). Expected bookings over this time are c. $100mn; the
business is valued at the market at 1.4x the revenue run rate. Revenues that are currently growing
c.40% pa.
ASPU’s owner earnings are above reported levels due to reinvestment of discretionary profits
into additional income statement costs, such as new enrolment advisors, associated with the
expansion of their education programmes. The excellent unit economics associated with these
programmes make this the obvious capital allocation choice to maximise long-term SVA of the
business. Given increasing LTV/CAC and contribution margin per marketing dollar ratios, either
marketing dollar reinvestment rates can be maintained in order to accelerate revenue growth, or
revenue growth can be maintained or moderated, but marketing dollar reinvestment rates
should decline over time. As such we expect our equity in ASPU to compound at 25-40% pa.
72 ASPU defines this as total revenue per enrolment.

85
Franklin Covey Co (FC)
FC offers training courses, consulting services and training-

related products focused on solving organisational problems
through changes in human behaviour. Prior to 2015 FC sold
engagement-by-engagement training and performance
solutions to companies seeking to develop leadership,
improve employee productivity, develop internal and
external trust, augment sales performance and build
customer loyalty. Clients can purchase complete offerings
such as The 7 Habits of Highly Effective People and The 5
Choices to Extraordinary Productivity, or use individual
concepts to create a custom solution. To increase customer
lifetime value, FC began the process of converting to a SaaS model, called All Access Pass (AAP).
AAP allows FC customers unlimited access to all of FC’s content. This content can be delivered
though a broad range of modalities in 16 languages. The cost per population trained is equal to or
less than single competitive courses delivered through single modalities. FC did therefore not
undergo this business model change to inflate price per population trained, but rather to
maximise customer lifetime value73. Retention rates under the subscription model are superior
because: decisions to renew are passive rather than active; commitment bias may play a role
given the more arduous initial approval process; and customer value is superior – more content
is available for the same price.
Five years into the business model transition FC is better positioned to weather event driven
macro crises such as these. More than half of revenues are now subscriptions. This subscription
model provides clients with the ability to access FC content and solutions across a wide variety
of delivery channels, including digital and live on-line, allowing FC to help clients who are now
working remotely. The company recently reported that AAP revenues in North America
increased year on year in March. Significant income statement expenses are associated with
client and revenue growth, such as client partner variable compensation. We estimate that FC
could fund its monthly cash burn for over a year, even if all non-recurring income stopped
immediately.
The goals of improving performance in the areas of leadership, execution, trust and productivity
have been enduring business objectives regardless of time, geography or industry. And the
human behavioural flaws at the root of these challenges seem equally perpetual in nature. The
goal of achieving permanent human behavioural change is a long lasting one likely to need
ongoing, multi-year support. A typical AAP subscription is in the region of $200 pa per
employee. At this price point, it really doesn’t take much in the way of employee productivity
73More people trained initially and subsequently, lower customer acquisition costs/higher retention rates, greater
capacity to sell add-on implementation services
86
improvement for an investment to be worthwhile. This low cost/high utility proposition 74 makes
switching unlikely once a client chooses to partner with FC; FC’s techniques and processes are
deeply integrated with the way that clients conduct their business. High retention rates are
evidence of this.
FC recently acquired the licence rights to intellectual property based on the works of Clayton
Christiansen. Creators of compelling content seek to monetise that content further than the
obvious avenues of book sales and keynote speeches (think Jim Collins licensing Good to Great
and Built to Last to FC). As a global leader FC seems to be an obvious company with whom to
have a conversation about licencing that content. FC then turns it into learning modules,
classrooms and course components, and helps companies establish facilitators to teach the
content internally. The transition to AAP enables a potential platform business model
opportunity. Previously, under an engagement-by-engagement model, the introduction of new
content served largely to cannibalise the available options, of which companies typically chose
one solution. Under AAP, with access to all available content for an annual subscription, there is
room to licence world-class content and differentiate vs. competitive offerings.
The US training industry is estimated to be worth >$90 bn, >$350bn on a global basis. In the US
$6-7bn of training spend is currently outsourced. FC’s largest competitor is therefore the client
base itself. The outsourced providers are fragmented, with few large competitors. Industry
fragmentation is testament to the low barriers to entry allowing potential new entrants to
participate. FC’s target market is companies with >200 employees, of which there are nearly 100k
in the US alone. Of these, just over 10% are assigned to current client partners, and of those 4%
are clients. The remaining 83k companies will be assigned to new client partners as they are
hired. We estimate that, in normal times, the business can grow revenues at a mid-single-digit
pace without hiring any new client partners, due to the ramp up in profitability and maturity of
existing sales staff. Incorporating the company’s hiring ambitions, we think the business could
enjoy low double-digit growth with high incremental profit margins and low capital intensity.
Management is not content to rest on its laurels but is clearly willing to make uncomfortable
strategic decisions to improve the long-term prospects of the business. Despite the large
opportunity to deploy capital into value-accretive growth, management’s number one capital
allocation priority has been balance sheet strength. FC has almost $50mn of deferred revenue on
the balance sheet, all of which will be released with minimal associated cost to the income
statement. In addition, the company has $35mn of unbilled deferred revenue due to multi-year
contracts with its clients. These economic characteristic result in a business in which more than
90% of revenue growth flows through to EBITDA. So, while customers delaying the renewal of
contracts will inevitably slow new sales and depress margins, this operating leverage will work in
reverse as FC benefits from clients getting their bearings. It took four months after the Global
Financial Crisis for the pace of bookings to return to normal levels. FC suffered from clients
74Say an average employee earns $50k pa and generates 2x revenues for her employer. If FC improves productivity
by 5%, her employer has earned an extra $5k of revenues for a $200 investment, an ROI of 25x.
87
postponing purchase decisions, but the company did not lose significant business – training and
consultancy revenues declined 17% in 2009, when there was no subscription product. Yet in one
month FC’s quoted share price declined 63%, and the stock has lost more than half is quoted
value in the first quarter.
Today FC trades on 1.2x EV/reported sales. The enterprise value is c. 85% of sales including billed
and unbilled deferred revenue. We expect the current cash flow of the business in normal times
to be c.$30mn, a 9% yield to the current quoted price, and we think this could compound at rates
of c.20% p.a. for a long time. We purchased more shares, lowering our cost of ownership, and
therefore our risk of permanent capital impairment, and improving the IRR we expect to earn on
this investment.
Gym Group plc (GYM)
The Gym was founded in 2007 and is today the second largest
low-cost gym operator in the UK. It is a simple one service, one
market business, offering services with enduring utility. It
employs a recurring revenue model with a predictable cost
base and strong FCF generation, but high operational and
financial gearing.
Despite prices 60-70% cheaper than traditional gyms, GYM’s

margins are materially higher, weakening peers’ ability to
compete on price. Cost advantage is facilitated by superior
asset utilisation. This is driven by fewer staff, more equipment
stations and fewer wet facilities, tennis/squash courts and 24/7 opening. Economies of scale
relating to equipment purchases as well as property developer relationships and site flexibility
are lowering average site investment costs as the estate grows.
GYM has multiple shots on goal to deploy capital into high returns projects, including market
share take from public and traditional private gyms, increased penetration of UK gym
memberships, lower entry prices and lower minimum age requirements increasing the
addressable market. These volume drivers are expected to lead to 15-20% site growth per year.
There are also multiple avenues to increase gym yields, including estate maturation, ancillary
revenue development including potentially targeted advertising and customer data
monetisation, and better amortisation of marketing spend lowering customer acquisition and
retention costs.
GYM’s business will be strongly adversely affected by the lockdown in the UK. Gyms have been
ordered to shut down and will therefore earn no revenues from customers, none of whom are on
contracts. GYM has frozen all memberships for free. 100% revenue loss during periods of
shutdown is therefore a reasonable assumption and one we have assumed in our ‘survive and
thrive’ analysis. In addition, two other factors will make this period challenging: (1) this is a
88
business with high fixed costs, and (2) GYM will not enjoy pent up demand for its services 75. The
market (over)reacted swiftly in recognition of these challenges; GYM stock lost half its quoted
value in the first quarter, enduring a peak to trough decline of 75% over a one-month period.
Given the operating leverage and off-balance-sheet financing in the business model, our original
research on the company presented scenario analyses which stress tested the point at which
GYM could no longer service its maintenance capex and operating lease commitments. The
following excerpt from our December 2018 letter to partners describes some of this work:
“Leverage/recession risk: In the event of a recession prices and memberships may erode. GYM
has high financial gearing in the form of operating lease obligations. Given GYM’s operating
leverage mature site profitability would fall significantly should the business experience a
marked revenue decline. We estimate that a 20% drop in revenues would reduce the owner
earnings to £17mn from £40mn in a no-growth scenario. This is still a 5% yield to the current
market cap. Current adjusted net debt/EBITDAR = 3.5x; assuming the maintenance profitability
of the estate this is 2.5x. This would be 6x with a 20% revenue decline assuming zero capacity to
cut 100% fixed costs, or 3.8x based on maintenance profitability. Under these assumptions
GYM’s EBITDA margin would be 13%/ and maintenance EBITDA margin 35% vs. the current 30%
rate (and 47% mature gym rate). It is difficult to envisage a 20% revenue decline due to the
immaturity and increasing penetration of low-cost gyms, as well as the 60-70% price discounts
vs. mid-tier competitors. When US membership growth was negative in 2012, Planet Fitness still
grew its memberships by 28% yoy. This might suggest that the large discount lowers the price
elasticity of demand for budget gyms.”
Admittedly we did not go far enough; we did not envisage a circumstance in which GYM would,
overnight, lose 100% of its revenues. We tweaked our original assumptions as part of the re-
underwriting of our ownership of GYM shares. Assuming GYM continues to earn no revenue,
ceases all maintenance capital investment programmes, and obtains business rates and
furloughed staff relief from the UK government, we estimate GYM could fund its rate of cash
burn for five months, or ten months should it draw on its existing accordion credit facility. This
gives no credit for GYM’s ability to renegotiate rental agreements with landlords to obtain
deferrals or discounts, despite this being likely due to the paucity of creditworthy tenant
alternatives. We think more broadly that access to low-cost health and fitness for all is a mission
the UK government is likely to support. In short, we think the business has adequate liquidity to
survive. Meanwhile we see several opportunities to thrive once gyms re-open. The transition
from mid-tier to low cost may accelerate if recessionary conditions follow. Smaller scale, more
levered and less financially sound independent gyms may close.
More so than any other stock we own, we moved aggressively to purchase many more shares,
substantially reducing our cost of ownership. Unfortunately, because we conducted the above
calculations swiftly, we purchased more shares too readily; we would have served our partners
75 When gyms reopen, customers will not be paying more than 100% of their previous monthly membership fees.
89
better by waiting. But frankly we could not have anticipated the market would present the
outstanding opportunity it did in the middle of March. When we finished averaging down at
76.9p, the stock was trading at an owner earnings yield of 35%. The company is investing
practically all these owner earnings into new site development yielding cash on cash after tax
returns of c.20%. At any reasonable cost of capital, we expect these additional purchases to
generate an annual investment return of c.60-70% for our investment partners.
Sea Limited (SE)
SE operates three platform businesses in gaming, ecommerce

and digital payments, primarily in seven Southeast Asian
markets. Garena distributes mobile and PC online games in its
markets. Shopee is a platform for connecting buyers and sellers
of long tail products across fashion, health and beauty, home
and living, and baby products. SeaMoney is a digital payments
provider; consumers use it as an e-wallet to pay for products and
services. SeaMoney is integrated into the Garena and Shopee
platforms. All three businesses are platform business models which take much investment to
drive scale and barriers to entry but have winner-take-most potential economic outcomes.
Shopee is the largest ecommerce platform in SE’s markets. As the number of buyers increases,
Shopee attracts an increasing number of sellers, resulting in increases in SKU variety available on
the platform, which increases the purchasing opportunities, and therefore monetisation
potential, or value, for each of those buyers. Ecommerce penetration is materially below global
averages in almost all Shopee’s markets. But ecommerce and m-commerce engagement in
Indonesia, Shopee’s largest market representing almost half of GMV, is the highest in the world.
The GMV of the internet economy is c.3% of SE Asia’s GDP, almost 10 years behind the U.S.
Shopee’s take rate (revenues/GMV) is currently suppressed by efforts to build scale and market
leadership, entrenching the network effects’ barriers to entry of the business. But the take rate
has been consistently increasing due to higher commissions and advertising fees. In addition,
sales and marketing efficiency has been suppressed by efforts to build platform scale and
liquidity. Shopee has heavily subsidised the cost of shipping for its sellers in order to build
supply scale. The extent of these subsidies has been declining without any noticeable impact of
GMV growth, resulting in sales and marketing expenses declining as a percentage of revenues.
Garena is a platform business with network effects dynamics due to the social, multi-player
nature of the games distributed and self-developed. Each new gamer increases the value of the
platform for existing users. Garena’s success in distributing games for local game players
facilitated relationships with international game developers, which has allowed Garena to source
high quality games from world class developers, many of whom work as exclusive partners.
Southeast Asia is the fastest growing games market in the world; growing 20-25% p.a. Despite
65% internet penetration, the region has the most engaged mobile internet users in the world.
Video gaming has been growing its share of consumer time and entertainment spend for years.
90
In 2019, consumers spent $120bn on video games (excluding hardware), up from $35bn fifteen
years ago. This makes the category more than three times larger than the global box office (two
times if including home video and SVOD revenues) and four times the recorded music industry.
It is also growing 2.5–3.5x faster and on a larger base.
A wide variety of merchants are on the SeaMoney platform. SE is focused on increasing the
number or type of merchant to increase the number of users on the platform, principally
through the growth of the Garena and Shopee user base. SeaMoney has thus far mainly served to
lower the costs of doing business for Shopee and Garena. By launching this business in 2014, SE
reduced payment channel costs for Garena and Shopee and captured value that previously
flowed to third-party payment services.
SE’s stock price increased 10% in the first quarter. Its weighting in our portfolio therefore
increased by virtue of substantially outperforming all our other investments. With more than
$3bn of liquidity and gaming and ecommerce markets unlikely to be materially negatively
affected directly by quarantine, we feel very comfortable with SE’s capacity to weather this crisis.
We also see ways in which SE might bloom as the world’s governments and citizens get to grips
with COVID-19. Coronavirus has led to a 12% increase in digital video traffic and a 20% increase
in web traffic. While Nielsen reports total TV time (digital and linear) is up 20%, video gaming is
up 75%. Gaming is one of the few ways to socialise today. During a time in which hundreds of
millions of people are forced into physical isolation, the ability to co-experience content from a
safe distance becomes particularly important. Players play immersive multi-player games
primarily to spend time with their friends. Games such as Fortnite and Free Fire are immersive
social networks. COVID-19 may also accelerate the migration to digital gaming by forcing
physical game retailers to shut down. Esports may also benefit from coronavirus. Several sports
leagues, from the NBA to America’s NASCAR and Europe’s F1, have sought to replace their
cancelled seasons with purely digital events. As for ecommerce, this is now the only way to shop
for non-essential items for many of the world’s population. Supply may also increase as retailers
rely entirely on online purchasing and delivery vs. store browsing. This period may therefore
ignite a 'kink' in ecommerce penetration that is a temporary accelerant.
Decision making in bear markets
Don’t be a bystander
Credit conditions of recent times have possibly allowed weak companies to thrive. This event-
driven macro crisis may expose those firms who have been swimming naked. Likewise, investors,
seduced by a decade of low finance costs, have geared up their investments, leaving them ill-
equipped to invest countercyclically. Both dynamics may exacerbate quoted price declines. We
must not be a bystander. Knowing what to do in a crisis and being able to do it are two different
things. Our companies are appropriately capitalised, we do not employ leverage at the portfolio
level, and our investment process, capacity constraints and working environment allow us to act
decisively when the odds are overwhelmingly in our favour.
91
We are optimists. We think that makes us sound less smart than pessimists, but we are not in the
business of convincing others of our intellect. As optimists, we believe in the forward progress of
humans and humanity. But we believe that bear markets present special opportunities to make
money, although it may not feel that way at the time. Decision making in periods of collapsing
share prices determines long run investment success; such environments offer up choices which
can materially alter the course of future investment results. In times of economic stress and
market dislocation, sound judgment has the potential to be most lucrative. Stress biologically
shortens one’s horizon76. Myopic loss aversion increases in crises77; our ability to exploit the
institutional constraints of our competition matters now more than ever.
We present in the Appendix a transcript from a presentation and Q&A with students and alumni
from London Business School in February 2019. We presented a context for investment
managers to judge their prospects for success, a context that may allow investors to stay the
course in periods of self-doubt. We discussed the behavioural constraints facing the investment
management industry and some of the tactics and tools investors can employ to create long-term
value. The video can be found here and a copy of the presentation here.
We wish all our partners and their families good health. Your support at this time has been
outstanding, for which we are most grateful.
Yours sincerely,
Mark
76 Stress causes the release of hormones, such as adrenaline and cortisol, in all vertebrates. These hormones evolved
to help facilitate behaviours that help vertebrates survive when they were about to become dinner.
77 Myopic loss aversion occurs when investors take a view of their investments that is strongly focused on the short
term, leading them to react too negatively to recent losses, which may be at the expense of long-term benefits (Thaler
et al., 1997). Loss aversion refers to the disproportionate responses to wins and losses. Specifically, we regret losses
two and half times more than we celebrate similar sized gains. Myopia refers to how often we evaluate our
portfolios. The more frequently we check, the higher the odds we will find a loss and experience loss aversion. On the
other hand, the less frequently we check, the greater the likelihood of finding gains. As investors check their security
prices more frequently in periods of high volatility, this behavioural bias bids up equity premia and bids down equity
prices, reinforcing myopic loss aversion and creating a downward spiral intensifying price declines.
92
Appendix: Tollymore presentation and Q&A at London Business School
Mark Walker: The title of this talk is “Investing is not as Hard as We Think.” That’s a bit tongue-in-
cheek; it’s a provocative title. This is not a presentation about how easy investing is. It’s hard. And
if any of we professional fund managers would remotely pay attention to the base rate of success
in the industry, we wouldn’t do what we’re doing. Thankfully, a healthy level of hubris prevents
us from being rational about that decision. What I want to do is share a bit of my experience in
providing a context to help us judge our decision-making and investment management and to
assess our prospects for success in the industry.
Tollymore Investment Partners is a private investment partnership here in London. It invests in a

concentrated portfolio of global public equities with the goal of compounding investors’ capital
over the long-term. I spent five years in generalist global equity mandates on the buy side, and
before that, five or six years on the sell side working on pan-European telecoms research for
Redburn Partners and Goldman Sachs. It’s that experience which has given me an overview of
the institutional money management industry, how the actors in that industry behave, how
they’re incentivized, and about several of the behavioural constraints they face in making good
investment decisions.
Investing is not easy. Only one in 13, one in 19, and one in 23 managers, depending on your
investment universe, outperforms their benchmarks. However, what we need is a framework to
be able to stay the course in periods of underperformance or other source of self-doubt. We
already know investing is hard. I spend time in a reflective, introspective, and self-aware group of
value investors across the globe, and we constantly remind ourselves how difficult this industry
is. Indeed, spending time with those types of thoughtful, talented investors can lead one to feel
like a fraud and to doubt our own beliefs and our capacity to add value for our clients. Therefore,
we need a broader context in which to judge our prospects for success.
Through my experience of speaking to key decision-makers in institutional money management

firms on the sell side, I want to present a collection of observations from that experience. There
are eight observations. For each one, I’ll give you a flavour of my personal experience. I will
discuss why I think certain behavioural constraints are antithetical to good investment practice.
I’ll highlight some of the literature and the research which provide the outside view of that
particular constraint, and then I’ll suggest some measures we can take to avoid, and hopefully,
exploit those behavioural errors.
My first observation concerns a pursuit of informational edge. I saw investment processes led by
corporate access where investment managers travel the world trying to find the best investment
ideas. They speak to the managers at the companies they own and the companies they look at.
And frequently, that’s right at the front of an investment process. The first time they are
introduced to a company is from the words of a talented salesman in the form of a CEO. What I
also saw was this focused effort on building complex and detailed financial models of companies
which were predicated on knowing everything about everything, and therefore, having a broad
93
knowledge rather than a deep understanding of some of the key factors that matter in
understanding business drivers. This led to an inability to calibrate some of the forecasting errors
back into improving our investment processes. What’s the problem of this approach? It’s that
information gathering I saw led to increasing confidence in fund managers’ expectations of the
future and into sell side capacity to forecast businesses but not leading to any better accuracy.
And whereas the sell side and their buy side counterparts undoubtedly have clearly above
average IQs and are well motivated and are well incentivized, there’s a lot of misdirected
intellectual effort in terms of trying to play this difficult game.
What does the literature say about this pursuit of informational edge leading to overconfidence?
It can have profound consequences. In the investment management industry, it can inflate our
perceived valuation of the investments we own, and it has several ramifications outside of
investing. Several studies have shown little correlation between confidence and accuracy. One
of those is related to a study in which professional horse handicappers estimated the outcome of
a series of 10 horse races, with each race having 10 horses. In each subsequent round of the
game, they were given more information about the jockeys, about the horses, about their form.
And initially, when they were given a small subset of information, they were asked to predict the
outcome and to give an estimate of their confidence in the outcome. They closely matched. A
random guess would give an outcome probability of 10%. They were accurate about 17% of the
time, and their estimation of their confidence was also around 17% of 100%.
As the rounds of games went on, they were given more information. Their accuracy stayed
roughly flat at about 17%, but their estimated confidence went up massively to 40% to 50%. That
study suggests that there are, at best, flat returns to increased information. This problem
manifests itself in position sizing in a portfolio. It causes us to dramatically increase the size of
our bets. James Montier popularised a study which shows there are diminishing returns to extra
information through the mechanism of overloading our cognitive capacities. In his book, Value
Investing, he shows a study in which participants were asked to estimate or choose the best car,
which can be objectively measured. When they were given just four aspects of the car, they
guessed correctly 60% of the time. When they were given 12 aspects, that dropped to 20% of the
time.
How can we avoid this or exploit it? In terms of corporate access, we need to be cognizant of the
sales pitches from CEOs. Consider the role corporate access plays in the investment process. I
suggest right at the front of the investment process is not the best place. Give yourself time to
carve out independent thinking and study around the business and the industry so you can
shape the meeting with management and are more insulated from sales pitches. Conduct work
on the handful of drivers you’ve identified rather than striving to know everything about
everything. And construct an investment program with a large opportunity set and a
concentrated portfolio. That avoids the temptation to be precise about your estimations of
intrinsic value for the companies you own or would like to own. Consider a mental model of
equal weighted portfolios. As investment managers, we like to think our portfolio management
94
skills are strong. But if this overconfidence tendency causes us to inappropriately outsize our
bets, then consider a constraint in which we have an equal weighted portfolio.
The second observation is the pursuit of analytical edge. I observed how large money
management firms tend to be specialists, either geographically or by sector rather than have
generalist mandates. Therefore, sector teams are forced to pick winners and losers within narrow
universes. This creates the perception, real or otherwise, of deep intellectual expertise which
leads to hubris and a level of certainty which might not be appropriate and is a barrier to
probabilistic thinking. It’s probabilistic thinking that might help us to reason to be accurate
rather than reasoning to be right. This deep self-perceived expertise potentially makes us
unreceptive to conversations with people with dissenting opinions. That’s a barrier to decision-
making. Large asset managers tend to organise themselves in terms of analysts and portfolio
managers, where analysts do deep fundamental work on companies and portfolio managers
typically do not. Therefore, there’s a conviction difference between those two groups. When
share prices move around a lot, portfolio managers may lack the conviction to know how to act
optimally in the face of that volatility because they haven’t organically done the work
themselves. In these kinds of large teams in the pursuit of analytical edge, hierarchical structures
in larger organisations introduce authority bias. Committee-led decision-making typically slows
decision-making and is a form of loss aversion because we cannot act decisively when odds are
clearly in our favour.
What you often see is investment managers and their counterparts on the sell side who need to
sound smart to justify high fee structures. You rarely hear the words, “I don’t know” when “I
don’t know”, is, in most instances, the correct answer. It’s a barrier to epistemic humility, which is
necessary for good decision-making. These fee structures create the temptation to conceal
ignorance rather than to openly acknowledge it. The large teams create the perception of
intellectual expertise, which makes us less receptive to perceived non-experts.
Large teams increase the complexity of organisations. As you arithmetically add individuals to
your investment team, you’re geometrically increasing the number of relationships and
communication lines between those individuals. You’re also making it more difficult for
individuals to be more accountable for their work. There are more places to hide. Your ability to
discern the difference between luck and skill goes down. The bad news for everyone in this
room is that smarter people are worse at combatting these behavioural biases. They are better at
assembling the information, assembling the narrative in their minds that fits their preconceived
ideas.
Does the literature support this personal experience? Milgram conducted experiments on
authority bias in which participants were asked to administer electric shocks to actors. Half of
those participants, in response to a person in a white coat with a clipboard asking them to
increase the level of electric shocks, administered what would be fatal electric shocks. This
conflicted with participants’ personal conscience. Authority is useful in social systems, but it can
clearly be a barrier to decision-making. Michael Mauboussin suggests that a quarter of the time,
95
professional decision-makers, including investment managers, calibrate their processes and
learn from experience. The rest of the time they do something which feels more like progress,
like gathering and analysing information.
There’s something called a HiPPO effect, which is the highest paid person’s opinion. That can be
a barrier to evidence-based decision-making. A couple of high-profile examples come from
Apple and Amazon. Ron Johnson was a guy who headed Apple retail; he was responsible for the
successful rollout of the Apple stores. He left to become CEO of JC Penney. In a nutshell, he
created this culture in which people were disincentivised to be sceptics, to dissent. And
therefore, no one was willing to question his unsuccessful strategy of rolling out JC Penney
stores without testing the retail strategy first.
At Amazon, Jeff Bezos became obsessed with a feature of the Fire Phone called dynamic
perspective. Like Johnson at JC Penney, he created an environment in which engineers were
unwilling to air their concern about the tactical and financial viability of this feature.
If you’re in large groups, you could succumb to something called group think. There were some
Asch conformity experiments conducted, one of which was called “face the rear.” Imagine a
scene in an elevator. Of course, nearly everybody faces the door when they ride an elevator.
However, in this illustration, four actors entered an elevator in which a non-actor, a naïve
person, had already entered and took their position facing the door. The four actors all faced the
rear after they entered. The naïve person tried to maintain his individuality. But little by little, he
turned slightly more to the wall.
In a second illustration of conformity, a naïve person had occupied the elevator facing toward
the door, then four actors entered the elevator and faced the rear wall. The naïve person
conformed by facing the rear wall. During the elevator ride, the actors all turned facing the same
side wall. When the elevator door opened to reveal its occupants, everybody had changed
positions to face the same side wall, including the naïve person.
The point in these elevator illustrations is that group pressure is subconsciously strong. Groups
are like mind-altering substances. Other studies asked subjects to do a simple task, like to choose
which line of three lines is the longest. When the experiment is performed with subjects in
isolation, 99% of people get it right. When they’re put in a group setting where everyone else is
planted and chooses a different option, that rate drops to about 50% or 60%. When those people
are interviewed afterwards, they did not say they felt pressure to conform to the group. They said
they genuinely believed that was the right answer.
How can we exploit this? Investment teams should be constructed with small numbers. Mine is
small. I’m a single-person investment committee. Small teams help to retain the accountability of
decision-makers and provide for simple communication structures. It should be diverse. If the
collective decision of a group is to be wise, individual decisions should be independently
reached. The solution for a number of these constraints is to have a large investment universe
and a concentrated portfolio, which obviates the need for precise valuation mechanisms.
96
Embrace uncertainty. Be willing to say you don’t know. Consider the exercise of being asked to
bet on something. If you express an opinion and you were forced to bet on it, to what extent
would that calibrate the certainty with which you expressed that opinion?
Embrace the spirit of intellectual generosity with peers. In this business, I have ongoing
relationships with peer investment managers. There is no sentiment for shielding our intellectual
property. We’re there to help one another. We share the belief that execution of a long-term
investment strategy is our competitive advantage, not the mere fact that we are long-term
investors. Consider the usefulness of anonymous voting mechanisms to avoid authority bias
when reaching decisions. Portfolio managers should be analysts, and analysts should be
portfolio managers. Everyone is doing the work, and you can have a shared conviction. Consider
the incentives of people outside your organisation giving you advice.
Number three is pitchbook mentality. Large asset managers have marketing-led investment
strategies. They’re trying to create an investment strategy they think will sell. That tends to lead
them down a path of having niche investment universes predicated on analytical edge using
scuttlebutt or primary research methods which are potentially into better investment returns.
But often, they reduce to marketing presentations. I also see all the time proprietary idea
generation funnels into marketing presentations in large research teams to advertise their
analytical edge. There’s also a bias to action. Managers need to feel they are earning their fees by
doing stuff.
What’s the issue? There’s an asset gathering imperative in the asset management industry. This is
a scalable model. It’s tempting to grow assets and rapidly increase the profitability of the
business model. This imperative can misalign with the providers of the capital you manage. Our
strategy is tailored to the investment manager’s temperament and what you think will work
rather than what you think will sell. The outside view supports this tendency for action.
Goalkeepers facing penalty kicks will dive almost every time, which is at odds with what they
should do given where the ball goes. But they just can’t stand the potential situation in which
they stand still and the ball whizzes by them. If only to avoid embarrassment, there’s a bias for
action. In some other examples of action bias, some research suggests trading frequency is
negatively correlated with returns. Warren Buffett certainly aims to profit from that action bias.
If you have an investment management business owned by shareholders or equity owners in a

business, there’s clearly an obligation to those owners to maximize the profitability of the
investment management business. There are ways to enrich the owners of an asset management
business that are consistent with compounding the capital of your investors, and there are ways
that are inconsistent. An asset gathering imperative is mostly inconsistent. Having managers
with high levels of insider ownership and their own funds is consistent. Having a performance
fee-led structure is consistent.
Number four is the folly of forecasting. When I was on the sell side, there was a strong focus in
having the most accurate numbers. StarMine collects all the analyst estimates for the companies
they cover, and it ranks analysts and investment banks according to the analysts with the most
97
accurate numbers. My firm had the most accurate numbers, but our numbers are still horribly
inaccurate. The Ashton Partners study showed virtually everyone in the asset management
industry thinks that consensus estimates are important.
What’s the issue with this? We’re not operating in a linear system. In meteorology, if someone
opens an umbrella, it doesn’t affect the accuracy of the weather forecast. Markets and economies
are reflexive systems. They’re not like meteorology. If a macroeconomist advocates for raising
interest rates because of his expectation that inflation will increase, to what extent is he
incorporating the feedback loop of employees asking for higher wages in response to that
expectation of high inflation? And the volatility of stocks in the near-term certainly makes
predicting share price movements exceptionally difficult. The outside view, in a nutshell,
suggests strategists cannot predict markets. Macroeconomists can’t predict economies. They
predicted two of the last 150 recessions. Chess, which is a reflexive game materially simpler than
an economy or a financial market, has 10 to the 120th possible moves.
Dick Thaler popularised a game, which was a version of Keynes’ beauty contest. He asked
participants to guess a number between 0 and 100. The person who guessed closer to 2/3 of the
answer of the average would win a prize. This is a reflexive model. There’s a wide variety of
guesses, and there are certain spikes in guesses. A spike occurs at 33; this is a first level thinker
guessing the average guess would be 50 and the 33 guess would be 2/3 of 50. Another spike
occurs at 22; this is a second level thinker. He thinks everyone else will guess 33. Another spike
occurs at 15 by third level thinkers. Finally, there’s the Nash equilibrium at zero.
Guessing share price movements is difficult. Stock market volatility in the near-term is extremely
high. If you just play at the end where you lengthen your investment horizon and you’re
estimating the long-term prospects of businesses and not the short-term movements of their
quoted prices, your prospects for a good outcome are better.
I suggest that trying to understand and analyse a reflexive complex system such as the financial
market is a low return on investment exercise. Better directed intellectual effort is in
independent research on the fundamentals of companies. Having a large investment universe
and a concentrated portfolio lets you focus on the large gaps between price and value. In this
approach, you don’t need to find a company with 10% upside and be guessing EPS estimates.
My fifth observation is manager/investor misalignment. The observation is that fund managers

don’t invest in their own funds. Management fees are typically high. Performance fees seem to
be justified by this legacy expectation that strategies drive fees rather than any consideration of
the power of incentives. The issue is that managers’ and investors’ fortunes are not aligned.
Particularly, managers without insider ownership are much less incentivised to limit the size of
their fund or even consider the capacity of their investment strategy, which is a barrier to the
time-weighted returns on their portfolios. Two observations emerge from the research. There
are potentially no statistical indicators of future performance. Maybe there is one, which is
insider ownership. Evidence supports the conclusion that companies, not just investment
management firms, outperform businesses that are not founder led in the long run.
98
How should we avoid the constraint of misalignment? Have skin in the game. Your incentive
structure should enrich and reward the managers of capital and for doing a good job with that
capital. Your enrichment should, in no small part, come from the compounding of your own
capital. It should come from a thoughtful performance fee-led fee structure. Consider the
employment of hurdles, so charging a certain performance fee over a hurdle which has the effect
of making the early performance cheaper for investors and strong performance more expensive.
If the manager does well, the investors do well.
There’s a debate whether management fees should be charged at all, or whether it’s your Buffett
partnership fee structure of not having a management fee. The Mohnish Pabrai fee structure is
the ultimate alignment of interest. I’m a subscale emerging manager, and the management fees
certainly help me to make stress-free business and investment decisions. If I didn’t have a
management fee structure and re-mortgaged my house and made all kinds of personal financial
decisions, I would not make good decisions for my investors. Also, tell your partners what’s in
the portfolio. Be transparent with the economic prospects of the portfolio holding companies.
You ultimately want an LP base that can invest counter-cyclically. They can give you money
when the markets are down. The vast majority of investors invest pro-cyclically, and that is a
barrier to the sustainability of the money management business model.
My sixth observation concerns career risk and loss aversion. There is no shortage of rhetoric in
the conversations I have with fund managers and large global asset managers about being a long-
term fundamental stock picker in one breath, and with the next breath, asking you what
Vodafone’s Q3 EPS will be. These fund managers may well have a similar investment philosophy
to long-term investors, but their capacity to execute that investment philosophy, their capacity
to have an investment process consistent with that philosophy is inhibited by all of these things,
by career risk, by short-term capital. All of these fund managers have a big budget, so they spend
their big budgets on Bloomberg terminals, which encourages them to look at share prices, to
look at near-term news flow, to look at markets, and ultimately to try to predict share price
movements. This essentially leads to benchmark hugging.
The European Securities and Markets Authority (ESMA) discovered one in six funds overcharges
clients because of benchmark hugging. The Financial Conduct Authority (FCA) estimates over
£100 billion of investors’ money is in benchmark-hugging funds run by apparent active
managers. This study essentially shows that by being selective, having a high active share and
being long-term, you’re already on the path to outperform your benchmark-hugging short-term
peers. But you need capacity for periods of dislocation from the benchmark.
How can we exploit career risk and loss aversion? As an independent emerging management
firm, Tollymore has a disadvantage because we’re subscale. However, we have an advantage
because we can organically build a limited partner base sympathetic with the investment
strategy that’s intellectually generous, who can have conversations with us about business
prospects at the underlying holding companies and not challenge us on why we underperform
or overperform over short periods. In a nutshell, our job as investment managers is to tread this
99
line between conviction and dogma. We want the conviction to average down. That helps us
outperform if we execute correctly. But we want to have the humility to acknowledge our
mistakes. We don’t want an environment causing us to conceal our mistakes or our ignorance.
The seventh observation concerns short-term capital. Large asset managers with an entrenched
LP base are not positioned to improve the quality of the LP base organically. They’re stuck with
it. Asset gathering mandates have typically led to unidirectional marketing processes; that is, the
manager pitches his fund, pitches his ideas to prospective allocators and current investors. That
potentially leads to situations in which the investor and manager are incompatible. That can lead
to sullied reputations in the investment management business and what is typically the case,
procyclical capital flows. You haven’t made your loss on your investments permanent until a
redemption forces you to crystallise that loss. What’s the issue? Short-term capital leads to short
holding periods, and again, pressure to predict share price movements. Annual incentives at
large investment management firms cause fund managers to try to lock in their bonuses, to make
short-term investment decisions, and to trade frequently. Average holding periods are now
around eight months.
Having a performance fee-led structure avoids the temptation to risk substantial assets. We
smaller managers have an opportunity to create organically an aligned LP base. That aligned LP
base allows us to buy securities from sellers who are selling for non-fundamental reasons. That
could be because they have procyclical investors, and therefore, they’re funding redemptions, or
because they think the news flow will be negative over the short term. That’s because they have
eight-month holding periods. Avoid these two situations by having a sympathetic, aligned LP
base and the capacity to invest for the long-term.
Communication style is my final observation. Attention spans in this industry are incredibly
short. There’s constant competition for attention. When I was writing research on the sell side,
no matter how long the research note was, the most important part was a three- or four-line
eloquent investment thesis on the front. That’s what my clients, if anything, read. I have limited
time, either on the phone with my clients, at meetings with fund managers, or within the
investment bank, to communicate my thesis, to convince other people of my ideas, to have them
believe me.
In a trading firm in an investment bank, there’s a podium with a microphone. When a company
reported results or when I wrote a piece of research, I jumped on the microphone. I literally
shouted over a massive room of hundreds of sales traders to get their attention.
Analysts on the buy side also try to get the attention of their PMs. They have limited time to do
this. The typical practice in the industry involves writing monthly letters offering a backward-
looking introspection of what happened to markets and why. The letters reduce complex
systems to simple eloquent theses.
There is a problem about stories in this industry. In life, stories are useful. They help us to make
sense of the world. Perhaps you’ve read the book, Built to Last, by Jim Collins. It’s a nicely written,
100
entertaining book in which he dissects the sources of several outperforming companies. The
conflation with the historical causality of that analysis with its predictive power would have
caused you to potentially make some grave investment errors because around half of those
companies, since he wrote that book, have suffered an enormous fall from grace. Making stories
compelling might be entertaining, but it’s a barrier to finding the truth. Stories are simple. The
world is complex. Stories tend to describe outcomes to talent and stupidity while they ignore
luck.
The literature refers to this as “narrative fallacy.” It’s our constant strive to attribute a cause and
effect chain to historical events. This is useful in everyday life, but it can cause us to think in
terms that violate logic. In his book, Thinking Fast and Slow, Daniel Kahneman gave an example
in which he posed a question to several participants. The subject of the question was Linda, age
31. She was single, outspoken, and bright. She majored in philosophy. As a student, she was
deeply concerned with issues of discrimination and social justice, and she participated in anti-
nuclear demonstrations. We asked which alternative is more probable? Is Linda a bank teller, or
is Linda a bank teller active in the feminist movement? Despite the second alternative including
all the information of the first and additional information, therefore being the illogical choice,
90% of undergraduates chose the second option. That’s the power of stories.
How can we avoid and exploit this? Try to write involved, detailed, long shelf life letters to
partners. We investment managers try to create not only an investment track record, but a
decision-making track record, a journal of our decisions. Letters can show our current investors
and our prospective allocators how we think. They can give them comfort that we’re doing what
we say we’re doing. They can give them comfort that our process is consistent with our stated
investment philosophy.
There’s little value in writing retrospectively about market or stock price movements. It’s
reductive. Ideally, write infrequently or irregularly. Write when there’s something to write about.
Avoid pitching. This is incredibly difficult, but write in almost a bland way. Here are the facts.
Here are the strands of logic. Here’s the data that support a set of conclusions. A reader of those
facts, data, and logic, if they come to the same conclusion as you and you presented that
information objectively, is more powerful than pitching them an eloquent thesis that leads them
to merely believe because of your sales skills.
How do we profit from this? Actively seek out stocks with bad stories. They might be
opportunities the market is extrapolating from negative news flow. For me, the essence of value
versus glamour investing is what James Montier discusses. It’s finding the stocks, the
opportunities where the investment merits are attractive, but where the stories and the news
flow is negative.
The following are excerpts of the Q&A:
Participant: Mark, you just started your own fund. What advice would you give to someone
who’s just starting his own fund? What should he make sure of before jumping into this?
101
Walker: One is to recognise the capacity to do well is limited, so you’ve got to love this for the
intellectual exercise. To build a fund, you need to build an investment track record and a
decision-making track record. You need to build trust with a series of hopefully aligned
investment partners, and you need to embark on a journey of building relationships with those
aligned investment partners. For me, it was important to enjoy that journey and enjoy those
conversations for what they were so that I wasn’t going to look back in five years and regret
having spent all my time in building these relationships. Enjoy them for what they are. When
you’re a subscale investment manager, the temptation to take on assets through potentially the
wrong type of capital is high. Try to be disciplined about the capital you take on, about the claim
on your resources those investors might take. Think about what they might add to your
investment organisation in addition to capital. Some thoughtful investors and family offices have
a value philosophy and investment teams who, for me as a single-person investment committee,
their time and their intellectual generosity is valuable to me outside of the capital that they can
offer me. In the early days, I would suggest any desire to be rich is not a good source of
inspiration. You will typically earn a lot less doing this than you can by being part of a large fund.
It’s a hugely scalable business model.
Participant: Why did you choose to start your own fund rather than work for a small fund you
admired?
Walker: What you’ll find is that it’s important that everyone develops their own investment
philosophy and their own principles and process that you think are consistent. No matter how
aligned you are with a small fund, there will always be some inconsistencies. Temperamentally,
I’m someone who enjoys collaboration but not a level of forced collaboration. Working in even
small teams, I find it can be frustrating for me as a barrier to progress either in the investment
management business or in investing in the portfolio. Aside from that, for the type of investing I
wanted to do, for the investment philosophy I had, there’s an incredibly tiny number of suitably
aligned investment firms I would consider. Those firms are small firms with low levels of churn,
long tenures of experience, and talented people. Getting opportunities are not always there.
Several of my peers are open about the fact they don’t like marketing. They’re horrible
marketeers, and they love business and investment, or they love the practice of investing. I quite
like the process of building relationships with potentially aligned partners, of building
something, building a brand, building a business, like a project outside of an investment
portfolio as well.
Participant: I completely agree with your approach about not looking at the market price, the
noise, and to completely focus on the business. But then, given that most of the funds across the
board are focused on that, how do you suggest we present our facts and figures? How do you
suggest we go about it, because not everybody can start their own fund, like yourself, and
succeed?
Walker: It’s not right for everyone. I’m just sharing my experience and the way smaller
independent managers potentially have an edge over larger asset managers. There are smaller
102
independent managers that sympathise with this approach. It’s about being cognisant of the
differences in the investment approach, about simply asking a question in your mind or
explicitly to someone that you’re interviewing with about how they think about aligning their
interest. It’s about how they think about decision-making, how they think about the elements of
the process that are consistent with an investment philosophy. Closed funds are generally more
able to do this. They have fewer marketing pressures, and they intend to have a quite sticky
investor base. I encourage you to try to find compatible investment organisations to work with,
to ensure there’s a two-way dialogue in the relationship building process, and that you’re not
just trying to pitch yourself as a good employee without asking any difficult questions about how
they conduct themselves.
It took me quite a long time, if I’m being honest, to be introspective enough to interrogate my
own personality and temperament. Typically, in the early days of my career, I worked for big
organisations, and I was an introvert in an extrovert’s business. I was shoehorning my personality
into something that would fit the mould of a large organisation, adorning this gregarious,
charismatic personality and then finding myself in working environments that suit gregarious
personalities with open plan, noisy, trading floor environments. I found myself increasingly
going to a quiet room to try to carve out time for independent thinking. I’m not suggesting one is
better than the other, but a few people are genuinely introspective about what makes them tick
and how many people are caught up in this process of building a shiny CV. Warren Buffett refers
to this as an inner scorecard versus an external scorecard. The sooner you recognise that, the
more useful it is in your career choices.
Participant: Instead of wasting time trying to get the right price by forecasting, what else do you
suggest? What are the tools you suggest we use to make our case?
Walker: We are in the business of understanding the long-term prospects of businesses. That is,
for us, a forward-looking exercise. When I talk about the value of forecasting, I’m talking about
the futility of precisely estimating a complex range of outputs. Rather than build a financial
model where, line by line, you’re going through volumes, and pricing, and variable and fixed
cost, capital intensity, and tax rates, capital structure in 50 different markets for a company, and
you’re trying to do that in a quarterly basis and getting to an EPS estimate, then you’re putting in
a multiple on the EPS estimate. I suggest that is a low ROI exercise.
I prefer to do a lot of fundamental work to answer a handful of questions. In terms of a business’s

financial future, I tend to try to understand what I call “owner earnings” or the “maintenance-free
cash flow” of the business. If the business were to stand still today and invest what it needed to
invest to maintain its unit output and its competitive position, how much cash can you put in the
pockets of owners of that business today? That’s a starting point that reflects a normalization
exercise of reported earnings. For free cash flow, it reflects an understanding of what the normal
working capital intensity or maintenance versus growth capex of a business is. It’s a reflection of
the cyclicality of the business. That’s the normalisation process. Then I try to understand the
103
sustainable economic returns on capital and to what extent we reinvest that capital into valuable
projects for a long time.
If you have these owner earnings of 100, how much of that 100 will be reinvested into projects?
What will they earn? Will they earn the cost of capital? Will they earn materially more than the
cost of capital? Does the business enjoy lasting unfair advantages offering super-normal high
profit capacity and profit sustainability? That ultimately gets me to what is a prospective IRR on
an equity investment today. What are owner earnings as a percentage of the market cap? That’s
the yield. Say that’s 10%. Say management will reinvest all of that into projects with incremental
yields of 20%. I should be able to get a 20%-plus IRR on an equity investment barring any
material derailing of the owner earnings of the business. It’s a simple framework, but there’s a lot
of fundamental work required to understand those numbers. If my opportunity cost of investing
is, say, 10%, I want something materially higher. Tollymore has a concentrated portfolio. It holds
10 to 15 securities. It has a global remit, so I could invest in 50,000 companies. Therefore, I’m
typically fully invested. I don’t have cash because I should be able to, at any point in the cycle,
find 10 to 15 of these securities with an IRR materially higher than an opportunity cost at any one
time.
Participant: Can you talk about the companies which screen well? They’ve got great return on
capital, great earning growth historically, but, fundamentally, they might be a bit weak.
Walker: I don’t do a lot of screening. I’m a fan of the opposite of what you were asking about.
Companies that might screen well are cyclical companies a lot of the time. Cyclical companies
might be cheap when their earnings multiples are high. That’s one reason I don’t use screens.
Cyclical companies have high returns on capital at the top of a cycle, but in the absence of
sustainable competitive advantages, those returns on capital revert. That’s a value trap. I suggest
a good majority of screening based on returns and cheapness throws up those types of
companies. It’s the corollary I’ve been interested in and several companies in the portfolio
exhibit, which is companies that screen terrible generally because their reinvestment rates are
extremely high. Companies with negative free cash flow but attractive operating cash flow yield
and enterprise value are attractive. Their capacity to invest in projects is high. I want them to
have as much growth capex as possible if the capital structure of the business can support it.
High reinvestment rates can explain why they screen poorly. It could be accounting measures,
the accounting treatment of amortisation of quasi-permanent assets, for example. It can be the
reason they screen poorly and the reason the economic, financial picture of a company is
different than the reported accounting picture of a company.
We’re in a world where you typically see some proprietary screening process. This is how we
generate ideas. They look like a funnel. By and large, they’re not proprietary, and by and large
most people in the industry do screening. Maybe, unfortunately for me, having a more
serendipitous process, which is like engaging with management teams, reading filings, engaging
with my peers on the buy side and opening myself up to serendipitous conversations with
companies and industries, isn’t a mechanism to find ideas. That is perfectly fine if you’re
104
passionate and competent. It’s not marketable, I can’t put that succinctly and eloquently on a
slide and show it to an investor, but it’s a good source of investment edge.
Participant: Mark, you spoke about focusing on a handful of questions in an investment idea.
How do you get to this handful of questions? How do you identify them?
Walker: The difference is they’re not the same handful of questions other than when it comes to
valuation. I want to understand the normal earnings part of the company. I want to understand
the reinvestment rate and the potential economic incremental returns in capital. When I
conduct due diligence on a company, I have a number of headings where I go through filings,
speak to management, speak to competitors, speak to peers, read industry investing
publications, and try to gather facts and data to help me form a conclusion on a number of
aspects in a research report. Those aspects are business simplicity. What is the business model?
Can I get my arms around this business quickly? Is it complex? Does it have a simple
organisational structure? Does it have complicated associates and minor interests?
Second, what is the capital structure of the business? How does it finance its activities? What are
the uses and sources of capital? Is the capital structure appropriate? Are there a lot of off-balance
sheet items? Does it carry the leverage given the recurring nature of the revenues? Another is
competitive positioning. Is there evidence for a moat? What are the potential sources of the
moat? Is the moat widening or narrowing? Another is growth opportunity, a decent return on
investment. What is the addressable market of the business? For how long can it keep
reinvesting in projects? Does it have adjacencies it can invest in and expand the addressable
market? There’s stewardship. How are the decision-makers in the company aligned to make
decisions consistent with value creation for owners of the company? Is there insider ownership
in the company? Does the long-term incentive plan reflect management’s ability to make
potentially near-term difficult decisions for long-term value creation? Or do near-term decisions
make a company’s reported financials look weak but create value on the surface?
It’s a fact-finding mission. I might get to a point where I just don’t understand the business; it fails
immediately at the business simplicity standard, and I just move on. I might get to a point where
I’m uncomfortable with the stewardship of the company or the competitive positioning of the
company to an extent that I don’t think it’s merited to continue doing research. I might get to a
point where I write a complete research report and love the company and love the management
and its stewardship and its prospects, but the recent valuation, the recent owner earnings yield
doesn’t give me a good shot at an attractive IRR; therefore, it might go on a watchlist over time.
Those things change.
Participant: What is the limit to the number of holdings while also maintaining an information
advantage? Related to that, how long does it take you to get to know a company well?
Walker: It depends. For me, I feel if I have a portfolio of 20 to 25 names, I would struggle to know
those assets well enough to make optimal portfolio management decisions. Diversification to me
is not a mechanism to lower risk. If that diversification is inadequately resourced – take it to an
105
extreme example, if you’re one manager managing a one hundred-stock portfolio, your ability to
know a lot about those businesses is limited. Therefore, your ability to have conviction in what
they’re worth is limited. Therefore, your ability to act based on share price volatility is limited. In
terms of getting to know the businesses, that depends. I’ve invested after a couple of weeks of
work on one business. Some unique circumstances caused the asset to come to market at an
unduly distressed level. It was a simple business to get my arms around.
Sometimes, like the last company I invested in the portfolio, I looked at it for about six months.
The stock price declined by about 50%, and I did a little more work. Nothing had changed. It
didn’t require too much more work before investing in the company. It’s unusual for me to go
straight into an average-sized position. For me, an average-sized position is 7% or 8%. The
smallest equity positions typically approximate 5% unless I feel exceptionally comfortable with
the business model. Usually, I invest in negative stories and carry with them the potential for
further share price declines. For that reason and because six months is still a short time to get to
know a company, and because of the long holding periods my portfolio has, getting to know
these businesses is important in a multiyear, potentially multi-decade context. As I grow
comfortable with the business and its potential and its management, I can potentially increase
the size of that holding.
Participant: Do you go in with your desired sizing? Or do you scale in? How do you do that?
Walker: The last portfolio addition was at a 5% position.
Participant: Did you scale that over time? Or did you just go in?
Walker: I went straight in at 5% because it’s concentrated, but that’s the smallest position in the
portfolio.
Participant: It doesn’t create any market impact?
Walker: No. When I say going in at 5%, I’m not necessarily going in in half an hour. What I mean
is, I’m not going in at 1%, and then a year later, maybe increase it to 3%. When portfolios are
concentrated and you have a global or a large opportunity set, the question is, is this so massively
undervalued that it should be in the portfolio? If that’s the case, put it in and put it in
meaningfully. Or if it is not, don’t put it in.
Participant: Do you lock your investors’ access for a period because you’re so concentrated?
Walker: No. My investors can have access to their capital whenever they want. I don’t want to
create unhappy investors. There may be external circumstances outside of Tollymore that lead
investors to sell positions. I want them to be able to have access to it. There is a process of
education and relationship-building about the merits of long-term investing, and investors get it.
They don’t always act consistently with what they say. That’s just part of being an investment
manager.
106
Participant: You mentioned short attention spans and communication style. Could you give us
examples of how you manage to capture people’s attention in your letters?
Walker: I was suggesting you avoid the temptation to try to grab people’s attention. The
incentives in this industry are to create sound bites that truncate messaging and communication.
They’re meant to grab people’s attention. The incentive to pitch is overwhelming for investment
managers. That is a barrier to good decision-making. There’s a lady called Annie Duke who wrote
a book called Thinking in Bets. She was a professional poker player. I met her recently, and she
gave me simple advice about how to make better decisions. If you want to open your work up to
intelligent interrogation and analytical inquiry, you should not impart your opinion on
whoever’s receiving the information. It will influence their perception of the problem. It will
make them less inclined to give you their dissenting opinion, which might be valuable. Their
dissenting opinion might cause you to do more work. The additional work might lead you to
acquire more conviction in your opinion or change your mind. You should be able to change
your mind.
When I wrote research for the sell side, I created eloquent theses. I’m not saying that’s all there
was. There was proper fundamental work underlying the research, but the focus was on
grabbing people’s attention. The focus on the microphone was on grabbing traders’ attention.
You’re in competition with several other sources of stimuli devoted to the goal of grabbing
others’ attention. I suggest having a self-selection mechanism for the right types of investors by
writing involved letters. I have a lot of information on my website. My last letter was 17 pages. I
include many interviews, articles, and presentations. Therefore, if I have a conversation with
investors who have gone through all that material, that is a higher intention conversation and
potentially a more valuable relationship. By making people do the work and jump through
hoops, you’re almost creating this self-selection solution to creating the right relationships with
the right investment partners.
Participant: Mark, from the investor’s point of view, you’ve got into a lot of points about what to
be wary of. Do you have some red flags when you’re receiving pitches or looking at firms for us as
potential investors to managers?
Walker: Here’s the barrier: When I speak to investors, high net worth people, their family offices,
or their advisers, we have a conversation about the things to look for and how they allocate the
liquid public market capital of their portfolios. There are things we have discussed that they’re
thinking about like alignment of interest, thinking about under what circumstances the
investment manager is getting rich. They’re thinking about under what circumstances the
investor is getting rich. What are the incentive frameworks for them to make decisions? Do they
have someone tapping on their shoulder, warning them that their deviation from a benchmark is
not tolerated by the internal risk policy of the company?
Technically, they get all this, but my barrier is the IBM effect, or the McKinsey effect. No one will
get fired for recommending McKinsey to do a consulting project. If they do a bad job, they won’t
get fired. If they recommend an unknown consultancy firm or if the company hires an unknown
107
auditing firm and is a fraud, that person will get into a lot of trouble. If they hire Deloitte, they
probably won’t. My problem is that even though they recognise a thoughtful way to think about
enriching the providers of capital over the long-term and there’s a less thoughtful way that is a
by-product of some of these behavioural constraints that we’ve discussed – despite recognising
that – they’re open about this McKinsey effect, the idea that allocating Tollymore Investment
Partners capital is a career risk decision for them. It’s my job to get their attention and be patient
about building a multi-year relationship and trust profile that gradually overcomes that
McKinsey effect.
Participant: How did you get over the infrastructure barrier?
Walker: Infrastructure requirements and the compliance requirements are onerous in the U.K.
compared, for example, to the U.S. My challenge was in trying to develop a proposition and an
infrastructure that is the right trade-off between credibility and cost effectiveness. On one hand,
I want quality, credible counterparties. On the other hand, I want to be able to survive as long as
possible in a position of being subscale and coming back to not being under pressure to make
stressful investment or business decisions about Tollymore and who you’re partners with.
Tollymore is effectively an appointed representative of a directly authorized FCA firm. That firm
has a board of directors with career compliance people who fulfil my regulatory compliance
obligations and help me to stay on the right side of the FCA. That’s one option. The other option
involves hiring a full-time compliance person. That’s a costly source of fixed expense for a newly
formed independent manager. Another option is that you go down the road of becoming
directly FCA authorized, and I will then manage that myself. I’m not a lawyer, and I don’t want to
be spending all my time managing compliance. The FCA’s handbook is enormous. I set it up by
outsourcing several operational functions. I have legal counsel, accountants, and a compliance
firm. My primary responsibilities are managing the portfolio, marketing the fund, and building
relationships with other peers. This is no small undertaking, but it is incredibly enjoyable.
108
A letter to partners, June 2020: four unpopular opinions
Dear partners,
Tollymore’s raison d’être
Tollymore is a partnership which manages capital on behalf of its principals and a small, special
cohort of investment partners who have demonstrated the ability to think unconventionally and
act countercyclically. In doing so we make decisions in the interests of long-term results. We
expect most of these investment results to come from the internal earnings power growth of the
companies we hold, augmented by occasional and sensible portfolio management decisions.
We do not expect superior investment results to come from any IQ advantage, but from the
implementation of factors that will allow us to acknowledge ignorance or exercise conviction
better than our peers. These factors relate to the consistency of investment horizon,
temperament, working environment, incentives, and investment partners’ beliefs and actions.
We believe in the incomparable importance of flourishing relationships in determining enduring

contentment. We try to uncover win-win in investing and life.
Four unpopular opinions
1. Holding cash is imprudent
“Far more money has been lost by investors preparing for corrections, or trying to anticipate
corrections, than has been lost in corrections themselves.”
Peter Lynch
2020 has so far been one of the most tumultuous periods in public market history. These
episodes are reliable opportunities for clear long-term thinkers to exploit quoted price volatility
to improve future investment results. Yet misdirected intellectual energy in the form of market
commentary tends to accelerate in periods of market dislocation. The inability to act decisively
in these periods, due to capital redemptions, inappropriate capacity constraints, committee-led
decision-making models, lack of conviction or other source of self-doubt is a barrier to value
creation and one of the great shames of our industry.
We have been struck by the broad consternation, and even irritation, at the recovery of equity
markets since the nadir in March. A recovery which many market commentators have professed
to be disconnected from economic reality. In a scramble to overlay current circumstances over
historical precedents78, market strategists and the media have concluded that the market is
78The utility of this has always escaped us. Every crisis or crash has been unanticipated. Therefore, they are not
predictable. Yet the practice of incorporating prior known unknowns into future investment decisions is common.
109
dangerously ignoring economic fundamentals. It may be interesting to consider the incentives
behind such frustration, particularly for those investors who use cash to time markets, and have
been walloped by a double whammy of having fully invested portfolios in the first quarter and
meaningful cash holdings in the second. These cash flows are part of the reflexive nature of
markets; they contribute to the very factors that drive them. Tollymore’s concentrated portfolio
and global opportunity set obviate the need to use a cash weighting as a tool to predict stock
market movements. A fully invested portfolio is better for profiting from non-fundamental
exacerbation of quoted price changes. Increasing cash allocations without demonstrably reliable
market insight is not lowering risk. Especially when the trigger for its subsequent redeployment
is typically lower uncertainty, the cost of which is higher prices. The uncontested premise that
this sell-low buy-high behaviour is consistent with risk mitigation is staggering. We contend that
those with fully invested mandates can profit from it.
It is this ‘de-risking’ behaviour that depresses prices and creates opportunities, some of which we
described in our March 2020 letter to partners. The economic consequences of the
circumstances that led to this selling come after. And it seems sensible to assume that future
price movements will be a function of (1) the second derivative; that is, the relative severity of
economic conditions relative to expectations, not relative to the past, and (2) the perceived level
of uncertainty, which declines as unknown unknowns become known. To expect weak markets
to coincide with weak economies vastly oversimplifies a complex system and contradicts
financial history. To scoff at the disconnection between the two, to assume that the aggregate
sentiment of the market is missing something so obvious that it is being highlighted by every
pundit and commentator, is hubristic and naïve. We can do better by being adequately self-
aware to respect the complexity of market and economic forecasting and acknowledge the
terrible base rate of success in this endeavour.
2. Dispassionately cutting unprofitable investments is consistent with long-term ownership
"The best money managers are also the best quitters. They quit early and they quit often. As soon
as they see things turning for the worse, they don't wait around, they bail."
Scott Fearon
We are comfortable in averaging down when collapsing share prices clearly improve potential
future investment results. But one of the challenges of a fully invested investment programme is
identifying the most profitable source of funding for this activity. Without cash holdings, we
have two funding options: reduce our ownership of successful investments or exit unsuccessful
ones. The former may be the theoretically sound choice. Rising stock prices lower prospective
stock returns, all else equal. Yet all else is rarely equal, and it is often a mistake to materially pare
back investments in companies which have demonstrated strong fundamental business
progress. We have become increasingly comfortable with utilising the latter funding source; that
Reasoning by analogy creates a false sense of security; reasoning from first principles is harder work, but in our view
the only way to judge the investment merits of individual businesses under conditions of uncertainty.
110
is, to distinguish between those losing investments that deserve our patient capital, and those
that do not.
In the absence of clear and substantial overvaluation, selling due to valuation implies an ability to
predict near-term price movements. In addition, selling a familiar business making positive
fundamental strides for an unfamiliar one creates reinvestment risk. We are not playing for 20%
upside; we will continue to own businesses which we believe will be worth a lot more in five to
ten years. We expect Tollymore’s aggregate long-term results to be determined by a small
number of outsized winners and a tong tail of (many) below opportunity cost mistakes.
3. ESG ≠ sustainable investing
“For business to survive and prosper, it must create real long-term value in society through
principled behaviour.”
Charles Koch
Institutional investment management is broken. Asset gathering business objectives and gold-
plated cost structures magnify the imperative to grow AuM. Investment firms led by marketers
rather than investment managers create strategies tailored to what will sell rather than what
works. Large pools of management fees are required to fund large, expensive, teams designed to
convey analytical edge. Pressures to justify high management fees discourage investment
professionals from acknowledging ignorance or mistakes and create an action bias that is
antithetical to good investment outcomes.
Money management is a very scalable business model. It is possible to raise significant capital
without commensurate increases in resources, rapidly increasing the profitability of the
investment firm and substantially enriching its owners. Managers’ and investors’ fortunes are
typically not aligned. This is a barrier to sound investment decision making. A simple way to
weed out the managers that back themselves is to consider the presence and power of
incentives. Managers without insider ownership are less incentivised to limit the size of their
fund, therefore limiting their achievable investment results. In our view, and in the case of most
institutional money managers, the components of stewardship reflect a product to be sold rather
than a strong belief in the strategy.
There are ways to enrich the owners of an asset management business that are consistent with
compounding investors’ capital. Incentive principles should allow for the enrichment of GPs
together with, and not at the expense of, LPs. The manager’s compensation should be driven by
the compounding of his own capital in lieu of fees on additional capital. Sharing of investment
returns in excess of appropriate hurdles lowers pressure to grow beyond the strategy’s
investment capacity. The creation and promotion of investment strategies that will sell result in
niche investment universes predicated on analytical edge, primary research methods, or
proprietary idea generation funnels, all of which look ‘differentiated’ in a pitchbook. The latest
incarnation of this pitchbook mentality comes in the form of ESG investing mandates.
111
Investors last year ploughed a record $21bn into ‘socially-responsible’ investment funds in the
US, almost quadrupling the rate of inflows in 2018. The surge in popularity of companies with
the best social, environmental, and governance scores in recent times has resulted in a crop of
ESG funds, eager to attract some of the cyclical capital flows to this bucket.
Attempts to quantify art through an obsession with measurement create bubbles, disincentivise
first principles thinking, and make capital allocation and manager selection processes less
efficient by making them more data driven. But data can be falsely empowering, unaccompanied
by thoughtful qualitative review of a manager’s capacity and incentives to do a good job.
The shoehorning of ESG frameworks into existing asset management programmes and the surge
in new ESG funds extend the already warped incentives in the investment management industry.
Stocks with strong ESG scores are bid up as these capital flows are put to work. Are investment
managers, guided by the desire to grow assets and the employment of a quantitative ESG scoring
framework, and absent the facility to think independently about the sustainability and
reasonable governance of companies, and the investment merits of their equity, being socially
responsible by directing endowments’, charities’ and pension funds’ capital to this more richly
valued subset of the global investment universe?
ESG’s commonly accepted synonymity with ‘sustainable investing’ is especially perplexing. Who
wants to own unsustainable businesses? What is investing if not the purchase of part-ownership
of sustainable, flourishing enterprises? What long-term business owner, managing capital for
investors with long-term, sometimes perpetual, investment horizons, would like the company
she owns to be run by dishonest, self-serving managers misaligned with company owners, to
treat employees poorly, create negative externalities, and exploit its customers? The studious
long-term investor must consider the company’s relationship with all its stakeholders. At a
minimum, understanding the strength of these relationships is central to mitigating risk of
permanent capital erosion. Ideally, these relationships confer a lasting unfair business advantage,
difficult to replicate by peers merely paying lip service to regulators and investors requiring the
completion of ESG checklists. The items on these checklists must be measurable. This
encourages the quantification of factors which are qualitative. How can one quantify, or score,
whether, and the extent to which, social media companies exploit their users to the detriment of
their health? Can these checklists consider and measure the enormous and compounding
positive externalities created by modern platform businesses? An appreciation of the
relationship between a company and its stakeholders is a qualitative, subtle, and effortful
endeavour. The interrogation of business ethics and the capacity for company managers to make
the right long-term decisions are at the heart of serious investment programmes. The
employment of an outside analytics firm to bestow credibility on an ESG framework should be
no part of this.
Any investor with an interest in owning sustainable businesses must recognise that company
boards have broader responsibilities to consider environmental and social issues. They will also
understand that these duties do not dilute their fiduciary obligations to equity holders. Rather,
112
they are a necessary part of satisfying them. Decisions to create large positive externalities and
forge win-win outcomes – non-zero-sumness – take time and often come at the expense of
traditional short-term measures of shareholder return. But these are the decisions that
determine corporate vitality and staying power.
The broad disregard for long-term win-win outcomes is a function of short holding periods.
Company ownership changes hands on average every few months, vs. eight years in the 1960s.
This is a by-product of investment constraints deeply embedded in the institutional money
management industry, such as short-term capital, manager/investor misalignment, career risk
and asset-growth agendas. These constraints cannot be addressed with an ESG checklist.
Just as truly special corporations seek to thrive under the principle of win-win, so must
investment management firms create more value that they consume. Incentive structures should
make it sensible to forgo additional management fees in lieu of excess returns on insider capital.
Frameworks for economic profit-sharing should reward acceptable performance and coordinate
manager and investor enrichment. These include the employment of hurdles to make weak
performance cheap and strong performance expensive. Investors should receive most of any
outperformance generated, not just most of the absolute return. Fee structures should therefore
allow fund managers to talk about integrity with some legitimacy. Too many managers charge
egregious fees with the sole purpose of enriching themselves at the expense of clients.
Tollymore enjoys trusted relationships with long-term investment partners. Partners who can
think like business owners rather than stock market traders and understand that a common
stock represents a fractional interest in real assets. Great investment partners are a competitive
advantage: an evolving, mutually appreciative, and increasingly resilient relationship is an
enabler of greater and more sustainable value creation. Our goal is to accumulate wealth by
investing in high quality businesses for the long-term; it is not to predict share price movements.
With business ownership in the public markets so fleeting, and given the strong economic
incentive to raise capital, it is unsurprising that the objective of investing in viable, durable,
companies is not evergreen nor widely practised. This is a sombre impeachment of the asset
management industry. But it creates opportunity for those investors who recognise that treating
stakeholders well is good for business owners. We look for symbiotic value chains; there need
not be a trade-off between compounding owners’ capital and the principled consideration of all
other lives our companies affect. Managers of the companies we own must look beyond the
direct or immediate consequences of their actions to create lasting value. And a long-term
outlook is essential, because shorter term sacrifices – ‘the capacity to suffer’ – are often required
to create value of any permanence.
113
4. Successful investing requires intuition
"One's investment approach [should] be intuitive and adaptive rather than fixed and
mechanistic."
Howard Marks
The emulation of successful investors is dangerous and reductive. Blindly copying the factors
that may have driven others’ success is less likely to lead to good outcomes than the
introspective examination of one’s own temperament and qualities, and a careful matching of
those qualities with our environment.
We believe a behavioural advantage is possible by coordinating the disposition of the firm’s

principals, the mentality of its investment partners, physical working environment, methods of
internal communication, the time horizon of the strategy, and the implementation of the
programme. We spend time thinking about the complementary nature of these aspects of the
ecosystem. We spend less time thinking about the individual merits of concentrated vs.
diversified portfolios, noisy vs. quiet offices, or short vs. long-term decision making. There are
lots of ways to invest, but the best results come from thinking about the interdependence of
every aspect of an investment organisation.
Investing is an art. Yet money managers attempt to quantify and measure progress and decision
making due to an incentive to raise large amounts of institutional capital through the suggestion
of repeatability. History’s greatest investors have used intuition to guide their decision making,
particularly with respect to idea generation and portfolio management. The highest achievers in
another complex, reflexive game – chess grand masters – rely heavily on intuition.
It seems popular nowadays to espouse the benefits of quelling emotion to make more rational
choices, thanks to the fantastic work of Daniel Kahneman and Amos Tversky, and studies such as
“Lessons from the Brain-Damaged Investor”79. While Tollymore is an effort to exploit the
behavioural shortcomings and constraints of peers, we are not sure it is possible or even
desirable to simply banish emotion. A life without empathy, sorrow, passion, excitement, or joy
sounds horribly unfulfilling80. Even the desire to avoid negative emotions such as shame or
regret make it more difficult to acknowledge and learn from mistakes and can lead to suboptimal
investment decisions81. Finally, attempts to suppress emotion are highly energy-depleting. This
energy is required to form the insights and make the judgements necessary for sound investing
practice.
Emotional management is clearly sensible in investing. Market cycles can boil down to
fluctuations between participants’ states of happiness and sadness. So, an ability to regulate
79 Lessons from the Brain-Damaged Investor

80 Granted efforts to be less angry seem to have little downside.
81 We must be clear about the reasons for lowering our cost basis in an investment: to improve future investment
results, or to avoid being ‘wrong’?

114
these emotions is plainly helpful. But there are other aspects to emotional intelligence such as
empathy and self-awareness that do not seem to receive the same attention. Avoiding the
discomfort associated with an honest evaluation of oneself is a barrier to self-awareness. In this
way a military focus on eliminating negative emotions can be counterproductive in making good
decisions.
The art of using and perceiving emotion is something we can become better at over time. Trying
to improve in these areas in the pursuit of a good life will also make us better investors. By
dismissing the opportunity to develop in these areas, we run the risk of abandoning intuition in
favour of purely quantitative information processing. This is likely to lower a tolerance for
uncertainty which may be one of the only advantages we have as public equity managers. This
seems especially relevant today when most ‘investment’ decisions are being made by computers.
We spend most of our time on rational reasoning through independent study of companies and
examination of their investment virtues. But fundamental business analysis is table stakes; and it
is more commoditised than money managers realise or admit. A disciplined investment process
is required to prevent otherwise helpful emotions from becoming unregulated. So too are the
people that will help us to stay humble in periods of success and support us in periods of self-
doubt.
Back to intuition. We think it is a mistake to dismiss gut feelings. They are not random but the
product of pattern recognition. They can lead to poor decision making when not subject to
analytical scrutiny. But when operating in a culture that promotes introspection, mental
flexibility, and humility, and accompanied by relevant experience, these feelings can improve
judgement. The low utilisation of intuition in investment management is not due to the dismissal
of its efficacy, but rather the difficulty in underwriting its deployment; that is – it is tough to
market a credible investment strategy that relies on ‘feelings’. With qualitative judgement in
short supply, we see great opportunity for those willing to embrace intuition to guide idea
generation and portfolio management. Neither rational analysis nor intuition are sufficient, but
both are important components of sensible investing.
Yours sincerely,
Mark
115
A letter to partners, September 2020: profiting from change
without predicting it
Dear partners,
To manage capital patiently and successfully, it is not desirable to know everything about
everything. Rather, our job is to express, through the direction of capital, insights that are
anomalous or mispriced.
Our inability to predict the future severely limits the value of anticipating unknown unknowns.
And it is unknown unknowns that move markets in aggregate. Due to the market’s efficiency in
pricing forward expectations, making investment decisions according to forecasts about known
unknowns is also a low ROI exercise. Rather, all we can do is to prepare for uncertain shocks
through a general risk framework, one that allows us to own resilient, or antifragile businesses,
with adaptive and innovative business leaders.
We are not trying to express anticipated thematic changes through our investment process. We
are not trying to react to macro or cyclical shocks by predicting winners and losers. Of course,
distressed prices due to these shocks may create opportunities, but these are unearthed via a
bottom up discovery process.
The holdings in our portfolio have different business models and economic characteristics, but
all are consistent with a goal to maximise the intrinsic value of a portfolio of companies,
principally through the compounding of each company’s enterprise value through its own
efforts.
We try to describe using first principles language whether a business is ‘high quality’ along the
vectors we use to define such a concept. Without such a technique, we might fall into two traps:
(a) we become enamoured with certain business characteristics that are likely to be overvalued
and lead to a non-diverse portfolio, and (b) we dismiss hidden gems that may be in
unfashionable industries conventionally perceived as low quality, but when examined
idiosyncratically are actually wonderful businesses.
Exploiting faulty heuristics: an example
Our ownership of Trupanion is an instructive

case study of: interrogating investment merits
from first principles; profiting from short sellers’
different investment agenda; the value of steady
business progress accompanied by fluctuating
quoted prices; and the embodiment of the
symbiotic value chain.
116
The mental shortcut
The heuristic that comes to mind for insurance companies is that they are typically mature,
capital intensive businesses whose internal investments yield a cost of capital for owners. So
embedded is this heuristic that when an insurance company’s management describes the
company in terms of recurring revenue and maximising the lifetime value of its customers, it
leads to a high level of short interest claiming a disingenuous management team
mischaracterising the business.
Short sellers’ different investment agenda.
We do not short securities. A capped upside and unlimited downside 82 are inconsistent with the
risk framework we have described in past letters. But TRUP remains one of the most shorted
stocks we own, and it serves us well to contemplate the incentives governing short seller
behaviour. Because the upside is capped, short sellers are not long-term investors. We can look
well beyond five years in the pursuit of satisfactory results. Even if a short seller believes a
company is worth zero, an investment horizon of five years would still require an annual
investment result below what we as long-term long only investors might be hoping to achieve.
Indeed, if Tollymore’s portfolio returns to date are the opportunity cost, their zero would need to
be reached in around three years (ignoring the borrow cost). If they are aiming for a 50% stock
decline, they would need to achieve this in two years, before diverting resources to the next
target. It all just seems very labour intensive. And hard.
With an understanding of short sellers’ agenda, and while recognising the behavioural risks of
being defensive83, let’s use their thesis to scrutinise our own set of insights:
“TRUP has an adverse selection problem. That is, by distributing insurance through the vets, they
are selling insurance to sick animals. This will lead to a rate spiral.” TRUP’s competitors rely on
online marketing initiatives to acquire customers. It seems logical to us that pet owners will be
searching the internet for pet insurance when their pet is injured or unwell. TRUP focuses only
on puppies and kittens, acquiring them when they first visit the vet. Those pet owners that take
out a TRUP policy to treat a pre-existing condition quickly churn out when they realise it is not
covered.
“Churn is very high”. Again, once normalising for early churn due to uncovered pre-existing
conditions, the majority of churn is explained by the pet’s average life. We think TRUP has room
to better educate customers in making it clear that pre-existing conditions are not covered. But
in general, they do a good job at keeping customers. The lack of coverage for pre-existing
conditions makes changing insurance provider costly.
82 For those still short the stock since our original purchase, they have lost more than three times their original
investment in under two years, excluding the cost to borrow the stock.
83 Surrendering the freedom to change one’s mind, objectivity, and even-keeled investment temperament.
117
“Management is overly promotional and disingenuous.” Short sellers have claimed that the CEO
has referred to TRUP as a SaaS business. This isn’t true to our knowledge. The CEO has
specifically stated that TRUP is NOT a SaaS business because it has low gross margins.
“The business is egregiously overvalued and should be valued like an insurance company”. We
agree that it should be valued like an insurance company, because it is an insurance company.
But it is growing its discretionary profit pool 30% per annum and has an RoE approaching 50%
on an owner earnings basis. So, calls for 1x BV make no sense to us. After a tripling of the stock
the P/B implies 7-8% sustainable growth. While no longer clearly deeply discounted to today’s
intrinsic value, neither is this business clearly overvalued and we continue to own the company
as a core investment.
The value of steady business progress accompanied by fluctuating quoted prices
In a decade of public documents, TRUP has never reported a quarter of sales that was lower than
the previous quarter. This is not due to extraordinary execution but the result of a monthly
recurring revenue model and a large underpenetrated addressable market. In the US around 2%
of cats and dogs are insured. In the UK, a market in which Patsy Bloom’s PetPlan business
increased penetration via a similar comprehensive product and vet distribution, one quarter of
dogs and cats are insured. In that market it took 20 years to reach 5% penetration and another 20
years to reach 25% penetration. This potential for accelerating growth is not reflected in TRUP’s
quoted price.
At the same time TRUP is one of the most volatile stocks we have owned, allowing us to
occasionally pare back or add to our ownership when the quoted price makes such actions likely
to improve our equity investment results.
The embodiment of a symbiotic value chain
A symbiotic value chain is one in which multiple stakeholders participate in a company’s value
creation; success is shared with customers, employees and owners through thoughtful and
transparent incentives. When the mission goes beyond profit maximisation, profit maximisation
is often the result. When all stakeholders are inspired by a mission, short term sacrifices in the
interests of long-term outcomes become easier. This makes for a more defensible, less replicable
business as stakeholders adhere to the mission though good times and bad.
There is something special about the incentives of TRUP’s stakeholders and the economic
characteristics of the services TRUP sells. Consider what a 70% loss ratio implies for the
customer’s value proposition. Customers are paying an average mark-up of 43% for each dollar of
premium. This would seem uncompelling for a customer seeking an ROI on their premiums. But
pet owners are clearly not seeking a return; in fact, they hope that their pets are never sick or
injured. Further, competitors’ loss ratios imply mark-ups in the order of 100%, highlighting
TRUP’s superior value, if not cheaper price.
118
Even if pet owners were seeking a return on their paid premiums, there is an asymmetry in the
value chain that makes it difficult to self-fund the medical needs of individual pets. Pet owners
are not playing a repeated game; the average experience of the average pet, even if it were
known to individual pet owners, is of limited use in underwriting the potential costs of
individual pets. TRUP on the other hand is playing a repeated game using pooled data to
accurately underwrite a group of pets. It is this aggregated data that constitutes a barrier to
profitable participation that would make it difficult for vets to offer their own insurance for
example.
Other components of synergistic value chain: (1) Trupanion Express allows claims to be paid
directly by Trupanion within minutes. This is great for TRUP as it allows access to more data from
other insurance providers. It is great for customers there is no need to settle expenses out of
pocket. And it is great for vets because it lowers economic euthanasia and allows for Plan A
treatment options, and saves credit card fees, which might be 20% of their profits. An important
insight here is that vets are paid a fraction of doctors’ salaries. Their careers are intrinsically
motivated and mission driven. (2) Enterprise value growth is shared between employees and
owners according to a transparent formula. And all employees are owners.
Multiple small insights can confer lasting unfair advantage
The discovery of these insights requires a bottom up research process and first principles
thinking. They would have been missed by the quality business investor seeking ‘network
effects’, ‘switching costs’, ‘capital light business models’, ‘SaaS’, ‘platforms’, ‘high net retention’,
‘high gross margin’, ‘WFH stocks’ or other heuristic du jour. TRUP has certainly been written off
by many investors perplexed at why an insurance company’s price to book might be in the
double digits. TRUP’s investment merits cannot be punchily summarised; unearthing them
requires effortful digging and a capacity to see the wood for the trees.
119
Learning from mistakes: an example
Betting on revisions of progress
We used to believe that buying companies that have attractive long-term prospects, but which
are facing short term, but surmountable, business problems was an attractive source of superior
investment results. This may still be the case, but our investment history has demonstrated an
inability to consistently profit from this.
The investments we made into TripAdvisor and Grubhub were, with hindsight, bets on a revision
of fundamental business progress that did not materialise. In both cases there exist winner take
most potential economic outcomes, with demonstrable barriers to entry and an owner-operator
business ethos.
Yet in both cases it was our misassumption that a monopolistic outcome was unnecessary for
outsized value creation. This was a philosophically inconsistent premise.
We considered TRIP part of a global duopoly in hotel meta. And in our view the principal
competitor to TRIP’s product offering was the large portion of travel bookings and advertising
still taking place offline. But we were too dismissive of the value that Google commands by being
right at the top of the funnel for most hotel booking experiences. From this position Google has
the power to inflate OTA and meta companies’ customer acquisition costs by replacing their
organic results with ads or Google’s own inventory occupying the most valuable top-of-page real
estate. As TRIP’s core hotel business stagnated, we were the proverbial frog in the boiling water.
Labels are bad for mental flexibility
The broader lesson here is that being a long-term investor does not mean you should not quit
when you are wrong. The long-term investor badge of honour that many of us self-righteously
parade around can really put our investment results in jeopardy by inhibiting the objective
reasoning we are so fond of telling people we
possess.
The more subtle observation from these

types of errors is that for businesses with
potentially exponential success outcomes,
oligopolistic participation in a particular
digital migration may not be enough. As
such, we expect several of our current
holdings to be the outright winners in their
respective industries.
One example is one of Tollymore’s largest holdings, Farfetch, a global luxury digital marketplace
for brands, retailers, and consumers. We acquired our interest in FTCH for $14.4/share in May
2020.
120
FTCH exists because of a compelling consumer and supplier proposition. 80% of products sold
online are in a multi-brand environment. This is evident in brands’ historic willingness to have
concessions in department stores – that is where the footfall is. Luxury brands, in return, gain
access to 2.5mn luxury customers across the globe. Given consumers’ preference for shopping in
a multi-brand environment, brands have a choice to either partner with a vertically integrated
retailer, or a marketplace. The clear preference for marketplace partnership is a function of (1)
unit economics and (2) control.
FTCH’s take rate is c. 30%. Under traditional linear industry economics, a product costing $20
may sell for $100. The retailer typically keeps most of the $80 mark-up, perhaps around two
thirds of it. The brand margin is therefore c. 25%. For a brand selling directly on FTCH, this might
double to 50%, which is what is left after the product cost and FTCH’s 30% take rate are deducted
from the retail price.
In addition to this gross margin advantage, brands can achieve incremental sales by making their
inventory available to a global luxury audience without increasing their invested capital or
diminishing their returns on capital. The high take rate is also supported by special
characteristics of luxury products – high gross profit products and a fragmented supply base
characterised by family-controlled companies seeking to protect brand integrity. By partnering
with FTCH, brands retain control over visual representation and pricing, mitigating potential
concerns around brand dilution.
The right model for industry domination
FTCH does not compete with any other large luxury marketplace. If the same pattern of
disruption happening in music, travel and communication, is repeated here, then the leading
player will be a platform, not a retailer or a brand. CEO and founder Jose Neves is playing for
winner take all. His strategy is predicated on the logical assumption that consumers will always
gravitate to one single app, forcing vendors to gravitate to one single platform. And we think that
this is most likely to be FTCH given its scale advantage and given it has found itself, through
good fortune or managerial foresight, to have the right model at the right time.
With my best wishes,
Mark
121
A letter to partners, December 2020: operating systems
Dear partners,
Our investments are a product of a bottom-up organic investment process, and an examination,
one by one, of each company’s investment merits. We are deliberately not limiting our
investment universe by restricting what we can look at. We do not create rules relating to the
business models, economic characteristics or geographic end markets our holding companies
must possess. We have a simple research process requiring our companies to meet high hurdles
for business quality as we define it. Having said this, we can make broad observations about the
features of the businesses we own. Two features have stood out to us in the context of our
preference for businesses with symbiotic value chains84: operating systems and shrinking supply
chains.
Value chain symbiosis cannot be achieved while one component’s profits are another
component’s losses. Operating systems create value for multiple parts of a value chain. By
helping one component, they can simultaneously create value for another. For example, retail
operating systems such as Shopify allow sellers to focus on doing what they love – delighting
customers by creating amazing products. Everything else, including website design and hosting,
payments, inventory management and fulfilment is taken care of through a single counterparty
relationship. By levelling the playing field with larger merchants such operating systems allow
smaller retailers to access larger markets. Customers benefit from better choice, superior
products and lower prices.
There are two limitations of Shopify as an operating system for retail. The first is its lack of B2C
brand (which of course is its appeal to brands seeking DTC routes to market) and by extension its
inability to help sellers reach the large majority of customers shopping in multi-brand
environments. This is both a demand aggregation and a marketing efficiency shortcoming. And
the second is its inability to help sellers sell offline, still the mode through which three quarters
of commerce is conducted. These limitations inhibit Shopify’s ability to optimise an end-to-end
customer experience.
This letter discusses why we think two of our holding companies are uniquely positioned to
become operating systems of their respective fields. In so doing we believe they can create vastly
more value than they consume via highly compelling supplier and customer propositions. Before
we proceed to some detailed case studies of retail operating systems, a brief note on shrinking
supply chains. When the value created by a linear supply chain is shared amongst fewer
components, win-win outcomes are more achievable. We will discuss shrinking supply chains in
a future letter to partners, with a focus on the conditions under which shortening the supply
84A symbiotic value chain is one in which multiple stakeholders participate in a company’s value creation; success is
shared with customers, employees and owners through thoughtful and transparent incentives.
122
chain can lead to substantial value creation for owners, suppliers and customers, again using
case studies of companies we own as well as those in which we decided not to invest.
The appendix to the letter is a transcript of an interview we did with Good Investing TV, in which
we discussed our business quality criteria, investing in companies with emerging moats, the
determinants of antifragility within Tollymore, as well as the individual investment cases for two
of our holdings.
Operating systems: Next plc and Farfetch Ltd
Our most recent investment has quietly but purposefully been building the means to be the
operating system of multi-brand, omnichannel retail in the UK and beyond. We acquired an
interest in Next plc (NXT:LN) for £61 per share last quarter.
Next is a multi-geography, multi-brand, omnichannel retail business. The company sells own-
branded NEXT products, as well as c. 1,000 third party brands via Next’s online aggregation
business LABEL, a £500mn business in the UK. Next has 6mn UK online customers and 2mn
overseas online customers. Since 2019 Next has also provided customers access to partner
brands’ products stocked in those brands’ own warehouses (‘Platform Plus’). This has broadened
the range of third-party brands offered. Online sales are routed through Next’s own website and
app, as well as brand.com websites. The store network is an important part of the online value
chain; half of online orders are collected from Next stores, and 80% of returns are processed
through the stores.
Next offers nextpay, a credit facility for UK online customers, and next3step, a credit account
which allows customers to spread the cost of orders over three months interest free. Next’s
£1bn+ lending book is not just a sales enabler/friction reduction tool but is also highly profitable
(60-70% return on equity).
Under ‘Total Platform’ Next makes its warehouses, call centres, distribution networks, customer
relationships, marketing engine and lending business available to third party brands, including
next day delivery services and store collections and returns. Next also builds and operates
brand.com, all in return for a 39% commission on brand partner sales.
Next has a forty-year business history of innovation and successful strategic pivots including
pioneering out of town stores when competitors remain focused on the high street, expanding
apparel categories from adult wear to children’s clothes and homewares, and expanding beyond
own-label online to a multi-brand marketplace, and now becoming the operating system for
third party brands via Total Platform. Throughout this period Next has earned consistently high
operating margins in the high teens, three to four times higher than high street market leader
Marks & Spencer, or online pure plays ASOS and Zalando. One can see how continued progress
in widening Next’s SKU variety advantage might drive increasingly attractive basket economics
and higher customer loyalty.
123
The value created by Next over the last fifteen years against a backdrop of the rapid proliferation
of ecommerce pure plays, supermarkets’ success in penetrating the UK discount apparel market
and declining high street footfall, has been commendable; free cash flow per share has
compounded in the mid to high teens. One could have bought the business for £4bn 15 years
ago. Since then, Next would have paid the owner roughly that amount in dividends, and today
the business could be sold in the public markets for more than double this purchase price.
Distinctively positioned to be the operating system of multi-brand omnichannel retail.
Today, Next is perhaps uniquely positioned to invest in and benefit from delighted customers
and suppliers. The strength of the supplier proposition strengthens the consumer proposition
and vice versa. This positions Next well to create more value that it consumes.
Online shopping has two characteristics that distinguish it from traditional brick and mortar
retailing: (1) product choice: customers can now access products from all over the world, vs. a
few physical shopping locations; and (2) convenience: customers can now have goods shipped
directly to their door.
We can clearly observe the increases in online penetration of retail sales across the world. But as
the retail industry is divided between largely pure play online players, and largely physical
retailers, it may be difficult to ascribe relative value to these two propositions. Next is perhaps
the only scaled multichannel home and clothing retailer in the UK. The behaviour of its
customers is therefore instructive; half of online orders are delivered to Next stores. This might
suggest that SKU variety is the most valuable advantage that online shopping has over brick and
mortar. But there is another reason: for many, delivery to store is more convenient. In Northern
Ireland, delivery is free, yet 28% of orders are still delivered to stores. This would suggest that
over half of customers picking up in store are doing so due to convenience rather than cost.
Then there are returns. Rather than attempt to suppress returns rates, Next has embraced the
customer’s right to return unwanted items. Again, the stores play an important role in providing
an efficient returns process. In some categories of online apparel sales return rates are as high as
50%. And for Next 80% of returns are conducted through stores.
The store network is part of the distribution infrastructure; store stock operates as an online
reserve. Stock can be delivered from store to home when there is no warehouse inventory, or
stock can be moved from store A to store B for collection by the customer. This omnichannel
aspect of the customer proposition may be more defensible than crude subsidised shipping
strategies. And the ability to collect in store is more valuable in the context of a multi-brand
shopping service – there are not many places in which one can shop across many different
stores and pick up once in one convenient location.
Two conclusions might reasonably be drawn from this observation. The first is that online
penetration has some way to go before it stabilises, as product choice is a more sustainable
advantage vs. offline than home delivery. The second is that retailers with a physical store
124
footprint may, by avoiding the material costs of last mile delivery, be best positioned to offer the
most valuable consumer proposition and the most profitable route to market for third party
brands.
The supplier proposition is the provision of the most profitable route to market. The internet has
lowered the barriers to entry for brands seeking market access by removing the upfront cash
commitment associated with retail stores and lowering the operating leverage of such
businesses. And e-commerce platforms such as Shopify have reduced these barriers further by
turning yet more fixed costs into variable expenses for online brands. But given that 80% of retail
shopping takes place in multi-brand environments, brands can access many more customers by
trading on aggregation marketplaces such as Next.
Our assumption that the multi-brand omnichannel shopping proposition is valuable is

supported by most customers choosing to pick up and return in store and most customers
shopping in multiband (online and offline) environments. In addition, many of the world’s most
successful pure play ecommerce businesses, such as Amazon, Alibaba, JD.com, Wayfair and
Farfetch have been investing in the integration of offline and online shopping experiences.
If we accept this consumer preference for shopping in multi-brand environments, and the
sustainable product choice advantage of online, it seems that two categories of sellers will find it
hard to generate and retain value for owners. The first is established physical retailers, and the
second is brand.com destinations lacking the brand equity required to stand alone and compete
for the 20% of consumers shopping in such places.
Next’s traditional business falls in to the first high risk category: a high street retailer with
burdensome fixed cash obligations such as operating leases and wages. Next’s investments in
next.co.uk over the last 20 years are evidence of customer-centricity. But its more recent efforts
to help competing brands increase their addressable markets and safeguard their own
businesses against the more level playing field in an online world seem to be particularly forward
looking. And necessary. Since 2014 Next has featured third party brands alongside its own Next
and Lipsy brands through its online platform LABEL. The first order implication of this move has
been to increase competition with, and potentially cannibalise sales of, Next’s higher margin
own brands.
But in the words of CEO Simon Wolfson “There is nowhere to hide on the internet, one way or
another our customers will find the brands they want. If they can find what they want on our
website they are more likely to come back to us, furthering our ambition to be our customers’
first choice for clothing and homeware online.”
Since the introduction of Platform Plus two years ago, the number of third-party brands available
on the Next platform has doubled to 1,000, and Next’s active online customer growth has
accelerated from mid to high single digit rates to mid-twenties in calendar 2019 and 2020, despite
the larger base. More recently, visits to next.co.uk have more than trebled since the first
125
lockdown in the UK at the end of March 2020. This is the highest growth rate amongst
established ecommerce peers such as H&M, Zara and Asos.
LABEL was originally a wholesale business model, but today more than half of this business is
done on a commission basis. This may increase as the rate of growth of brands paying a take rate
is three time those selling wholesale. LABEL represents just c. 13% of total Next sales currently.
In return for a flat all-inclusive 39% take rate brands receive the following benefits: 8mn active
customers, growing c. 25% pa.; price control – Next is a full price retailer – 15% of sales are on
markdown, vs. the industry average of 35-40%; Next day delivery network including returns and
collections from stores; and a large-scale consumer lending proposition.
The timing of LABEL’s quiet but notable business progress is fortuitous; it positions Next to gain
from the COVID lockdown-induced restructuring of the UK apparel industry. Cross-shop
between Next and retailers in company voluntary arrangements or administration such as
Debenhams and Arcadia Group is high. These retailers represent a tenth of UK market capacity,
vs. Next at c.7%, including LABEL.
In addition to the omnichannel distribution advantage, but much more nascent, is the
opportunity to use online information about consumer shopping habits and preferences to
optimise in store inventory and customer service. In store collection and returns also drives
footfall and therefore incremental selling opportunities.
Avenues for highly accretive internal reinvestment
The economics of online capacity growth are attractive. The current book value of Next’s online
warehouse and distribution plant and machinery assets carries annual depreciation of c. £20mn,
1% of online sales. Next is halfway through a six-year warehouse and logistics investment
programme costing £300mn; this will increase annual online sales capacity by £1.7bn. So, 18p of
investment is increasing annual sales capacity by £1. And the principal growth driver of online,
the LABEL aggregation business, carries no inventory and enjoys negative working capital.
Next’s multi-brand offering is exploiting historic distribution and marketing investments. Next is
leveraging existing assets to expand into close adjacencies. And as the scale of the business
increases, so too does the utility and value of Next Unlimited, which offers unlimited free home
delivery to customers for £20 per year.
The LABEL penetration opportunity is vast. LABEL sales still represent just 1% of the UK apparel
market, vs. 6% for the NEXT brand alone. Sales per LABEL brand are also still less than £1mn pa.
Given the strength of the LABEL proposition in a world in which consumers overwhelmingly
prefer to shop in multi-brand environments, engines of business growth may include adding
more brands, the expansion of product ranges for each brand that are available on LABEL, and
extensions into adjacent categories such as Beauty and Home. LABEL still sells mostly apparel;
home and beauty together are just 15% of revenues. The online penetration of beauty products is
126
still less than 10% in the UK, but the high value, small size and low returns rate of beauty
products make them well suited for online sales. Homeware online penetration is c. 15%, around
half of the proportion of apparel and footwear sold online.
Manufacturing is a nascent growth opportunity. Retail operating systems can further integrate
themselves into retail value chains by reorganising their partners’ supply chains; that is, by
connecting merchants with manufacturers. Next sees an opportunity to leverage its sourcing
experience to create and distribute partner brands’ products. In 2020 it signed such licencing
agreements with Ted Baker, Oasis, Joules and Scion Living. These opportunities, and the
economic terms underpinning them, may only improve for Next given the COVID-induced
failures of large retailers.
Online business models enable geographic expansion. As rent is replaced with digital marketing
dollars, Next’s online assets have allowed the company to profitably enter new regions,
unencumbered by the customer density required to economically justify physical store
investments. Online sales of c. £600mn are growing 25%+ pa as Next adds >1mn new overseas
customers per year. NEXT brand accounts for c. 85% of overseas revenue and has been growing
20% pa. However, after initially modest take up, LABEL brands have started to exhibit
meaningful growth, +c. 70% yoy as the range of brands has increased substantially.
This overseas growth is valuable. Digital marketing spend has been doubling annually, with
sound return on investment; the business is generating £1.53 in incremental orders within the
first year for every £1 spent on digital marketing. Assuming a 40% gross profit margin implies a
20-month gross profit payback on each £1 spent on marketing. Overseas customers spend 20%
more each year on average; a customer spending £100 in year one spends >£350 p.a. in year
seven. Despite the attractive unit economics on digital marketing initiatives, almost 90% of new
customers are still acquired organically.
Total Platform is the most nascent, and potentially most valuable opportunity. The LABEL
leadership team is responsible for developing client relationships for Total Platform, leveraging
pre-existing relationships with LABEL partners. Next’s first Total Platform customer was
Childsplay Clothing, which sells luxury childrenswear. While www.childsplayclothing.co.uk
looks like an independent brand, part of the Total Platform proposition is to reduce friction of
signing up to a new website. Next account owners can sign in using their Next login credentials
and Next credit customers can pay using nextpay.
Next’s CEO summarised his thoughts on the prospects for Total Platform as follows:
“Total Platform is potentially a very exciting business because what it does is it takes a partner
brand and it allows them to plug in to 20 years' worth of investment that NEXT has made in its
web systems, its marketing, its online warehousing, its distribution, whether that distribution be
through couriers, through NEXT stores, through overseas networks, it plugs them into our call
centres and gives them our credit offer. And that integration allows them to have a much more
powerful and robust presence both in terms of U.K. online presence, their online presence
127
overseas where they can tap into all the work that we have done to develop 70 different
international websites for our business. We can put their product into those countries on
websites that are translated, that deal in local currency and it takes specialist forms of local
tender types, for example, cash on delivery in Russia.
They also have the advantage, if they want it, of being able to deliver their stock through our
stores. We know that 50% of our orders are picked up in our stores. So, we think, the use of
NEXT stores as a post office is actually a very powerful marketing tool, even more powerful as an
easy way to return stock.”
Next’s brand aggregation business LABEL has made this possible. Next has been delivering
LABEL stock from hundreds of third-party brands into its own warehouse for six years. To
integrate that stock into Next’s system, it is given a single unique identifier. This identifier allows
Next to know who bought returned items and how they bought them so that customers can be
refunded accurately, regardless of whether returned online or in a Next or brand partner store.
Contrast the alternative available to brand.com seeking to grow sales profitability via
outsourcing aspects of the value chain independently. A new website built from scratch needs to
be integrated into the seller’s warehouses, credit systems, customer account systems and call
centres.
The cornerstone of the Total Platform proposition is the reduction of business risk facilitated by
swapping fixed costs for a single commission structure. A commission structure which aligns the
interests of Next and the partner brand. 2020, perhaps more than any year in recent history, may
have brought the benefits of such a cost structure into sharp focus for retailers mandated to shut
their doors all over the globe. And even in times of rapid growth, Total Platform eliminates the
growing pains of step-ups in fixed costs and major capital investment projects to expand
capacity and customer support. In this way Next’s vertical integration can lower partner brands’
capital intensity. The utility of business resilience is particularly high in the volatile fashion
industry.
As Next scales this business, the proposition becomes increasingly compelling due to the size
difference between Next and the partner brands. A 50% step change in call centre or warehouse
capacity for a partner brand might be achieved by a mere 1% increase in Next’s volumes. In
addition, Next’s longstanding reputation, physical high street presence and high customer
satisfaction may serve to lower Next’s costs of customer acquisition and retention vs. third party
brand partners, particularly smaller and emerging brands for whom Next’s Total Platform
proposition may be especially compelling. It also allows newer and emerging brands to focus on
what they find interesting – designing and creating great products, and delighting customers,
rather than the unseen, but crucial, components of a retail supply chain, such as data security,
call centres or warehousing.
Capital will be increasingly directed to servicing online demand. Management forecasts

modestly lower capex over the next five years vs. the previous five. But warehouse and systems
128
capital requirements are set to rise dramatically with store capex falling to levels commensurate
with maintaining, rather than expanding, the store estate.
So, what kind of returns can these internal investments generate? 18p of capacity investment is
increasing annual sales capacity by £1, and the operating margin on online business is running at
c. 18%. Incremental returns of c. 100% are attractive indeed. Of course, this fully loads the
operating expense base into the new profit pools generated from these investments. If we use
the group gross margin of 40%, incremental returns are north of 200%. Achievable rates are
likely between these two percentages; the point is that incremental returns are very high, and
capital can be reinvested at substantially above opportunity cost levels. Next currently earns
returns on non-lending capital employed of more than 100%.
As for increasing the lending book, this too is valuable. Next uses group borrowings which cost
3-4% to lend to customers at a rate of 24%. Next generates almost £300mn in revenue pa from its
£1.2bn lending book. Assuming its current 85% debt funding mix, it earns c. 70% RoEs in this
business. The low cost of servicing this business is due to that fact that it serves existing retail
customers, three quarters of whom have been shopping with Next for more than five years.
Now for the bad news. We hope the priorities for the use of cash in this business will change. For
15 years Next has been repurchasing shares and issuing special dividends due to a limitation on
the amount of capital that can be internally deployed into value accretive projects. Today, Next
can use that capital to help smaller, more capital constrained brands with a higher funding cost,
and in so doing become the operating system of multi-brand, omnichannel retail. For many
publicly quoted UK businesses, capital allocation agendas are set by a low-quality institutional
investor base requiring sustained annual dividends to satisfy their marketing-led investment
remits.
Prior to 2019, Next’s preference for returning excess cash to shareholders was a result of its
inability to deploy capital profitably in overseas markets. But two years ago, that changed. The
company is getting better at targeting those customers who will like Next’s clothes. And Next has
found distribution partners capable of delivering high quality service at attractive unit
economics. The result is that from 2019 Next has been earning supernormal incremental returns
on investments in these areas; c.100% IRR for UK digital marketing campaigns and almost 400%
for Overseas campaigns. Extraordinary circumstances such as these provide the perfect
opportunity to reset the capital allocation agenda and really lean into the online opportunity
that few others are so well positioned to exploit.
Scale is a more important driver of success in online vs. offline business models, especially for
working capital light marketplace models. Scale drives advertising efficiencies: large scale,
reputable retailers will win Google and Facebook auctions with lower bids due to better quality
scores – that is, high customer satisfaction and strong brands increase click through rates. Scale
also lowers shipping costs via volume discounts on last mile delivery. Scale begets scale; this is
responsible for the winner take most market structures of most online categories.
129
Unfortunately comments from management such as “We recognise the importance of providing
our shareholders with consistent and reliable dividend returns” and “We remain committed to
paying dividends to our shareholders once the crisis has passed and do not anticipate any long-
term change to our dividend policy” may mean that catering the existing shareholder register
inhibits total EVA that can be generated by the company over the long-term.
Given highly able management and the huge optionality embedded in this business; that is, the
opportunity to become the operating system of multi brand, omnichannel retail, why pay
dividends and buy back shares? The IRR on these activities is likely to be low compared to
internal business investments that could lead to quite extraordinary industry dominance. This
may be particularly true at this juncture, in such a period of competitor distress and disarray.
See for example the considerable business progress of Farfetch, progress that accelerated in
moments of maximum stress, such as 2009 and 2020. That company’s capital allocation agenda is
designed to widen a set of unfair business advantages so that their assets become increasingly
hard to replicate. It is this type of anti-fragility that we see with Next. A company whose
competitive position has been strengthened by COVID-induced retail business failures and
accelerated online shopping habits.
Perhaps we have misappraised this opportunity. Or perhaps capital allocation choices at Next are
governed by factors other than simply long-term business value compounding. Presupposing
directionally right analysis of Next’s opportunities for value accretive investments, capital
returns to shareholders could diminish the value of optionality embedded in Next’s equity price.
Dividends are typically paid by businesses caught on the wrong side of time. Next could be a
special business for the future.
Next’s market cap was £8.5bn when we acquired our interest; the quoted value of the enterprise
including finance receivables was £9.5bn, vs. FY20 EBITDA of £1.1bn, or 8.5x EV/EBITDA. In FY20
Next generated £6 in FCF per share; we paid a 9% yield for our shares. Of the c. £800mn Next
generates in FCF, around £130-140mn is spent on sales and marketing. We estimate Next
comfortably generates more than £1bn p.a. in pre-tax owner earnings, a 12% yield to current
price available to equity owners.
This valuation, which anticipates no growth or value neutral reinvestments, together with Next’s
outstanding track record of economic progression and shareholder value creation amidst a
challenged industry backdrop, mitigate against substantial loss of capital for equity owners.
Meanwhile Next may be uniquely placed to leverage existing assets to exploit optionality and
become the operating system of multichannel omnichannel apparel, beauty and homeware. And
to do so on terms that are extremely valuable to owners, customers and sellers alike.
One of Tollymore’s largest holdings, Farfetch (FTCH), is a global luxury digital marketplace for
brands, retailers, and consumers. The marketplace connects 2.5mn active consumers in 200
countries to 1.3k sellers across 50 countries, including 500 direct brand e-concessions. FTCH is
the biggest online destination for luxury in the world.
130
The consumer proposition is predominantly range of merchandise and curation of supply. But
also localised websites, multilingual customer support and same day delivery in 18 major cities.
Accessibility to global products is evidenced by consistently >90% of transactions being cross-
border. Shopping for luxury products via FTCH’s website or app enables customers to search by
designer, category or keyword, is available in multiple languages and accepts multiple payment
methods.
80% of products sold online are in a multi-brand environment. This is evident in brands’ historic
willingness to have e-concessions in department stores – that is where the footfall is. But the
value of a fashion marketplace goes beyond choice. As a marketplace, rather than an inventory-
owning retailer, FTCH can better curate products for consumers. FTCH can more rapidly test and
iterate messaging, content and SKUs; purchasing inventory would slow this down. A large SKU
count also makes it easier to satisfy consumers’ desire for ‘newness’ via daily product additions
and the ability to offer new/more extravagant products without inventory risk.
A compelling supplier proposition seems to be evidenced by: the large number and strong
growth of brands and boutiques on the platform; that 98% of retailers have done so exclusively;
and FTCH has retained all top 100 retailers and all but one top 100 brands over the last five years
despite strong supply growth.
What are the reasons for this? Access to 2.5mn customers across the globe, on a fully managed
basis; that is, everything from content creation to last mile delivery. An online offering helps
brands reach new audiences, increase visibility, and expand their presence into new
geographical locations without the need to invest in retail infrastructure.
The price for this is a 30% take rate to FTCH. This is high by the standards of marketplace
business models. But it seems to economically be a better deal for brands, no longer required to
pay the retailer margin. Under traditional linear industry economics, a product costing $20 may
sell for $100. The retailer typically keeps most of the $80 mark-up, perhaps around two thirds of
it. The brand margin is therefore c. 25%. For a brand selling directly on FTCH, this might double
to 50%, what is left after the product cost and FTCH’s 30% take rate are deducted from the retail
price. Brands can achieve incremental sales by making their inventory available to a global
luxury audience without increasing their invested capital or diminishing their returns on capital.
Brands have full control over visual representation and pricing. FTCH allows access to a targeted
luxury customer base, mitigating potential concerns around brand dilution. Increasingly FTCH is
overcoming challenges faced by emerging brands such as brand exposure and limited reach.
FTCH’s marketplace offers a place alongside leading luxury brands, lending credibility to the
emerging brand and providing access to luxury consumers, who, in turn, are attracted by the
opportunity to discover under-the-radar, exclusive, emerging brands. This potential flywheel has
opened because of FTCH’s purchase of New Guards Group, a brand incubator. The market
reacted to this acquisition by cutting the stock in half, a price decline that piqued our initial
interest. The NGG acquisition helps FTCH to provide more choice, including exclusive products
from emerging brands. That is, to drive engagement through limited supply.
131
Like Next plc, Farfetch is embracing the omnichannel advantage. But unlike Next, FTCH’s
distribution is largely decentralised and offline sales are made via partners’ stores. FTCH has c.
$5bn of inventory available to its customers, which has two channels. For the online channel
FTCH represents most online sales for sellers. There is also a compelling offline advantage to
being on the FTCH platform: you become a distribution partner. That is, FTCH will distribute
through those fashion boutiques that are on their platform, sending footfall into the stores,
creating offline selling opportunities not available to non-FTCH partners.
Customer orders are directed to a boutique based on proximity, cost of delivery and fulfilment
record. This model is scaling successfully due to the absence of strong local incumbents. FTCH
has not needed to build supply and demand in each market; the distributed model arises from
the value of SKU range. It is also resilient: 85% of SKUs are available from multiple sellers within
FTCH’s supplier network. FTCH can therefore tap supply from wherever it is. However, it is more
expensive than centralised logistics, in which multiple items ordered on the FTCH website arrive
in one box when processed through a Fulfilment by Farfetch facility. FTCH has six Fulfilment by
Farfetch warehouses, which are 3PL and therefore have no associated capex obligation. The
main 10-20% of SKUs are available in these centralised warehouses.
The high take rate is supported by high gross profit products and a fragmented supply base
characterised by family-controlled companies seeking to protect brand integrity. In addition,
there are some more subtle characteristics of the luxury fashion industry that may be creating
special unit economics. Typically, ecommerce marketplaces grow scale and compete partly on
price. This is not how to build a sustainable ecosystem in the luxury industry, in which decisions
are a function of emotion, personal identity and scarcity. This translates to a $600 AOV, a 30%
take rate, and mature cohorts delivering > $100 per order contribution/55% order contribution
margin. LTV to CAC payback of fewer than six months makes investing to grow the active
customer base by 50% each year the right capital allocation choice. Finding luxury customers is a
costly endeavour, but when they are found there is an immediate payback on CAC. FTCH has run
promotions in the past, but today seems to be increasingly focused on increasing the loyalty and
therefore value of the customer base and allowing the brand full control over price setting. This
seems consistent with sustainable value creation vs. ephemeral but expensive revenue support.
Is this a symbiotic value chain? That is, is there a non-zero-sumness to the sharing of value
creation in which suppliers, customers and owners enjoy win-win-win dynamics? It is not
immediately obvious FTCH passes this test. The superior economics of FTCH to brands may
result in pressure to disintermediate boutiques. There seems to be a clearly compelling
proposition for brands and consumers, but the proposition for retailers seems to be less clear. In
particular, the margin that needs to be surrendered (and ceded to brands) by being on FTCH’s
platform vs. the traditional linear industry model. In return for this margin dilution the retailer
receives a range of valuable services such as access to a global luxury audience (which also
increases the probability of full price sell through), data insights, logistics, outsourced customer
service, and the omnichannel offline footfall described earlier. But given the opportunity for the
132
brand to disintermediate the boutique and take all the margin, why would a retailer happily sign
up for this, and what sustains their future in the value chain under a marketplace model?
More brands seem to be bypassing retail partners and exploring a direct-to-consumer model.
The reasons: larger profit margins and stronger customer relationships/insight. Brands have
increased in FTCH’s mix, now 50:50, a trend which management expects to continue.
One aspect of value chain symbiosis is FTCH’s supply chain capabilities provided to platform
partners, such as content creation and global fulfilment, integrating global logistics partners in a
single interface. Thus, as with Next plc, there are interesting comparisons to both Amazon and
Shopify here.
The following anecdote, from Frederic Court of Felix Capital, FTCH’s earliest VC investors,
suggests the presence of a non-zero-sum value chain:
“I am often asked for anecdotes about Farfetch and the story so far, there are many special
moments but here is one that resonated particularly strongly when it happened, at one of the
company’s “Gatherings”. José had pioneered the event early in the life of the company, to cement
the relationship with our key partners, and create a stronger sense of community, gathering all
the boutique owners and brands to Porto or London, once or twice a year for a couple of days
(hundreds of people connecting). This was a great opportunity to meet many boutiques and
brand owners, to explain the development of the Farfetch platform, and get feedback from users.
A couple of years into the investment, one of the first boutique owners to have worked with
Farfetch (and still a partner to this date) came to me at a Gathering and said: “thank you, for
investing in Farfetch and saving my store”. Without Farfetch his store would have closed
because of the recession and he was grateful that we had funded Farfetch, which ended up
rescuing his business from the recession. This was an emotional moment illustrating how critical
the platform had been for many boutiques at difficult times, reinforcing the mission of the
company, and all its stakeholders, to connect fashion lovers and make those products more
accessible”.
Unfair advantages are numerous and growing.
Network effects are two sided and global. More brands and boutiques on the marketplace
increase the choices available to consumers, and more consumers increases the potential sales
for sellers. The network effects are global. That is, the total global SKU variety is a factor in the
consumer proposition, and the total luxury consumer base is a relevant factor in determining
value to a seller. This makes these global network effects more difficult for entrants to overcome
via clustering strategies. Consumers prefer to shop in a multi-brand environment, and FTCH
offers more brands to consumers than any other luxury destination. In addition, the economics
of distributing via FTCH are more appealing to brands than retail, which typically will take half of
the retail price as a gross profit, vs. FTCH’s 30% take rate. Finally, they retain control over how
their products are presented and priced with FTCH vs. a multi-brand retailer. The appeal of
control is particularly strong in luxury because in times of depressed cyclical demand, retailers
133
will seek to unload stock via discounting which erodes luxury brand value. Luxury brands
therefore seeking to complement their brand.com presence with a multi-brand channel find
FTCH to be the most appealing option. As more brands make this choice FTCH’s multi-brand
proposition strengthens. Recent evidence supports the notion that this advantage is widening:
FTCH’s top 10 brands, which includes Gucci, Prada and Fendi, have more than doubled their
direct stock on FTCH yoy. And on the demand side active customers are increasing >50% yoy.
Fragmented supply increases value of a marketplace: 19 of the 20 largest luxury fashion brands
by revenue are in Europe yet address global demand. Large brands access demand by building
expansive networks of directly operated stores and through department stores. Emerging brands
typically have no route to the global market. They rely on wholesale distribution through a
network of independent fashion boutiques. As a result, luxury fashion inventory, from both
larger and smaller brands, is distributed across a highly fragmented network of luxury sellers.
FTCH has c. $5bn of third-party inventory (not owned but available for sale) sitting in thousands
of stores across the >50 countries from which supply is sourced, leading to multiple
combinations of shipping routes and logistics providers. The e-concession model means that the
same warehouses that serve brand.com serve FTCH. This happens with 500 concessions and
FTCH taps directly into the brand’s inventory without acquiring it. Thus, some of the
acceleration in revenue that FTCH has recently experienced was a result of these brands
prioritising their direct-to-consumer sales.
Established partner relationships are hard to dislodge. In seeking to preserve brand integrity,
luxury brands have historically been reluctant to partner with an online retailer. FTCH has
developed relationships with the leading luxury families to overcome this reluctance by, for
example, allowing control over visual representation and pricing. As a result, these brands have
agreed to partner with FTCH on an exclusive basis – 80% of items listed are exclusive to FTCH –
and FTCH has 8x the number of SKUs of its nearest competitor. It is now likely much more
difficult for a new entrant prise these relationships away.
FTCH is the only scaled luxury marketplace. However, it faces competition from technology
companies enabling ecommerce e.g., Shopify; online luxury retailers which hold inventory and
ship from centralised warehouses, such as YNAP.com; and multichannel retailers. Richemont,
the owner of YNAP, removed YNAP founder and CEO Federico Marchetti in March 2020 amidst
reported frustration in YNAP’s lack of profits. With seeming competitive disarray, FTCH
continues to invest ever increasing dollars at widening its business model advantage.
There is another subtle but significant advantage of marketplace over retail: FTCH’s marketplace
model confers an important and widening advantage vs. linear retail peers. FTCH’s technology is
integrated with the stockpiles of its sellers. FTCH therefore has access to data from traffic and
sales not only made through the FTCH platform, but also offline in its boutique and brand
partners’ stores. This information can be used to make FTCH’s services more valuable to more
engaged consumers, but also to provide aggregated feedback to sellers.
134
Like Next, Farfetch is investing in further integrating itself into the luxury value chain through
Farfetch Platform Solutions (FPS), a white label ecommerce offering. On the front end, FPS
creates global websites, apps or WeChat stores. Back-end services allow retailers and brands to
synchronize their websites with in-store and warehouse inventory, both from mono-brand
stores and other suppliers in their distribution network and facilitate in-store pick-up and
consumer returns. Layered on top are services such as marketing, localisation, production and
warehousing. These white label solutions are being provided for Harrods in the UK and 20 other
clients.
Also like Next, FTCH is demonstrating anti-fragile business qualities. FTCH’s boutique suppliers
have faced mandated retail store closures due to COVID 19 restrictions, and have suffered from a
material decline in tourists, particularly from China. FTCH’s revenue growth accelerated in 2Q20
as these boutique partners sought to access luxury fashion consumers across the globe. COVID
has helped to increase demand too: FTCH saw a spike in app downloads in the first quarter,
especially in China in which app downloads more than doubled yoy. Chinese consumers
represent more than one third of luxury consumption, of which typically 70% is made while
traveling. That is $70 bn of personal luxury goods purchased by Chinese nationals while traveling
outside Mainland China. Demand, which, with all else being equal, seems to have been
repatriated online with huge drops in international travel in 2020. Likewise, FTCH’s FPS
customers, such as Harrods, were able to continue serving customers during lockdowns.
Expect high reinvestment rates and incremental returns.
FTCH has a large, growing addressable market driven by online penetration, opportunities to
carry adjacent products, exceptional unit economics and evidence of operating leverage.
Luxury fashion transactions are increasingly online. Over the last decade online penetration of
luxury sales was c. 2% and has grown 20-25% pa. According to Bain online now accounts for 12%
of the global $330bn luxury sales market, with customers increasingly influenced and enabled by
digital channels, including in their physical purchases. 75% of luxury transactions were
influenced by the online channel, and almost a quarter of purchases were digitally enabled. Bain
expects online to become the number one channel with 28-30% market share by 2025.
Bain expects the global luxury market to be worth >€350bn by 2025, driven by growth in Chinese
consumers, online channel and consumers younger than 45. This might imply a >$30bn online
commission pool, c. 85x FTCH’s revenues. FTCH’s current share of the global market for personal
luxury goods is c. 0.3% of the estimated total addressable market.
FTCH has the right model for industry domination. FTCH does not compete with any other large
luxury marketplace. Its competitors are online brand ecommerce, omnichannel multi-brand
stores and online multi-brand retailers. Jose Neves is playing for winner take all: “I believe a
single company will orchestrate this revolution in the conversion of offline and online luxury
retail because, even if multiple retail-tech vendors emerge, the new technology will have to be
adopted both by retailers and consumers. We believe consumers will always gravitate to one
135
single app, forcing vendors to gravitate to one single platform, most likely a platform that has
already built consumer-side critical mass and benefits the entire ecosystem. This all translates
into a potential $450 billion addressable market for Farfetch, which, as the operating system for
luxury, we want to transform, empowering individuality for consumers, curators and creators of
fashion.”
Despite a clear understanding of FTCH’s advantages, attractive unit economics and long runway
for demand generation investments, management does not have a ‘growth at all costs’ mindset.
Jose and Jordan have frequently reiterated their focus on protecting margins and strongly
growing the profitability of the business, which they expect to be EBITDA break even this year.
So far, the demonstrable operating leverage in the business lends credibility to these ideals.
Despite more than quadrupling in value since we originally acquired shares at $14, the rates of
returns we expect to earn on our ownership remain high. An estimated $500mn of owner
earnings are being invested into demand generation efforts that are yielding materials ROIs.
Assuming a three-year customer lifetime and one order per year would imply incremental
returns north of 30% and an IRR on the equity of $20bn north of 20%. Assuming a five-year
customer lifetime and one order per year would imply incremental returns north of 50% and an
IRR on the equity of around 40%.
The quoted price remains inconsistent with the deep reality. FTCH has a large and growing
addressable market, a differentiated business model, high barriers to profitable participation, a
win-win value chain, great unit economics and a capable, aligned leader.
Thank you for your partnership,
With my best wishes.
Mark
136
Appendix: Interview with Good Investing TV
The following transcript is from Tollymore’s interview with Good Investing TV in September
2020, which is available to watch here.
Tilman Versch 00:05
Hello everyone! Welcome back to our livestream session. This time I’m very happy to have Mark
Walker of Tollymore here. Hello, Mark! How are you doing today?
Mark Walker 00:18
Hi Tilman, I’m really good. Thank you for inviting me.
Tilman Versch 00:22
You are located in London, right?
Mark Walker 00:26
Yes, my offices are in London. But I actually just moved my family out of London to a place called
Kent, which is just outside. Because we decided that a global pandemic wasn’t really stressful
enough for us. So, we decided to move house at the same time to spice things up a little bit. So,
I’m actually located at my house in Kent at the moment. But our offices are located in the centre
of London.
Tilman Versch 00:53
Brexit is also a topic we should discuss during our live stream. I also want to say “Hello” to all of
our viewers. As always, you are welcome to drop your questions into the chat box so I can pick
them up and pose them to Mark during our livestream. Mark, before I show the disclaimer, I want
to give you a question you can think about for a moment. The question is: What is good
investment for you? You have some time to think about that and for everyone else, you can again
enjoy the disclaimer. You can find it below in the links section. The main message is: “Do your
own work, this is no investment advice. We are having a qualified conversation. You always have
to do your own research and that is also the interesting part”. So, without further ado, what is
your answer, Mark?
Mark Walker 01:56
Well, first of all, thanks again, Tilman, for asking me to have this conversation. I think what you
do is in a clear spirit of intellectual generosity. I’ve had this uneasy sense of embarrassment at
being part of an industry that I’ve felt for quite a long time now has really done a pretty bad job
for its customers, which are the owners of capital. What you and people like you are doing in a
very value-added way is to shine a light on those individuals, those investors, those managers,
that community that are trying to forge a different path, which in many ways is non-institutional.
137
They are trying to think carefully and thoughtfully about how to set themselves up to create
value over time. And so, this light that you are shining on them and the support that you are
giving them is very, very valuable. Because there are lots of barriers to small thoughtful managers
becoming sustainable. So, I just want to thank you on behalf of the community for that work that
you are doing.
Tilman Versch 03:20
Thank you. Thank you as well for coming here and sharing your insights and your wisdom.
The qualities of a good investment
Mark Walker 03:26
So, what makes a good investment? That is a huge question. I think that it’s helpful to understand
what Tollymore’s central philosophy is. It’s also instructive at the outset to clarify that
I don’t think that there is anything unique or exotic about this philosophy and I certainly
don’t consider it a source of edge. It’s a philosophy that has three principles.
1. One is that we are trying to exploit a time arbitrage over time in the public markets with a
very simple goal of maximizing intrinsic value of a group of assets over a long period of
time. What that typically involves and the reason why that is through a mechanism of
public equity investing rather than private business ownership is that we can buy
securities from sellers who are selling for non-fundamental reasons. Because they have
not set themselves up, they don’t have an ecosystem that’s consistent with executing a
long-term strategy. The obvious examples might be because they have misaligned LPs or
they have short investment horizons because of all manner of constraints on their ability
to make good decisions. And so, we want to buy the securities from those people.
2. Secondly, the implementation of that philosophy, the goal of maximizing intrinsic value
of a group of companies is predominantly through ownership of companies that can
compound their own values through their own efforts over the long run. That
necessitates a level of self-awareness and recognition that we might not be that great at
actually managing a portfolio, but if we can pick the businesses that can do the work for
us, that is a less labour-intensive and more efficacious route to a good outcome.
3. The third aspect of our philosophy, our goal is to maximize intrinsic value of a group of
companies. If our goal is to be in this IRR hall of fame, it’s absolutely imperative that we
understand and cater for how we permanently lose money. And so, there is a risk
framework that informs what we are and what we are not, that is from two angles that we
consider permanent loss of capital to be a potential outcome. One is through the holding
companies themselves and we look for businesses that have a number of lasting unfair
advantages, that is that they can protect and hopefully grow their enterprise value over
time. Secondly, we look for those companies who are appropriately capitalized so that
138
that intrinsic value growth over time accrues to owners rather than lenders. Finally, we
are trying to acquire those businesses at prices, which means that our IRR as an equity
owner of the business is at least as good as the intrinsic value compounding of the
company.
That is the overall philosophy, but as we might discuss, my experience to date has led me to
believe that there is a sort of lollapalooza of institutional constraint to just executing that
philosophy, even if the rhetoric of long-term investing is well-understood and well-versed. You
asked, “What makes the best investments?” Well, we’ve made a distinction over time in obvious
versus non-obvious business excellence. Frankly, we’ve done a pretty poor job of generating
acceptable results by investing in obvious excellence. Principally because that excellence is
generally reflected in asset prices. What I find is that it is often these kinds of multiple small,
mispriced insights that compound to form a business which is very defensible and very difficult
to replicate. The discovery of those multiple small insights really requires a bottom-up organic
idiosyncratic investment process.
What I find is that it is often these kinds of multiple small, mispriced insights that over time
compound to form a business which is very defensible and very difficult to replicate. The
discovery of those multiple small insights really requires a bottom-up organic idiosyncratic
investment process.
Those opportunities are really what is missed by some investors who label themselves as quality
managers, because they are inappropriately attaching labels such as ‘network effects’ or
‘platforms’ or ‘recurring revenues’ or ‘SaaS’ or ‘working from home stocks’ or whatever the
heuristic of the day might be. And so, what’s really crucial to us in unearthing these non-obvious
moats is a first principles approach to understanding business characteristics and unit
economics and the description in a first principles way in our research which leads us to
conclusion of whether or not this business is of high or low quality along the vectors we use to
define that. That is much broader than simply competitive positioning, for example.
Long-term investing
Tilman Versch 09:05
You mentioned the word ‘long-term’. How do you use it in your practical approach in investing?
What is long-term for you?
Mark Walker 09:15
One of the things I talked about was that we have a risk framework which means that long-term
is a risk mitigant, because it allows us to focus on those businesses that are very high quality and
be reasonably agnostic as to over which period of time the value is created and we can be very
patient for long-term holders of companies to realize that value. One implication of that is that
we for example don’t short securities.
139
I think that having an unlimited downside and uncapped upside is inconsistent with our ability
or our willingness to take a view, which in some cases is very well beyond five years. I think if you
are a short seller and you consider that a business is worth zero, effectively, then you need to be
right on a horizon which is shorter than five years if you are willing to accept an opportunity that
cost of capital which we are willing to accept. In fact, if Tollymore’s historical results are your
opportunity cost, then that is much shorter. If you are trying to achieve an outcome that is 50%
lower than the current price, again your horizon maybe turns into 18 or 24 months. To us, that is
a very labour-intensive and hard thing to do. We’ve learned over time to try not to disrupt the
positive fundamental business progress of the companies that we own. That has also then
informed the source of capital that we have used to average down in businesses whose quarter
prices aren’t doing so well over time.
Origin of the name Tollymore
Tilman Versch 11:12
What is the history of Tollymore and what is the meaning of the castle you have behind you?
Mark Walker 11:19
I grew up in Belfast in Northern Ireland and Tollymore is a forest that our extended family spent
a lot of time growing up in. It’s a really beautiful place, it’s at the foot of the Mourne Mountains in
Northern Ireland and overlooks the Irish sea. It’s a place that is at the tip of the hat from me to my
heritage. It’s something that holds very happy memories for me. There is no real clever
association with the business of investment management.
Tollymore’s success in 2020
Tilman Versch 12:05
There doesn’t need to be! You did a good performance until 2020. When I looked in your fact
sheets and letters, I saw this performance; it was good. But in 2020, your performance was even
better. What happened for you this year?
Mark Walker 12:26
While being very cognizant of trying to extrapolate or interpret short-term results in this way,
I’d say 2020 has been a really important year for Tollymore. Because what our investors are doing
is underwriting my business judgment and underwriting my investment judgment. But they are
fully aware that the business of investing, the process of investment learning is never finished.
We are hopefully on a treadmill forever, as long as our mental and physical capacity allows us to
do so. And so, this is a period of rapid learning for a lot of people, I think. While Tollymore has
had a consistent investment philosophy and a process, that helps us to govern our behaviour on
a day-to-day basis in a manner that is consistent with that philosophy. I’m very aware that that is
something that needs to evolve over time.
140
Investment decisions that we have made over time have evolved and the sources of investment
results have evolved. Some examples, I think that earlier this year was the most volatile market in
history. Clearly, it was quite a stressful time, but it was a really comforting time for me,
because I’ve worked quite hard over the last few years for Tollymore to be one of these very
defensible non-replicable anti-fragile businesses in which I seek to invest. Some of the ways that
I have sought to do that were to think about incentives within Tollymore. I think about the
components of Tollymore’s ecosystem, which include the physical working environment, the
working partners that we involve ourselves with and our investment partners, our LPs. And so,
I’ve tried to create a set of LPs who are very temperamentally aligned with the objectives. They
are very unconventional by the standards of institutional active asset management. They have
the capacity to certainly think and hopefully act in a counter-cyclical way. I talked before about
an investment philosophy not being exotic or unique, but what I’ve observed in my more
institutional experience prior to Tollymore is that there are so many constraints in large
institutional active money managers. Whether that is the imperative to grow assets, whether that
is the division of labour between portfolio managers and analysts, whether is specialization
according to sectors or geographies, whether that is internal incentives and bonus structures,
whether that is lack of insider ownership, whether that is misaligned LPs; there are just so many
that prohibit the ability to make sensible decisions.
One example is that in the face of quoted price volatility, as firm managers, have we set ourselves
up? Do we have an ecosystem that allows us to either acknowledge our ignorance and the fact
that we actually don’t know what we think we know about this investment opportunity? Or does
it allow us to recognize that we do know, and that the quoted price decline presents an
opportunity to improve our IRR by lowering our cost basis and we can exercise conviction?
In March and April this year, as you might imagine, we frankly didn’t need to look outside of the
portfolio for distressed investment opportunities. Because there were plenty within the portfolio
from which to pick. Some of our businesses were down 65-75% in a period of 4-5 weeks. A huge
source of comfort to me was that our capacity was appropriate for the strategy. This is a capacity
constrained mandate. Our investment partners are willing and able to think and act counter
cyclically and able to not only not redeem but to step up to the plate and lean in. And so, what
I realized was a real shame of asset managers; when they bend to the overwhelming incentive to
grow assets, an incredibly scalable business model, what that leads to is when we have these
retrospectively seemingly gifts, these opportunities to acquire businesses at lower prices, they
are constrained from doing so for non-fundamental reasons because of liquidity or because they
can’t be a certain size of a company. When you have a portfolio, the positions in which are
dictated by non-fundamental reasons outside of simply investment merits, you are doing a
disservice to your investment partners. So that was very instructive.
One of the other ways in which a portfolio has evolved is that some of the biggest holdings of the
portfolio are there for different reasons. One is there because we aggressively averaged down in
this period. Another is there because it has done very, very well.
141
Tollymore is a concentrated portfolio and has a global remit. I think the natural implication of
that is that we are a fully invested mandate. That’s great for me, because frankly I’m unable and
therefore unwilling to use cash as a mechanism to time markets. And I don’t have to, because of a
small portfolio of assets and a big opportunity set. But what that means is that when it comes to
averaging down into certain businesses, we don’t have cash with which to fund that activity. And
so, we need to decide between essentially cutting our winners or cutting our losers. I think
increasingly we’ve been more comfortable with avoiding the disruption of fundamental
business progress of the companies that frankly deserve to be larger portions of our portfolio
and forcing us to choose between the losing investments that are eroding value; those that
deserve our capital and those that don’t. I think that some of the businesses have done very well.
Some of the businesses are still 50-60% off their previous prices from the start of the year and
they remain very core positions for us.
How to recognize good businesses
Tilman Versch 19:43
That’s an interesting insight. You’re investing globally and this is a bit of a challenge to boil down
to the best ideas. What are your steps to boil down to the good businesses and the anti-fragile
businesses you mentioned?
Mark Walker 20:04
The opportunity set is global. We’re not global investors in the sense that we are trying to
identify thematic changes or geographic opportunities from a top-down level and using that as a
source of idea generation. As I said before, I think that it’s really important that opportunities are
discovered in a way that some might consider haphazard and because it’s very serendipitous in
nature. What we’re not doing is finding ideas according to a systematic quantitative method
such as screening. One of the reasons I’ve described, if you are looking for these businesses that
are high quality according to our vectors of quality and you are screening over a period of time,
what you are finding is obvious excellence and that is unlikely to be undervalued. But more
broadly, the problem that we have with screens is that those who use screens cynically – coming
back to this imperative to raise capital – are doing so partly to make the investment merits of
their organisation more marketable in an institutional sense.
Institutional investors are typically – and rightly so – trying to understand the repeatability of a
process to allow them to underwrite it. It seems that by showing a kind of quantitative or
proprietary idea generation funnel you’re helping allocators of capital to take tick that box of
repeatability.
Our process is serendipitous in nature and it’s really driven by energetic and passionate
engagement with written and verbal materials and conversations with management and our
peers and engagement with reports and transcripts and filings. If that is energetic and it is a
highly motivated endeavour and if you are passionate and reasonably competent, it’s a perfectly
142
adequate way to uncover investment opportunities. Our issue has never really been that dearth
of investment ideas. What we are trying to do and what we think should be repeatable is the
judgment with which you interrogate that idea. That again comes down to what your incentives
are and how you are resourcing the investment process.
Are you resourcing the investment process with a large highly pedigreed investment team?
Because that is the way you raise assets. Or are you resourcing in a way that is most efficacious
and likely to do the greatest good for your existing cohort of investors? People in the industry
have a debated over the merits of screening. They will suggest that screens will, just simply by
fishing in that pond, improve your chances of success because that pond will outperform in
itself. My argument is that it may well have done so in the past, but there is no real guarantee that
it will continue to do so in the future. Particularly when the screen is predicated on any kind of
style factor type of investing, such as growth or GARP or value as value is typically understood.
Especially considering these changing business models of the future versus the past.
If you look at our portfolio, there are a lot of companies that screen really badly there and some
of our largest positions screen really badly. That’s because the reported financials of the
business poorly reflect the economic reality or the economic prospects of the company
because of accounting convention or because of businesses in transition or simply because of
high levels of internal reinvestment opportunities. And so, we attempt to sort of disaggregate
between owner earnings and reported free cash flow, for example.
Criteria for high-quality businesses
Tilman Versch 24:42
What makes you kill an idea and what makes you like or love a company that you want to hold it
for years?
Mark Walker 24:51
The bar is high in that we have a research process that gathers hopefully objective
information, data, strands of logic which are designed to support or refute a contention that a
certain vector of quality is high or low. So again, these are not proprietary or rocket science. A lot
of them can really be found in The Warren Buffett Way:
• Is this business simple?
• Is it appropriately financed?
• Is the stewardship adequate or good?
• Is the competitive positioning of the business strong?
• Does it have avenues for internal redeployment?
143
• Does it have very value creative rates of return?
And so, we are collecting this information and trying to score in a pretty unquantitative but very
qualitative way whether we think this is a high-quality business among these vectors. An idea
will be killed if it becomes clear that this is not a high-quality business along these vectors. But if
it passes those hurdles and we are very excited about the quality of the company and its
prospects for internal capital and the excellence of its stewardship etc., we will assess the
valuation from multiple angles. But oftentimes, through a lens of a prospective IRR to equity
owners which starts with an understanding of what the owner earnings yield of the business is
and what the reinvestment rate and the incremental returns likely to earn are and that
triangulation helps to get a rough idea of whether the IRR is really high or really low, for
example. So, if we get this business that is screening to be a very exceptionally high IRR, that may
be because of the reason I talked about before, which is that these cumulative mispriced insights
are adding up into a business quality and opportunity is substantially undervalued.
Alignment of interest
Tilman Versch 27:10
Interesting. How are you aligned with your investors?
Mark Walker 27:20
Richard Lawrence of Overlook has this phrase he refers to: ‘Outlawing greed’, which is a major
competitive advantage in this industry but a path not really taken. I think if you can set up a set
of investor incentives that allow you to outlaw greed and focus on returns above asset growth,
you’ve taken a large step of the way there. And so, the way in which we’ve thought about that is
to be very rigorous and disciplined about the investment partners that we take on.
We want a very small number of partners over time. I think everybody talks about the value of
relationships in life but rarely actually implements that understanding and sacrifices it a lot of
the time for rapid wealth accumulation and then wonders why they’re not happy, given that
sacrifice they’ve made.
The majority of my capital is invested alongside investment partners. Currently, Tollymore’s

economics for me imply that I will accumulate more capital from the compounding of my
ownership of funds under management than from any fees that Tollymore has paid me. And
then we have a set of fee structures that have been stress-tested under a range of scenarios to
ensure that under reasonable circumstances investors receive not just the majority of the return
over time but a majority of the alpha over time, a majority of the excess return over whatever
reasonable benchmark they may decide. We have different fee structures; some with fixed
hurdles and some with benchmark hurdles that investors have chosen. And so, if you look at our
cumulative results, to date investors have received around almost three-quarters of the excess
144
return over the benchmark over time. One of the things I’ve been thinking about more since the
March and April period of 2020 is really “What is the appropriate capacity of the strategy?”
One of the ways to engender alignment is that you err on the side of caution and you close the
capacity of your strategy comfortably below what you think the ultimate capacity of the strategy
is. It’s our intention to close the capacity to external investors and then to periodically allow
existing investment partners to add if and when appropriate, if we have other periods of severe
dislocation like March and April.
How to invest in companies with emerging moats
Tilman Versch 30:38
We have a first question coming from the chat. If you have more questions, please send them in.
The question is: Do you invest in businesses in the process of building a moat but not there yet?
For pre-moat companies, historical numbers that you find in a screen look junky but the future
could be very bright. What is your take on this question?
Mark Walker 31:02
I think it’s completely consistent with what I was saying about obvious and non-obvious
excellence. If I look at the contributors to Tollymore’s results and the detractors from
Tollymore’s results, the major contributors are from businesses which at first sight are written off
as low-quality businesses. That is usually due to a faulty heuristic and an inability or
unwillingness to interrogate the business quality from first principles.
So, for example, many investors have written off Trupanion as an insurance company. The
heuristic associated with insurance companies is that they are capital intensive, mature
businesses with limited opportunities to internally redeploy capital and constrained addressable
markets. I think that heuristic has led to a really outsized investment opportunity. If you think
about the components of that company’s value chain and how they cumulatively manifest in this
really phenomenal business, you can avoid that heuristic.
Likewise, one of our largest holding, Gym Group, is an old economy operationally geared
financially leveraged business selling discretionary services with very high churn. In a world
enamoured with high multiple sales, SaaS companies and Big Tech, you could succumb to being
a victim of this potential bubble in buzzwords by becoming overly enamoured with these labels
if you are not interrogating from first principles what a good business is and what a poor
business is. Those types of opportunities have contributed strongly.
We have done badly in two main areas. One is we have identified obvious excellence and we’ve
become enamoured by this beautiful financial history going back 10-30 years of extraordinary
145
super normal - consistent, super normal profit generation. The second area is in what is often
called legacy moat type companies.
Back to our goal; our goal is to maximize the intrinsic value of a group of assets. There are two
ways of doing it, one is that you buy assets at very strong discounts to current intrinsic value.
That necessitates a re-rating of the asset or a turnaround of its fundamental prospects.
Potentially, that re-rating needs to be accompanied by a catalyst and that’s a very labour-
intensive method of generating acceptable results. When you’re trying to identify a catalyst,
I think that those catalysts are readily identifiable to most people. Therefore, it’s unlikely that
there’s a big gap between price and value or a big gap between likely future results and
expectations. And so, we’ve been poor at picking those deeply discounted legacy businesses
despite our contention at the time that they were extraordinarily discounted.
By the way, we are not investing in businesses that fail to meet our objective qualitative criteria
for quality. These are all extremely high-quality businesses, but they generally have limited
opportunities to reinvest capital and internally grow their own assets. Over time, the results have
been more and more driven by these businesses that at first sight – to answer the questioner’s
question – don’t have a moat that is widening but it’s not obviously huge, because there is no
eloquent way of describing it because it’s not an economy of scale or it’s not a switching cost or
whatever moat label you wish to use.
Reasons to invest in Gym Group
Tilman Versch 35:31
You already mentioned two names: Trupanion and Gym Group. Can you tell us a bit more why
you like these companies and why you invested in both? Also, how you managed your feelings
because on one side, Trupanion is a huge winner, this year as well. Gym Group is getting
squeezed and is one of the Covid losers, if you could call it like that. How do you manage to stick
to both and how do you manage the loser and winner side?
Mark Walker 36:10
Gym is a business that has a large position because we bought a lot more and it’s my view that
our final purchases in March and April were purchased at IRRs of 60-70% and the owner
earnings yields are at 35-40%.
Even pro-Covid, Gym is a good example of this very serendipitous idea of generation process. It
was really discovered initially as a consumer experience where having been a member of a mid-
tier large multi-site gym in the UK for a decade, I suddenly started seeing that mid-tier gyms
started closing down and these low-cost gyms from Gym Group and a business called Pure Gym
started opening up sites increasingly. When I just looked very roughly at the profitability profile
of these mid-tier private equity owned gyms and the low-cost gyms, it was very striking that
146
these low-cost gyms were incredibly more profitable and their super normal profits were off the
charts compared to mid-tier gyms. But their prices were 60-70% lower.
Having joined those gyms, it became clear to me that it was actually a superior proposition, it was
a more enjoyable experience and the question then becomes “As a potential business owner, is
this sustainable?” If you think about the positioning of the incumbent in this industry, what can
the mid-tier gyms do? It’s my view, looking at their margin profile, that they could cut their prices
10-15% before they become loss-making at EBITDA level in a capital-intensive industry. And so,
their competitive response has been really muted, there have been some small efforts to launch
low-cost alternatives that have been quickly wound down.
The source of the low-cost gym’s super normal profits is that there is a margin superiority which
is a function of lower labour intensity because there are only one or two employees per gym and
there is an asset turn advantage which is facilitated by a few things.
1. One is that they don’t have any wet facilities, so they don’t have any saunas or steam
rooms or swimming pools. They don’t have any general coffee or greeting areas which
improves the density of their site usage; their site utilization is improved. Another factor
that improves their asset turns is because of the labour intensity and the technological
tools they are using to admit people into the gym, they can open 24/7.
2. So, their assets are used more efficiently. The 24/7 capability is also expanding the
addressable market to those gym users who are shift workers or taxi drivers or previously
didn’t have access to gyms that constrained opening hours.
3. They also addressed the other barriers to gym participation, which are around cost and
people not being able to afford to pay 40-50 pounds a month for a gym and people are
not signed into contracts with low-cost.
It’s my view in a nutshell that this is a business with very, very high owner earnings as a
percentage of its market cap there were substantially all being reinvested in projects that I felt
were earning 20+% cash on cash returns in capital. Again, in this investment universe, these are
unconventional sources of a company’s competitive advantage. But Gym is also a really good
example of a business that is almost a holy grail of public equity ownership, because it’s a
business that has been steadily compounding its economic value and its owner earnings over
time. But there’s a very volatile public equity associated with it.
So, since our ownership, it’s basically doubled and halved twice and now it’s 60% below again.
So, it’s below our original cost but it has contributed positively to our cumulative results
because our average purchase price is lower than our cost price and it’s been at a larger position
in the portfolio when the stock has been beaten up. It presented a great opportunity in March
and April, because in many ways the market very efficiently and aggressively recognized the
shortcomings of this business model in a world where people are instructed not to leave their
house, in a world where governments are telling gyms to close down. So, it’s a business where its
147
revenues literally go to zero. They’re not selling services with any kind of pent up demand
phenomenon and they have a very high fixed cost base. It took some back-of-the-envelope
additional work to re-underwrite our ownership of the company through a lens of liquidity first.
My view was that this was a business that was just absolutely not going out of business.
Actually, their capacity to flourish at the other side of this is really, really attractive for a number
of reasons.
• One is that their relationship with their landlords is likely to improve because of the
dearth of retail or restaurants seeking leases.
• The other is that most gyms are either local authority gyms owned by the government
who have been heavily subsidizing whenever there is a funding crisis or they are mom-
and-pop fragmented highly levered single sites with weaker relationships with real estate
landlords.
• Finally, the recessionary or cautionary conditions that Covid is and is likely to continue
to confer could well accelerate this migration to low-cost gyms that don’t tie people into
contracts.
So, in a nutshell, you’ve got this capital cycle potential outcome where capacity is exiting and
demand is increasing. If you believe that they have the liquidity and balance sheet to survive a
period of extended lockdowns – back to one of your earlier questions: “What is long-term?” –
this is a classic time arbitrage opportunity where their results are likely to look pretty horrible for
a while. But to my mind, they’re one of the best capitalized players in the industry and their
prospects for incremental returns of capital have probably gone up over time. I’ll just pause there
without diving into Trupanion. But if you have any questions, we can talk about it.
Gym Group in a post-Covid world
Tilman Versch 44:07
Yeah, there were two questions coming up. One is: Do you see Gym Group as a winner in a post-
Covid world and do you see any other winners in the fitness space as well? Another question
was: Do people’s habits change? People get used to train outside the gym and after Covid, there
is a high chance they won’t come back to gyms, so this might be a problem for Gym Group.
Mark Walker 44:47
Other winners, the other main low-cost operator is Pure Gym, which is a private business. I think
that you may well have this sort of barbell evolution where there’s low-cost and then there are
premium gyms where people really, really want that swimming pool or that tennis court or that
boxing ring.
148
I think that there are potential parallels to be drawn with low-cost airlines or supermarkets
where these propositions are initially low-cost and then they simply become the mainstream. I
think the proposition gap between mid-tier and low-cost is so substantial that the market share
opportunity is very clear. And as I said before, there is a market share opportunity driving
volume and there is also an addressable market expansion opportunity, driving volume of a third
of customers. Every year, new customers for Gym and Pure Gym have never used a gym before.
Because they’re addressing these barriers. That addresses a little bit this challenge of a structural
change where people train at home. But I’ll come back to that.
So, the other winner I think is likely to be Pure Gym. What you see is that Pure Gym and Gym are
really accounting for the vast majority of new gyms and new members in the UK. The evidence
so far since gyms have opened is that very quickly, gym memberships and visitors are returning
to almost pre-Covid levels. I’m more comfortable betting on things that are likely to remain
reasonably unchanged rather than calling a structural inflection in human behaviour. I think that
Gym’s share price reflects either an anticipation that the business will not survive and lacks the
liquidity; I think that’s absolutely very comfortably refutable. Or that people just aren’t going to
visit gyms anymore. I think the evidence so far does not support that and I think the logic also
doesn’t support it.
As a user, my experience has been pretty consistent with other people’s experience when gyms
close down. Which is that I switch to an app, which by the way was offered in partnership with
Gym Group and I used that app to try and train at home. At the start I was very enamoured with
the experience, I thought it was a very high production quality experience. But after 3-4 weeks,
despite having this huge library of exercise classes and different trainers and so forth,
I basically became a bit bored. Unless you have the capital and the space and the time and the
willingness to set up a fully functioning gym with all manner of equipment and unless you are
permanently willing to forego the social experience involved in training with other people and
using that a source of inspiration, I think it’s unlikely that people won’t return to gyms.
Potentially, the addressable market doesn’t grow as quickly as it was growing before but I think
for the reasons I’ve articulated that the market share opportunity is larger than what it was
before.
And so, I think you have a future in which Gym’s competitors are very distressed. You’ve seen
that in high-profile bankruptcies over the globe and in some cases, you’ve seen low-cost
business models in continental Europe, such as McFit, acquiring those assets. I think Gym Group
as the best capitalized business in the sector certainly has on its radar opportunities to acquire
those businesses. It also has on its radar potential premium site acquisitions that were previously
not meeting its return invested capital hurdles. But now, because of the real estate situation, it
will be able to contribute to profitable site growth.
Tilman Versch 49:29
I think there’s also a certain desire to go back out again after Covid is over, because everybody is
frustrated being at home.
149
Mark Walker 49:40
Yeah, it’s just the value of having a long-term lens. In this case, I don’t even think it needs to be
that long-term. If you just think “What is the likelihood that in another 18 months Gym is not
somewhere around its previous level of utilization and unit economics, if not in a better
position?” The business today is 25% owner earnings yield, reinvesting all of that at 20% returns
in capital.
Tilman Versch 50:16
Interesting opportunity. Shall we switch to Trupanion?
Mark Walker 50:22
Sure.
Reasons to invest in Trupanion
Tilman Versch 50:27
What do you like about the company? And I still have this question open: Is it easier to keep a
loser or to keep the winner?
Mark Walker 50:43
My temperament is such that I’ve been historically very comfortable averaging down into
situations where I felt that the opportunity to improve my investors’ investment results
presented itself. I’ve had mixed success in doing that and sometimes I’ve been guilty of being a
proverbial boiling frog and being very slow to recognize structural deterioration. Either
structural deterioration the business or my ignorance or misunderstanding of the businesses’
prospects over time.
I’m trying to be more cautious and slower to average down. I think Gym was just really an
exceptional opportunity to do that in March and April. The corollary of that is that I’ve been
slower to cut profitable investments if those companies are demonstrating a level of execution
and fundamental economic earnings progress that justifies that. I think with Trupanion we have
lowered our position size. I think when we first acquired the shares a couple of years ago, this
was really a business that I felt was valued to not grow and I felt that the growth opportunity was
extraordinary. Essentially in the last couple of years, the stock has tripled and the owner earnings
of the business have also compounded quite strongly. But the current multiple of stockholder
equity implies a sustainable growth rate of 7-8%. That re-rating of the stock will typically not
cause us to exit investments but it may be a trigger to very occasionally and sensibly power back
our ownership if the owner earnings yield decline is a reason for the lower perspective IRR.
To go back and use Trupanion as an example of some of the things we talked about today,
Trupanion basically offers medical insurance to pets; cats and dogs in North America. As I’ve said
150
before, if you’re a quality business investor and you apply this heuristic that insurance
companies are low-quality generally and they should be valued at low single-digit multiples of
the book, as a long-only investor you’ll quickly dispense this is a potential opportunity. If you are
a long/short manager, you might be very interested in this business as a short. I think Trupanion
is one of the most shorted companies we own. It’s less now, but it still has a very high short
interest. I’ve described the reasons why we don’t short securities, but I think it’s useful to
understand the short seller agenda that I described earlier to examine the insights that we
believe we have about a company. The points of contention that bears on this company have and
why we think that through our lens is very long-term long investor, they don’t really make sense
to us – and I’m cognizant of the behavioural risks associated with what I’m about to do – it’s not
my business to be overly defensive and get into arguments about whether a company is
investable or not. But when you understand the reasons for the short interest in the business and
you disagree through a research process, which is bottom-up and hopefully objective in nature,
it can lend some conviction to your view which allows you to know what to do when the quota
price changes.
One of the things that it’s been argued by people who dislike Trupanion as an investment is that
the provision of medical insurance to cats and dogs that is distributed via vets has endemic to it,
an adverse selection issue: if you’re selling to animals in the vet, you are selling to sick animals
and that affects your risk pool and that will cause a rate spiral and that will cause an affordability
issue for your customers. Trupanion is the only pet insurance operator distributing its insurance
via vet relationships. Other competitors to Trupanion rely more readily on online marketing to
acquire customers. It strikes me as logical that if you are Googling for ‘pet insurance’ there may
be something wrong with your pet if that’s your first interaction with an insurance product.
Whereas if you, like Trupanion, are focused on puppies and kittens – animals in the very early
stages of their life when they first visit the vet in the first few months of their existence – I think
that your propensity to have an adverse selection problem is much lower.
One of the other issues in this world where net retention ratios of comfortably above 100% seem
to be the holy grail of investing in sustainable enterprises, a business that has an implied churn
of 16-17% a year seems like a low-quality business with a customer retention and problem which
increases the cost of doing business. Trupanion expresses its churn in on a monthly basis which
is 98-99%. This implies this annual churn of around 17%. That can really be explained by two
phenomena.
One is that the customers who churn early, which is most of the churn, are those that are under
the misperception that pre-existing conditions are covered. Once they realize that there’s no
coverage for such pre-existing conditions, they churn off. Clearly, that is an opportunity for
Trupanion and they can do better in educating their customers by that lack of coverage. But
normalizing for that factor, you have a churn level that is then predominantly explained by the
average life of a pet. I generally think that Trupanion – given its business model – does a
reasonable job of keeping its customers. Like Gym, Trupanion is this holy grail of investing in
many ways because it is a business demonstrating very consistent strong positive fundamental
151
progress but whose quoted equity is extremely volatile. Trupanion has never reported a quarter
of revenues that was not higher than the previous quarter in its 10 years of financials. That isn’t
because they have phenomenal executors in a challenging business, it’s the result of a recurring
revenue model and a very high and underpenetrated addressable market. I think the last bone of
contention for short sellers is two-pronged.
One is that the CEO and founder of Trupanion is overly promotional and disingenuous in his
characterization of the business and how it makes money and how they think about allocating
capital internally. I’ve specifically read their arguments that contend that the CEO has compared
Trupanion to SaaS business models in order to try and get this very high SaaS multiple. I haven’t
seen that anywhere, what I have seen is that in his shareholder letters he specifically said that
this is not like a SaaS company because it has a high cost of doing business. It has a high cost of
sales and a low gross margin. But he has used comparisons to companies like Netflix because
of the recurring revenue element and the growth of its key assets is around 30% and gross
margins are both around 30%. But because he is not describing this business in traditional
insurance terms, I don’t think that this characterization is disingenuous and that leads to where
we are in valuation which you’ve mentioned before, which is that if you have a heuristic that this
is an ex-growth insurance company, you will be very excited about potentially shorting this
company. But as I’ve described before, I think that despite a quite strong re-rating of the asset, is
pricing in a level of growth of 7-8% sustainably? Because on an owner earnings basis, if
Trupanion stops reinvesting internally to acquire new pets, the ROI of this business would be in
the order of 40-50%. What it’s doing with all those owner-earnings is directing them to invest in
new pets with investment returns which are 4-5 times any reasonable opportunity cost of
investing. So those are the reasons why it remains this core investment for us.
Consolidation of industries post-Covid
Tilman Versch 01:02:16
Those are good rates for sure. There is another question from the chat: Do you expect a
consolidation of certain industries after Covid? It’s a broader question, but referring to Gym?
Mark Walker 01:02:31
We don’t really think in this thematic way of “Let’s think thematically about Covid or the US
selection or Brexit and carve out winners and losers and try and invest according to those
paradigms”. But specifically with Gym, as I’ve outlined, I think it makes sense that a number of
competitors are not going to be in great shape to compete, like the local authority gyms; around
35-40% of them still haven’t reopened. Those gyms are charging 35-40 pounds a month. Gym is
charging 18 pounds a month and reinvesting substantially in their gyms. So, I think that value
proposition is going to maintain and as I said, I think that it’s very reasonable that capacity will
exit and demand will continue to increase.
152
On the broader point, you mentioned Brexit before as well, I have no special insight into
inferring consequences of macroeconomic or geopolitical events. I think that what this year has
shown and has reiterated is that it is unknown unknowns that in aggregate move markets.
Unknown unknowns are impossible to forecast. When it comes to known unknowns, I think the
market generally does a reasonable job of pricing expectations in an aggregate way.
Therefore, I think that trying to predict the future and trying to predict the evolution of known
unknowns is also a low ROI exercise. So, it comes back to “What can you do?” I think all you can
really do is own these companies that you think have business resilience and are led by
adaptable innovative managers and owners and to lead through crises. One of the things that we
look for and talk about or recognize in some of the companies we own – Trupanion is a good
example as well – is symbiotic value chains.
I’ve been quite dismissive of most ESG-programs and the value that can really be created by the
quantification of an ESG-checklist and the employment of consultants to help formulate ESG-
checklists and really the business of long-term sustainable investing should satisfy a reasonable
ESG-agenda. One of the ways to do that is through a symbiotic value chain, through a shared
mission which goes beyond profit maximization. If customers, if employees, if owners of the
company are united in a shared objective, the ironic consequence of that shared mission is
usually profit maximization over a long period of time. And so, the internal micro-economic
examination of Trupanion’s value chain is an interesting insight in itself.
If you consider what a 70% loss ratio implies for the customer proposition, it implies that the
customer is paying a 43% mark-up on every dollar of premium. The customer seeking an ROI on
this, he is going to find that very uncompelling. But of course, customers are not seeking an ROI,
they don’t want their pets to be sick or unhealthy. There is an asymmetry in the value chain that
is as follows: Pet owners don’t have access to the aggregate information of an insurance
company that the insurance company is using to accurately underwrite policies. That is a barrier
to self-funding the medical requirements of their pets. Even if they had access to that aggregate
information, they still don’t know if their pet is on the lucky or unlucky side of that distribution.
The value of 43% mark-up is actually very high to them. And by the way, Trupanion’s competitors
have 50% loss ratios, so their 100% mark-ups are even less uncompelling as a value proposition.
There are other components to the value chain that I think combined form what is a very
resilient and anti-fragile business. One of them is that the founder, Darryl Rawlings, has a very
transparently communicated formula for sharing enterprise value growth between employees
and owners of the company. And all of the employees are owners of the company and lots of
employees are customers of the company because they basically all own pets as well.
So that’s a form of symbiosis and then finally, Trupanion has this software product called
Trupanion Express. That is installed in a number of their vet partners and that allows Trupanion
to directly pay claims within minutes or seconds of a claim being made. This is a really interesting
development, because to me, it really cements this idea of symbiosis, it really cements this idea of
win-win-win. This is a great evolution for Trupanion, because it allows them to have access to
153
data from other insurance policies provided by other insurers that allows them to more
accurately price their products and it’s great for the vet because the vet can suggest and
implement plan A treatment options and it reduces economic euthanasia, which is sometimes a
function of the customer’s inability to fund out of pocket expenses. And also, for the vet, there’s
a direct financial impact, which is that the credit card fees are saved, which can be 15-20% of
their profits.
But it’s also interesting to understand that vets’ careers are intrinsically motivated and mission-
driven. They earn a fraction of what medical practitioners like doctors and surgeons in the
human field earn. So those are cumulative series of observations about the value chain. I think in
times of uncertainty or in times of distress the value chain is united by this common mission or
objective.
Gym Group prices versus mid-tier gyms
Interesting. In the chat, there is a former customer or maybe I could also say a short seller of Gym
Group, I’m not sure. His comments are a bit critical because he’s also mentioning that they
don’t charge 18 pounds per month but they are charging 22 pounds in the bigger cities and 15
pounds in the outskirt locations. He’s also having critiques with the person who is fitting the
gyms. It is the brother of the founder and that’s a critical point for him. Do you want to react on
this or is this getting too deep for you?
All I would say is that the point about pricing is that their average price is 17 pounds in ARPU.
They are the lowest cost national gym chain. Whether or not they are charging 18 or 25 pounds,
they are not charging the 45 pounds that the mid-tier gyms are charging for their inferior
experience. If you look at net promoter scores or customer feedback for gyms, it’s generally
terrible. For Gym it is reasonable. It’s not stand-out phenomenal. The negative feedback is
usually associated with gym members leaving and being charged and being difficult to cancel
their memberships. For example, for gym members who are tied into contracts. So, it’s not a
business that is universally extremely highly regarded by all of its customers. But I think that its
relative appreciation is pretty good.
Thank you. We are coming to an end with our livestream. At the end, do you want to mention
one point we haven’t discussed yet that might be interesting for the viewers and that describes
you, Tollymore or your view on investing in a good way?
Sorry, what do you want me to say?
154
Final notes from Mark
You have the opportunity to add something we haven’t discussed. If you have a point that comes
to your mind that is interesting for the viewers, adds value and gives insights, that might be
interesting.
Tollymore at the minute is a one-man band.
I know the feeling!
Exactly! There are shortcomings, there are pros and cons of that. There is an obvious
shortcoming in potentially the intellectual robustness of a process if you aren’t sharing the work
with people who are incentivized exactly the same way as you. I think that is a clear and obvious
shortcoming. I think there are lots of shortcomings of growing teams relating to generally
complexity of organizational decision-making that are under-appreciated and not often
discussed. And so, at the minute I attempt to plug the intellectual robustness gap by being part of
an intellectually generous community and having people like you to champion my efforts and to
bring this community together. It’s a really important safeguard against that risk for me and there
have been ideas that have come into my fold as a function of these types of relationships, so
I welcome people to reach out to you and reach out to me with feedback, with questions or with
observations or challenges, bearing in mind the respective contexts and lenses through which
we are trying to judge investment merits of companies. We can play a small but valuable part in
trying to create value in a very institutionally focused active management industry.
Thank you very much for your insights and sharing your ideas. Thank you also to our audience
for being with us and raising critical questions, that’s always helpful because critique makes
progress. Finally, I want to send you a “Hi’ from Dennis Hong. He will be on here in three weeks.
He is sending you a nice “Hi’ and I’m saying goodbye to everyone who watched the stream and
wish you a good night or day, wherever you are.
Thanks, Tilman. Thanks everyone, I appreciate the time!
155
A letter to partners, March 2021: collecting red flags
Dear partners,
“The only mistakes you can learn from are the ones you survive.”
Jim Collins
Tollymore’s investment partners are investing in my energy, willingness to learn, pursuit of

humility, reasonable judgement, and capacity for focused work. They are not investing in the
finished article. My decisions have been, and will continue to be, littered with errors. But
investment partners should look for progress. They should look for signs of improvement, of a
process that is evolving, while of course remaining faithful to our central philosophical tenets. I
hope their willingness to hold me to account will be an important part of Tollymore’s ability to
generate good long-term results. I hope to aid you in that task through transparent, objective
communication. Often my errors of misanalysis are idiosyncratic, but sometimes broader lessons
can be identified and internalised to improve our investment process.
In November 2017 I purchased shares of Wonderful Sky, a financial PR services business based in
Hong Kong85 for 1.88 HKD. I exited just over three years later at 0.41 HKD. Our ownership of these
shares is comfortably Tollymore biggest mistake, and permanently destroyed 4% of our capital.
One of the barriers Tollymore faces as an ‘emerging’ investment firm is the shortness of our
existence, and the smallness of our ‘brand’. Agents acting on behalf of principals when deciding
whether to become a Tollymore investment partner face a very real reputational consequence of
failing unconventionally. I enjoy the challenge of earning the trust of aligned investment
partners, and I believe the prize for overcoming this hurdle is stronger investment partner
relationships. So, it is indeed a challenge worth tackling.
Wonderful Sky, as an incumbent with a long-established reputation, is a beneficiary of this

reluctance to fail unconventionally. Financial PR and investor relations fees for IPOs are a small
percentage of the funds raised, but are critical to the successful execution of the IPO. An IPO
which only happens once. Given this is not a repeated game, deviating from the market leader is
a risk not worth taking, either by the companies being listed or the other service providers who
dare not risk their own reputations by recommending a PR firm other than the dominant market
leader with the long and successful delivery track record. This lowers customer price sensitivity
and allows IPOs to serve as a springboard for a longer-term relationship. Accumulated expertise,
relationships, and Wonderful Sky’s integrated offering all contribute to switching costs which
have preserved the company’s excellent financial track record.
85 The detailed thesis was explained in our December 2019 letter to partners.
156
But just as I believe that multiple small insights can compound to form a defensible business, so
too can multiple small fractures compound to unravel a company’s investment merits.
Red flag one: the company’s purchase of its office building and swelling fixed income security
investments on the balance sheet should have triggered a re-evaluation of capital allocation,
stewardship, and alignment of interests. I relied lazily on high insider ownership and dividends
to owners multiple times the CEO’s salary to mitigate this. As someone who talks a lot about the
virtues of first principles thinking, this sloppiness is hard to excuse.
Red flag two: the company issued a profit warning due to a soft IPO market in 2017 and 2018. It
should have become clearer that this company’s prospects are affected by factors outside of its
control.
Red flag three: management cut the dividend to zero, justified by the decline in market
conditions. However, the dividend was not reinstated as these conditions began to recover at the
end of 2018. I began to engage with management much more intensely to recommend, in
concert with other institutional owners, the payment of a large special dividend to address a
capital structure that was turning large returns on invested capital into paltry returns to equity
owners. While key executives sounded receptive to my pleas, ultimately my words fell on deaf
ears. Repeatedly.
Red flag four: as the stock price continued to fall precipitously, I became increasingly concerned
about management’s economic incentive to take the company private. I engaged with Hong
Kong legal counsel to better understand minority protections. It is bizarre that this is the course
of action I believed to be in the interests of Tollymore’s investment partners. A dispassionate exit,
unburdened by loss aversion, would have been a far more valuable decision.
Red flag five (I know…I am so sorry): the final straw came when my interviews with prior
executives revealed a potentially decaying company culture.
This is the opposite of fundamental business progress, and I cringe when I read the updates I
made to the investment memo. As the potential value of the company dwindled, upside
remained largely unchanged: a classic value trap. I was far too slow to incorporate new
information into my appraisal of the company’s equity prospects. I collected red flag after red
flag, painfully contorting the narrative to justify our continued, and in some instances, increased,
ownership. It should have been quite clear to a reasonable observer that business quality had
been misanalysed and was deteriorating. Despite espousing the value of intellectual honesty, our
ownership of Wonderful Sky was awfully dishonest. I was seduced by extreme cheapness: the
company’s bizarre capital structure meant that even modest equity declines had extreme
implications for business value, resulting in negative enterprise value in the order of hundreds of
millions of Hong Kong dollars. In the giddiness of finding this valuation ‘anomaly’ I lost sight of
the core principles of Tollymore’s investment philosophy.
157
Forward execution
I have been somewhat swifter in recent times to exit investments failing to demonstrate positive
fundamental business progress.
I acquired shares in Aspen Group (ASPU) in November 2019 for $7.38 86. I believed ASPU, an
online university for degree seeking nurses, was a mission-oriented cost leader in a defensive
industry. The provision of federal student loans had disturbed a properly functioning for-profit
education market, leading to high prices, poor outcomes, and substantial student debt. This
drove a decline in the for-profit education sector of c. 40% over the last decade. But ASPU was
bucking the trend, rapidly taking share due to a student proposition that was cheaper and better,
absent reliance on federal aid. ASPU benefitted from counter-positioning by offering courses c.
50% cheaper than competitors. Incumbents with flat growth and high profitability were unlikely
to match ASPU’s prices. We have observed similar incumbent behaviour with the rise of low-cost
gyms over the last decade.
Fuelled by a compelling student proposition, large addressable market, and extraordinary unit
economics, I believed that ASPU’s growth might be very valuable indeed. My thesis was that
profitability was inflecting as the business was scaling, and the quoted marketing efficiency
ratios were leading to very high revenue-to-profit flow through. Rising operating cash flow
supported management’s contention that the business was passing through a break-even point.
Increasing profit margins though 2019 and most of 2020 lent credence to management’s
presentation of mature cohort profitability. This played out for the first year of our ownership,
with the stock rising to a peak of c. $13.
But then came the red flags.
Firstly, ASPU reported two quarters of widening losses, explained by campus opening delays and
a COVID second wave delaying nurses’ return to education.
Secondly, the third CFO in as many years left the company, accompanied by assurances this was
for personal reasons.
Filing an S-3 was the final straw. If the cash flow reversal was temporary, why raise capital? We
exited in March 2021 at c. $6.8. There are absolutely risks to selling, but ultimately execution
disappointed and the business stopped demonstrating fundamental progress, undermining the
presentation of unit economics and shrinking my conviction.
86 Full thesis explained in our March 2020 letter to partners.

158
Lessons
There are some broad lessons I hope to learn from these mistakes and those discussed in prior
letters: be less willing to give the benefit of the doubt to management when aggregate
economics are not at least directionally supporting unit economics; do not pay to average into
businesses with temporary problems, funded by businesses earning the right to be larger
portions of the portfolio; be careful when fragmented supply consolidates, and the impact that
might have on the utility of aggregation87; do not get too close to management, a relationship
which may muddy the objective assessment of business prospects; and do not be hamstrung by
the label of long-term investor - cutting losses is consistent with maximizing long-term value of a
group of companies88.
At least some progress has been made; Wonderful Sky is my only significant dollar error. But the
opportunity cost of some of these mistakes has been large, and therefore much value can be
gained from not repeating them.
One final note: I am conscious that I owe you a discussion on shrinking value chains. The
conclusions of this work have still not fully been expressed in the Tollymore portfolio, but I hope
to discuss this enabler of value chain symbiosis in a future letter to partners.
Until then, thank you for your partnership.
Mark
87 See our previous discussion on TripAdvisor. This is something to watch with our current investment in Farfetch.
88 I am also open to the idea that I am a very poor micro-cap investor; I will keep an eye on this!
159
A letter to partners, June 2021: Tollymore - the next chapter
Dear partners,
On 1 July 2021 Tollymore launched a new fund structure89 to run alongside partners’ segregated
accounts. The investment results of the investment portfolio in the five years since inception are
presented below.
$1 invested in Tollymore at inception was worth $3.68 on 30 June 2021, after expenses but before
fees paid to Tollymore. Over the same period $1 invested in the MSCI All Country World Index
would have generated a return of 99 cents. Around three quarters of this benchmark
outperformance would have accrued to partners invested at inception. Tollymore has generated
returns of 26% per annum net of all fees and expenses90, vs. 14% for the MSCI ACWI:
USD Tollymore (gross) Tollymore (net) MSCI ACWI

2016 17% 15% 8%
2017 25% 22% 24%
2018 (2%) (3%) (10%)
2019 21% 20% 26%
2020 92% 81% 16%
2021 YTD 11% 10% 13%
Cumulative 268% 224% 99%

Annualised 29% 26% 14%
Inception 12 May 2016, unaudited.
Tollymore has been a joy. Our partners are special; they are a true source of strength, and a major
driver of the successful execution of Tollymore’s investment programme. My fiduciary
responsibility is to current, rather than prospective, partners. This means (1) reinvesting
Tollymore’s increasing scale into lowering the costs to participate. I don’t want the investment
vehicle to prevent the right investment partners from joining the partnership. And (2) new
investment partners must enhance, and not dilute, our collective prospects for success.
Tollymore has been, and must continue to be, compatible with a flourishing, examined, life. This
will never be sacrificed.
89 Cayman domiciled segregated portfolio company with UK reporting status. Subscriptions and redemptions are
monthly with 30 days’ notice.
90 Inception 12 May 2016. Source: MSCI, Interactive Brokers, managed account performance in USD, unaudited, net
of all expenses, 1% management fee and 10% incentive fee in excess of a 6% compounding hurdle, as of 30 June
2021.
160
I encourage potential future partners to read the letters on the following pages (armed with
coffee!) and to review the other content on our Investor Portal and Insights page. This content
will allow you to interrogate Tollymore’s investment approach and the consistency of my words
and actions, and to ultimately decide if we are right for one another. You can rely on me not to
pitch the merits of being part of our partnership! Judging by the actions of our current partners,
this has served us well.
Tollymore’s growth has opened the opportunity to launch a commingled fund, allowing our
investment partners to participate in a lower cost, more tax efficient, manner should they
choose. Segregated accounts will continue to run alongside the fund, and there will continue to
be a single portfolio and investment strategy.
The salient characteristics of the fund are:
Alignment: General and limited partner alignment is facilitated by: insider ownership;
performance fees and compounding hurdles which make strong performance expensive and
weak performance cheap; appropriate capacity constraints; soft lock ups to support Tollymore’s
potential for antifragility; and a commitment to reinvest operating leverage into resourcing
and/or lower fees.
Capacity constraints: Tollymore’s capacity is appropriate for the strategy. An appropriate

capacity will help us to maximise time and capital weighted returns by: (1) mitigating the risk of
being a bystander. That is, the risk that Tollymore’s portfolio is a function of liquidity or other
non-fundamental factors. This is a barrier to great investment results. And (2) allowing us to
reopen to existing partners in periods of severe dislocation, nudging our investors to behave
countercyclically.
Reasonable incentive principles: Sensible incentives allow for the enrichment of GPs together
with, and not at the expense of, LPs. The manager’s compensation should be driven by the
compounding of his own capital in lieu of fees on additional capital. Management fees should
create a stress-free environment for the investment manager to make the decisions most likely to
maximise long-term rates of capital compounding. Sharing of investment returns over
appropriate hurdles lowers pressure to grow beyond the strategy’s investment capacity. Finally,
transparent communication with an aligned investor base allows the manager to average down
and acknowledge mistakes. Doing both better than the competition is an important ingredient of
long-term outperformance.
The new fund will charge investors a 1% management fee and a 10% incentive fee above a 6%
compounding hurdle. I think this fee structure is consistent with Tollymore’s objective of non-
zero-sumness. I want to ensure that investors receive most of any outperformance generated,
not just most of the absolute return.
161
The business risk associated with a compounding hurdle – that is Tollymore fails to generate
incentive fees through a prolonged market correction or period of portfolio underperformance
– is mitigated by (1) a management fee – Tollymore should be adequately resourced to deliver
the best investment results possible, and this resourcing should not be a function of
affordability, and (2) early redemption charges – for the first three years, redemptions over 10%
of capital will be subject to a charge equal to two times the non-accrued management fee.
It has been my privilege to partner with you; I’m looking forward to the next five years and
beyond!
With my best wishes.
Mark
162
A letter to partners, September 2021: shrinking supply chains
Dear partners,
Value chain symbiosis vs. other business quality frameworks
Adversarial vs. cooperative value chains
We seek “non-zero-sumness”, or symbiotic value chains, in the small collection of companies we

own. A symbiotic value chain is one in which multiple stakeholders participate in a company’s
value creation; success is shared with customers, employees and owners through thoughtful and
transparent incentives. When the mission goes beyond profit maximisation, profit maximisation
is often the result. When all stakeholders are inspired by a mission, short term sacrifices in the
interests of long-term outcomes become easier. This makes for a more defensible, less replicable
business as stakeholders adhere to the mission through good times and bad. Value chain
symbiosis (VCS) cannot be achieved while one component’s profits are another component’s
losses.
We are value investors. But rather than seeking value in weak companies, we are trying to find it
in strong companies. We define business strength here in terms of durable corporate vitality;
the possibility for a company and its constituent parts to flourish for a long time.
Through this lens, conventional business quality frameworks such as Porter’s Five Forces seem
somewhat unstable. Concepts such as supplier and buyer bargaining power imply a zero-sum
value chain, whereas VCS has at its heart the idea of win-win-win. Rather than investing in the
customer and supplier proposition, Porter’s forces connote an adversarial relationship between
the company and its vendors and consumers.
And the first of the five, competitive rivalry91, has left most corporate leaders to obsess over
competitors at the expense of delighting customers. The popular adoption of military texts by
business executives is the result of, or perhaps driven by, an intense focus on existing
competitors92.
Game theory, or the economics of strategic behaviour, has little to say about value creation for
value chain constituents. Here too, there is an overwhelming focus on competitive behaviour
and incentives. The internet has lowered the barriers to entry across many industries. Incumbent
businesses compete with many challengers, often across the globe, lowering the practical use of
academic theories of competitive behaviour that assume a handful of rivals can be observed.
91 And, more subtly, the threat of new entrants and substitutes.

92 https://www.business2community.com/strategy/sun-tzu-and-the-art-of-business-strategy-0587525
163
Time arbitrage may be our only opportunity for great investment results. So, frameworks for
appraising corporate vitality should be more deep-rooted and durable. Thinking about these
types of business characteristics may help us to identify and recognise patterns of successful
behaviour93. But I do not wish to use them in a prescriptive way. To do so may restrict our
investment universe, in so doing lowering our hurdle for portfolio inclusion. The goal here is not
to find the best mental model, or the mental model that will unlock superior investment results.
Models will come and go. They will be relevant and then irrelevant. One hack that might help us
to find mispriced insights is to think from first principles. For example, I think that it is hard to
conclude, from first principles, that businesses which suppress competition with switching costs
or patents, have a moat over a very long period. This is because these situations lead to
disgruntled value chains. The goal of customer happiness may be consistent with VCS; the goal
of customer captivity is unlikely to be. Disgruntled value chains are a symptom of egregious and
ephemeral profit extraction and invite business model disruption 94.
Dominance acquired via customer captivity may last a long time. In the case of Gillette, it took
about a century before Dollar Shave Club offered a pay-as-you-go, cancel anytime razer supply
without the razor-and-blade pricing model95 that has been the subject of much acclaim. In the
thirty or so years prior to Dollar Shave Club, Gillette had around two thousand patents granted.
Patents that were used to protect evermore incremental tweaks to its product. That is a lot of
effort and capital directed at stifling competition rather than delighting customers. Even when
business dominance acquired via these means lasts a long time, it can unravel quickly when a
new proposition reveals the customer preferences that were not being met. I suspect that the
speed of Gillette’s market share reversal was not just down to DSC’s viral commercials, but also
the pent-up frustration from a customer base that felt overcharged and underserved. Goodwill
and reputation are destroyed faster than they are built and going against customers’ needs
should not be considered a reliably durable business strategy.
VCS and disruptive innovation
Today disruptors are not typically seeking to replace incumbents entirely. Rather, they break the
links in the customer journey, in doing so better aligning monetisation with value creation and
minimising externalities. For example, Amazon broke the link between product browsing and
purchasing, Zipcar broke the link between car purchasing and driving, Uber broke the link
between hailing a taxi and riding in one, Blue Apron broke the link between ingredient shopping
and meal preparation, Airbnb broke the link between staying in residential property and owning
it.
Zipcar, Uber and Airbnb are specific examples of business model innovation which separated
asset use from ownership. This is hardly a novel idea; it’s called renting. Rental models lend
93 Incidentally I think the ability to identify and apply generalisable patterns is a strong case for generalist investing.
94 Large and fast changes in market shares.
95 Other examples: Kodak cameras, HP printers, Nespresso coffee machines.
164
themselves to assets which are expensive and durable, and where usage is infrequent. Amerco
did this with trucks, and gyms have been doing it for thousands of years.
Of course, business and investment frameworks need not compete with one another. In business
ownership there are multiple ways to create value. Our ownership of low-cost gyms is an
expression of mispriced business model innovation first proposed by Clayton Christensen. That
is, incumbents’ neglect of less profitable customers creates an opportunity for challengers to
enter at the lower end of the market. Mid-tier gyms, focused on preserving existing profitability,
initially disregarded the growing traction of low-cost gyms. Despite this low-cost challenger
playbook in airlines and supermarkets resulting in large mid-tier market share loss, mid-tier gyms
ignored low-cost traction for the best part of a decade. The low-cost model increasingly appeals
to mainstream members due to increasing scale benefits being shared with the value chain. I
expect gym members to continue to migrate to these disruptive low-cost gyms; the incumbents
have nowhere to go, given the large price and profit differences vs. the low-cost alternatives.
Adopting the same business model would harm the incumbent’s existing business by requiring a
capital-intensive reengineering of their entire estate96.
Christensen’s frameworks are 25 years old. And there is plenty of evidence of this type of
business model disruption across industries. And yet it reliably continues to happen. Around the
same time as the Innovator’s Dilemma was published, Netflix launched its DVD-by-mail business.
One of the cornerstones of Netflix’s proposition was a promise not to charge late fees. The
incumbent, Blockbuster generated 16% of its revenue from this highly lucrative activity. With
slender operating margins, like mid-tier gyms, Blockbuster had nowhere to go. Hamilton Helmer
has referred to this as ‘counter-positioning’. Counter-positioning occurs when the NPV of the
challenger strategy is positive for the challenger but negative for the incumbent. Counter-
positioning, rather than managerial incompetence or technological inferiority 97, is responsible
for Blockbuster taking seven years to launch a DVD-by-mail business. Adopting the challenger
model becomes rational when the incumbent’s core business has shrunk enough to make this a
value accretive activity.
Often the incumbent’s response to being disrupted is to attempt to reclaim customer captivity
and therefore move even further away from VCS. A more sustainable approach is to identify the
incumbent’s value creating activities and monetise those – the slotting fees of supermarkets and
consumer electronics chains are one example of this, in which manufacturers are charged for
product placement.
96Incumbents have made perfunctory and unsuccessful efforts to do so.

97Technological disruption occurs when the incumbent lacks the expertise necessary to be successful in the
challenger’s business model. E.g., Kodak succumbed to digital photography challengers because it lacked
semiconductor memory expertise.
165
Next plc. (NXT.LN): a model for incumbent response
A focus on VCS will allow incumbents to successfully disrupt and defend against disruption.
Special managerial foresight, including an appreciation of VCS, a willingness and ability to self-
disrupt, and good luck, are required to successfully defend incumbent business models. Next
plc., skilfully stewarded by Simon Wolfson, ticks all these boxes. Since 2014 Next has featured
third party brands alongside its own brands through its online platform LABEL. This has been a
profitable strategy for Next. So why have ecommerce challengers not counter-positioned this
brick-and-mortar incumbent? It has been fortunate to own assets which are crucial to
consumers’ omnichannel shopping preferences. These include its stores which serve as
convenient pick up and return points, and its logistics infrastructure which is a legacy from the
catalogue business it launched 35 years ago. Lord Wolfson recognised the opportunity to
repurpose these assets in a more digital world. Next’s online investments over the last 20 years
are evidence of customer-centricity. But its more recent efforts to help competing brands
increase their addressable markets demonstrate VCS by treating both customers and suppliers as
clients, as partners. Today online is 70% of Next’s business98. A counter-positioned incumbent
may have arrived at this point due to core business erosion. But Next got here by profitably
growing online sales. By accepting the near-term cannibalisation of the traditional business in
return for becoming the operating system of multichannel ecommerce.
While the retail drag continues to diminish, online opportunities have materially increased, and
online customer growth is accelerating. This traction is galvanising management’s ambitions for
LABEL, in a manner consistent with VCS: “We recognise LABEL can only be successful if it treats
brands as valued clients rather than “suppliers”. With greater numbers of engaged customers,
Next can more easily uphold its promise of being the most profitable route to market 99.
Under ‘Total Platform’ Next makes its warehouses, call centres, distribution networks, customer
relationships, marketing engine and lending business available to third party brands, including
next day delivery services and store collections and returns. Next also builds and operates
brand.com, all in return for a 39% commission on brand partner sales. By creating substantial
value for brand partners, customer delight is achieved through better value products.
Next’s recent results demonstrated further progress in becoming the operating system of multi-
brand omnichannel retail. Next is becoming increasingly embedded in the value chains of its
brand partners; it is becoming the value chain within a value chain. Next is taking on more
functions of brand partners’ business models, allowing brands to de-risk their businesses and
focus on what they enjoy: design. Increasingly Next is manufacturing partner brands’ products
under licence100, further integrating itself into retail value chains by connecting merchants with
98 Next is a business in transition and may be considered at somewhat of a tipping point; today as many people are
employed in technology as buying/merchandise.
99 In August NXT lowered the commission rate on clothing for the third time in three years. Scale economies are being
shared with the entire value chain, not just customers.

100 Next now manufactures stock under licence with 12 partners, including Childsplay, Laura Ashley, Victoria’s Secret,
Aubin, Reiss and GAP

166
manufacturers. This is one less thing for the brand to worry about, and a relatively
straightforward activity for Next to leverage its sourcing experience to create and distribute
brands’ products.
For the first time, Next collaborated with LABEL brands to co-fund digital advertising campaigns
for third party brands on Next’s website. The company intends to aggressively extend this
programme to increase awareness of LABEL. This is another example of cooperative rather than
adversarial value chain behaviour. This cooperative behaviour is valuable for owners too.
Returns on digital marketing efforts are increasing, and this year the company will spend
nothing on catalogue production for the first time in 30 years. For every £1 Next has spent this
year, it has already generated £1 of profit. And this is before the positive impact of greater
partnership advertising.
So, LABEL unit economics are attractive. But LABEL’s success also improves the working capital
efficiency of the business. This is because with greater SKU variety, close substitutes can be
found for out-of-stock items, requiring less inventory cover.
LABEL added 320 new brands over the last two years with 60% of sales now generated via
commissions vs. wholesale. Next’s strong business progress contrasts with a UK High Street in
disarray. When stores reopened LABEL sales did not suffer, suggesting limited high street
competition for the LABEL business.
Business model innovation case studies: shrinking supply chains
Shrinking supply chains, or disintermediation, are consistent with both this idea of customer-
driven disruption, and value chain symbiosis.
Consider meal kit delivery business models such as HelloFresh (HFG.DE). The supply chain pulls
demand from the consumer. Orders are only fulfilled when the customer places their order, vs. a
traditional supply chain in which products are pushed from the producer. In the shortened
supply chain products are distributed direct to home, vs. the traditional model in which there are
multiple components such as wholesalers, warehousing and stores with whom value add must
be shared. Shortened supply chains are faster – typically a few days from order to delivery.
Because there are fewer steps, there is typically less waste and more margin, so they have the
capacity to be better for the environment101 and owners as well as suppliers and consumers.
Management also seems to understand VCS and intends to share HFG’s growing scale and better
margins into product and service improvements.
101Meal kit delivery supply chain CO2 emissions are 85% lower than supermarkets as there is no need to heat, light
and cool thousands of stores. One third of the world’s food production never reaches us and food production is
responsible for one quarter of greenhouse gas emissions. HFG’s food waste per € of revenue is 70% lower than
supermarkets.
167
HFG is a mission-oriented business102, which makes VCS more possible. HFG is disrupting its
existing market with business model rather than technological innovation. A potentially key
insight is there is only one brand (HelloFresh) which minimises economic leakage associated
with multiple brands and marketing dollars. A less complex value chain consists of the producer
and HFG only, in one vertically integrated model. There is much to like vs. the standard
ecommerce model which simply moves existing supply chains online: order frequency is high
and predictable, and working capital is negative, leading to attractive cash generation.
To my mind the customer proposition is extremely compelling: this is one of the very few
business models that is both highly convenient and highly affordable vs. the incumbent offering.
This is certainly the case when the incumbent offering is supermarket grocery shopping. But it is
arguably also true against grocery delivery, restaurants and food delivery. These alternatives take
more time and/or are more expensive than meal kit delivery.
Online food is decades behind other established ecommerce categories and COVID may have
allowed consumers enough time to better understand the options available to buy food online,
and to form lasting shopping habits. And HFG has audacious ambitions to disrupt the grocery
supply chain, one of the largest industries in the world.
I did not add HFG to Tollymore’s collection of businesses. As a customer of Gousto, an HFG rival,
I think this is the future of food consumption. Value is undeniably being created via business
model innovation.
But while this is a membership model, the offering is somewhat commoditised, and switching is
prevented largely by inertia. The recurring revenue nature of the business restricts customer’s
ability to shop around at the point of sale, as they are free to do with food delivery and ride
hailing for example. Customer churn has improved but is still very high. This may be down to
competitor promotions, which may elevate the cost of customer acquisition and cap achievable
long-term profitability (as may the replicability of the assets).
This is a better value, more convenient and more sustainable alternative to supermarket grocery
shopping. But in the absence of barriers to profitable participation, my sense is that most of this
value will accrue to the consumer. In the absence of a moat, it ultimately fails the VCS test. The
moat may be down to execution, which matters a lot. But it is harder to underwrite and there was
no shortage of VC dollars backing Blue Apron not so long ago.
One business that is disrupting existing markets via business model innovation with assets that
are very hard to replicate, and in a manner consistent with VCS, is Redbubble (RBL.AX). I have
been buying shares in RBL over the last year at an average price of AUD 4.40.
102 To change the way people eat forever; to be the world’s leading fully integrated food solutions group.
168
RBL is a three-sided marketplace connecting 700k independent artists, almost 10mn customers
and a third-party network of 44 global fulfilment partners. The business was founded in 2006 as a
social network for artists, but today is a platform centred around artists, onto which they can
upload their artwork which can then be printed onto a wide variety of merchandise such as t-
shirts, phone covers and scatter cushions. RBL caters to consumers and artists who value
individuality and find mass-production unappealing (‘anti-Amazon’). Products are made on
demand (POD) by third party fulfillers, who print designs onto standard ‘blanks’ (the white label
product e.g., hat, skirt, coaster) and ship the product directly to the consumer.
RBL is a leader in a new business model: on-demand retail. RBL is neither 1P nor 3P; POD is a
different business model that shortens the supply chain by creating and distributing inventory
only when the order is placed. This business model innovation seems to be occurring due to
changes in customer demand along the vectors of product uniqueness, variety, quality and
choice, delivery experience, and sustainability.
As products are manufactured and shipped on demand, RBL benefits from negative working
capital and cash inflows due to growing revenues. Customers pay upfront, while fulfillers and
artists are paid after shipment. This is a cost of growth advantage that may be missed by investors
preoccupied with profit margins. Zara, considered one of the world’s most efficient and capable
fashion retailers with a best in class supply chain, takes 60-90 days to complete a working capital
cycle; RBL’s is less than a week.
Often marketplaces are difficult environments to sustain VCS because sellers can view them as
necessary evils. Changing incentives are responsible for this. For example, to build liquidity and
scale a marketplace may offer very low take rates and advertising fees. One part of the value
chain, in this case sellers, may become more reliant on the platform to access customers, in the
process diluting its direct relationship with them. As power increasingly shifts to the platform
the temptation to exploit this grows. But long-term platform operators understand that whilst
moves to exploit suppliers, customers or employees may result in ephemeral economic
extraction, they open the whole ecosystem to business model innovation and disruption seeking
to solve a problem for the disgruntled part of the value chain.
Fragmented, global supply is one reason that sellers may not wish to secure direct customer
relationships. A local Chinese restaurant or plumber may find initial value in partnering with Just
Eat or Checkatrade.com, but they become incentivised to disintermediate the platform and
nurture local relationships with their customers directly.
Another reason is that the vendor proposition is highly compelling even as the marketplace
grows and the balance of power shifts. Marketplace owners that seek to equitably share the fruits
of operating leverage with value chain components are likely to be more defensible. Again,
switching costs cannot support truly sustainable excess profit creation.
169
RBL benefits from a fragmented, global artist community and a management team sympathetic
to the idea that sharing a larger pie in a way that delights all value chain constituents makes for a
very hard to replicate business. There is an obvious short-term incentive to monetise sellers by
charging them advertising fees for better discovery. Deeper pockets, rather than superior
products, drive success. This is not what customers, nor (talented and hardworking) sellers, want
and is therefore potentially an impediment to VCS. There will be a point when this urge will
become hard to resist. But for now, the organic nature of content discovery is an important
competitive differentiator and facilitator of win-win-win.
A key difference is that RBL is centred on the artist before the customer. Some feedback from a
seller which highlights some of the implications of this for value chain symbiosis:
“Redbubble (and Zazzle, one of their competitors) for example don't manipulate the visibility of
the seller’s listings like Amazon and Etsy do. Amazon and Etsy force their sellers to pay expensive
ads to get visibility for their products. The harder the competition gets on these two platforms,
the higher the marketing costs for the sellers rise. Additionally, Etsy significantly increased the
fees over the years, and they try to force their sellers to ship their products for free to the
customers. If you don't ship for free the visibility of your products gets reduced. Etsy and
Amazon earn tons of money selling prominent product listing places to sellers. So, you have high
marketing costs on their platforms, or your products nearly don't get seen by customers. Much
pressure. On Redbubble and Zazzle all listings show up organically. So, the most liked or the best
products automatically show up on the top and not the products for which the highest
marketing fees were paid (and that are often not what the customers are looking for). It's easy for
a seller to get in friendly contact with Redbubble when there is a problem - quite different from
the anonymous and uncaring companies Amazon and Etsy, who mostly only send you
prefabricated, delayed standard texts. On Amazon and Etsy, you often get your designs stolen
and copied and both companies don't care about that, or it takes a long time until they react. In
contrast Redbubble cares. Quite a few sellers now who left Amazon and Etsy and only sell on
Redbubble. I wish this sympathetic company much success”.
RBL receives a 30% take rate of the GMV in return for connecting the artist, consumer and
fulfiller. For every $100 of products sold, the artist makes $15, the fulfiller $44, the payment
platform $3 and the tax man $8, leaving $30 for RBL. This is a high take rate by the standards of
modern marketplaces103. The high take rate is justified by high GMV growth and greater scale
enabling lower per unit fulfilment costs.
Artists are not making huge sums from their participation on RBL. Yet their numbers are growing
strongly. In addition to a low-risk way to earn extra income, artists are seeking validation of their
talents and efforts. This search for recognition is a non-monetary stakeholder incentive that may
make non-zero-sumness easier to achieve and sustain. Similarly, the creative nature of this
industry and of the people involved – from programmers to business development personnel –
103 Like our other marketplace holding companies Next (39%) and Farfetch (30%).
170
may lend itself to mission-orientation and VCS. High revenue per employee of almost AUD 2mn
is perhaps testament to the idea that financial incentives are insufficient to bring out the best in
people. Management understands the need to keep various stakeholders happy, demonstrated
by the priorities for measuring success:
“The metrics by which we will measure our success are firstly Gross Transaction Value, Artist
Revenue and Marketplace Revenue”
Management’s forward projections imply that artists will receive the same percentage of GMV as
they do today (c.15%) despite the larger scale of the business104. They will be receiving more
value for the same price. Management seems to understand the value of delighting multiple
parts of the value chain. Even as they dedicate more focus to consumers, they are doing so by
making the artist’s life easier:
“Focusing on the consumer is not about not focusing on the artist. It’s actually about focusing on
the core thing that the artists want from the platform, which is to sell products featuring their
designs. Yes, we’re trying to get some balance into the way that we focus, to really focus on all
aspects of that consumer journey…The beauty of marketplaces is that you must be an ‘and’
business; you must be both. It can’t be artists or consumers; it must be both. Great marketplaces
know how to balance the needs of both sides of the marketplace, think about both
simultaneously and hold two competing ideas. On one side, I must think about it as an artist
services platform and, on the other side, I must think about consumers and all the different
places they could go to shop that have nothing to do with independent artists; it’s just, where do
they buy a cool hat. That’s the challenge and that’s what makes these marketplaces fun. It’s what
makes them unique when they work and it’s also what makes them hard to replicate. They’re not
easy businesses.”
RBL allows artists to focus on being creative; creating artwork without having to consider how to
market and distribute it105. Artists have the autonomy to set their margin and can choose which
of RBL’s 120 products onto which their designs are printed. Artists benefit from RBL’s collective
marketing dollars and demand generation engine. Finally, RBL arranges the manufacture and
shipping of these products.
RBL management refers to a desire to be artists’ first choice (onboarding is easy, low cost and low
risk), and best choice (return on effort is better with RBL than any other monetisation method,
such as Shopify or connecting with fulfillers directly). While RBL has done a good job of
establishing itself as the default partner for artists, more work is required to establish itself as the
default destination for consumers…
104 Artist margin has not scaled over time – it was c.15% in 2015, it is the same today and is expected to be the same
in the future. Trupanion is another business Tollymore owns that has enshrined its customer value promise in a
formula.
105 Solving a similar problem to Naked Wines and Next plc: allowing the vendor to focus on what they love.
171
RBL is a place for consumers to find original and meaningful designs on many different products.
RBL’s key customer desire is self-expression. The shortened supply chain means that customers
can access individual, personal products at reasonable prices.
The shortened supply chain also satisfies both customers’ desire to support small businesses, and
the increasing sustainability agenda of customers by reducing inventory waste and the energy
consumption of warehouses and distribution mileage, representing a refreshing antidote to the
environmental and social externalities associated with fast fashion.
Special business characteristics – SKU convexity – enable barriers to competition. Reinvestment

into greater supply has an especially outsized impact on increasing barriers to entry; new designs
are multiplied by the 100+ product range to drive exponential SKU growth. Therefore, the
number of unique items sold increases at a faster pace than the number of unique designs on the
platform. With 60mn designs106 and >100 products, RBL showcases billions of SKUs, without
holding inventory. RBL can immediately enter a new product category due to the existence of
millions of designs. RBL immediately becomes the largest provider on the internet with those
images. Scale begets scale. And with no capital cost or inventory risk, RBL can disrupt physical
retail with economics physical, or even online, retail cannot match 107.
This SKU convexity makes accelerating improvements in content discovery and customer
experience possible. This is probably very hard to replicate. And it is hard to appreciate the value
of nonlinear outcomes which are possible here to a an even greater degree vs. typical network
businesses.
A three-sided marketplace gives rise to two potential flywheels: (1) an artist-customer flywheel:
given there is no cost to artists to having their designs available on RBL’s products, the number
of SKUs is only likely to grow. RBL spends virtually nothing on artist acquisition. Meanwhile
customer acquisition costs are suppressed by long tail search: generic search terms are more
competitive, and therefore more expensive than specific, long tail, search terms. And as RBL’s
range of unique items grows, so too does the probability of conversion, improving the ROI on
paid search. This could be one factor supporting the sustainability of special take rates and unit
economics. Customers are increasingly able to find what they are looking for as the product
range increases. And (2) a customer-fulfiller flywheel: today there are 44 fulfilment partners
(printers) in the US, UK, Canada, Netherlands, Germany and Australia. Distribution is
decentralised108. That is, fulfilment centres closest to the customer are used to fulfil orders. As
the size and density of the fulfilment network increases, shipping times and costs improve,
increasing customer demand and pushing more volume through the fulfilment network. These
economies of scale are then used to suppress shipping costs and expedite deliveries to
customers. This is the essence of the customer-fulfiller flywheel. And it seems to be a flywheel
only enjoyed by RBL. While Etsy sellers can connect to third-party POD fulfillers via the Etsy
106 Much of this content is durable, perhaps evergreen.

107 A form of counter positioning.
108 Like Farfetch.
172
platform, they are striking fulfilment agreements as an individual creator, and not benefitting
from the collective bargaining power of the platform. Redbubble is the largest fulfiller customer
in the world109. With fulfilment costs representing almost half of GMV, this is an important
advantage. Etsy’s core customer proposition of unique and individual designs is a barrier to
production scale in a way that RBL’s product template standardisation is not.
This idea of collective bargaining power can be extended to RBL’s ability to use its large
fulfilment network to secure the best quality and prices from manufacturers of ‘blanks’ 110, and
perhaps could be considered a third flywheel not discussed as much by management. The
balance of power has shifted with RBL’s increasing scale: in the early days, fulfillers had the
stronger relationships with the manufacturers. This is an offline infrastructure advantage vs.
online pure play businesses.
Being the largest volume POD player, RBL not only will enjoy better rates than smaller peers, but
it will also be able to fill capacity more reliably at peak times. This supply chain robustness is a
customer service advantage. This lends antifragility to RBL’s model in times of stress. Like low-
cost gyms’ covenant strength and landlord relationships, Farfetch’s decentralised distribution
model, or Next’s use of partner warehouses via Platform Plus, the value chain strengthens when
under tension.
Can Etsy replicate this? While from a consumer perspective RBL and Etsy may seem similar, the
business models are rather different111. Both RBL and Etsy offer artists global demand generation
in return for a take rate. But Etsy’s creators are themselves responsible for not only designing but
also manufacturing and distributing the products. Likewise, Shopify may offer the online store
and third-party fulfilment integration, but it is not a marketplace. There are similarities to Next
plc112 and Farfetch here: business models which are marketplaces and operating systems, as such
addressing the shortcomings of global leaders in both, Amazon and Shopify.
Merch by Amazon has been in existence since 2015 and does not seem to have inhibited RBL’s
ability to grow customers, fulfilment partners or numbers of artists and their designs. Merch by
Amazon is a vertically integrated model with a narrower product set and more limited
geographic reach (US mainly). There is no separate identifiable consumer proposition and so it
doesn’t speak so well to the artist’s desire to be acknowledged, nor their ability to develop a
fanbase. As to the supplier proposition, Merch by Amazon has no artist profiles/avatars, perhaps
impeding consumers’ ability to develop an affinity for artists and their work. RBL’s decentralised
production and heavily localised fulfilment network are resilient and scalable. Sometimes
perceived as ‘inefficiency’, supply chain robustness is an underrated business quality in an
109 Merch by Amazon is vertically integrated; AMZN owns the fulfilment capability and manufactures close to
distribution. This makes sense for a player with such immense scale. For RBL owning the production assets would
mean a step change in capital intensity and a brake on discretionary cash flow generation.
110 The products onto which designs are printed.
111 Etsy is an indirect competitor. Redbubble, Zazzle and Society6 print visual art onto standardised blanks. Etsy is
not about standardised products but bespoke ones.

112 Next also has a mission to provide suppliers the lowest cost route to market.
173
uncertain world: the ability to be a reliable partner to suppliers and to delight customers in the
face of supply chain disruption, trade wars and pandemics!
RBL has benefited from competitor mis-execution. Zazzle launched before RBL but has
materially lagged RBL’s progress in artists, designs, customers and fulfilment network. Society6
has also lagged RBL’s product breadth and geographic reach. Zazzle and Society6 focused on
picking and promoting the artist designs they thought would sell the most, vs. organic discovery.
This upset the applecart, annoying artists and disturbing VCS. CafePress failed to scale due to a
vertically integrated model of owning the printing machines.
RBL competes with unlisted operators (Threadless, Zazzle, Design by Humans), which lack
access to capital markets, and companies which are smaller parts of larger groups (Merch by
Amazon, Society6/Graham Holdings, CafePress/PlanetArt), which compete against other
divisions for resources. Meanwhile RBL is a mission focused business solely dedicated to
delighting artists and customers.
RBL’s medium term revenue growth expectations are 20-30% pa. and EBITDA margin rising to 13-
18% over the next three years. These growth targets are low vs. RBL’s pre-pandemic five-year
CAGR of 35% and Etsy’s pre-pandemic five-year CAGR of 40%. RBL estimates that ecommerce
spend in current product categories and geographies to be c. US$300bn, of which “35-40% of
customers are seeking something unique and meaningful”. RBL’s GMV of c. AUD 630mn is <0.5%
of this. Discretionary spending on personalised mid-tier priced products across apparel and
accessories, and home décor and furnishings is clearly a huge market globally.
The potential for high incremental returns on investments are supported by RBL benefiting from
the capital investments of its third-party fulfilment network, and the marketing activities of its
artist community. With RBL’s greater scale and SKU range, it may enjoy better ROIs on
advertising spend. This is because greater brand awareness will aid conversion, and greater
design and product range increases the prospects for long tail conversion. So as the company
grows, it may be optimal to increase marketing spend, enjoying superior unit economics,
maximising SVA, and widening the distance between RBL and competitors. The ROI on ad spend
improves as SKU variety enhances search results and puts customers in front of the entire
product set, not just the products they searched for. Marketing spend should increase with
greater ‘surface area’.
Yet this course of action, despite being in the interests of long-term value creation, will impede
the demonstration of operating leverage and higher accounting profit margins. This seems to be
behind the large decline in RBL’s quoted price since the start of 2021, potentially presenting a
profitable time arbitrage opportunity to long duration investors.
RBL’s model is scalable. Consider similar models such as Fiverr and Etsy, which also connect
creative vendors and designers to consumers, but whose scalability is impeded by a one-to-one
mapping of time. A spectrum of scalability exists where the more bespoke the product or service,
the less scalable the model. Another example is expert networks and interviews. Bespoke one on
174
one interviews are less scalable than user generated content platforms such as Stream and
Tegus. Essentially, selling one’s time has a growth and economic value ceiling. Rather than
selling time, RBL’s artists are selling (sometimes quite evergreen) designs, creating the possibility
of very high and durable marginal gains per unit sold.
RBL has grown sales, artists, designs and customers strongly at relatively low cost; sales and
marketing expenses have been c. 10% vs. peers at 11-26%. A higher-than-average take rate of c.
30%113 also bodes well for operating leverage on SG&A114.
Avenues to improve the monetisation of RBL’s constantly growing content include improving
brand awareness115 and customer repeat rates. RBL’s customers transact 1.2x pa on average.
Repeat purchases are 42%116 of marketplace revenue (MPR) and are growing 67% pa (vs. +52% for
first time purchases)117. We might expect SKU variety improvements to enable repeat custom as
consumers are more likely to find what they are looking for. Brand awareness will lower the cost
of growth by increasing ROI on ad spend. Repeat rates will do the same and can be achieved by
improving the first-time customer’s understanding the depth and breadth of the marketplace.
Intent based search is responsible for most content discovery today; this is an impediment to
both customer acquisition cost (Google in the funnel) and customer lifetime value (repeat rates)
improvement. While this is an efficient process with decent conversion, it offers limited
opportunity for customers to become familiar with all RBL has to offer. The challenge for RBL
will be to better present that value proposition without distracting from the transaction funnel
and in the process harming conversion. Management has encouraged an experimental approach
to driving repeat custom. With 10mn customers, RBL seems quite well positioned to tease out
the most efficacious practices for driving repeat custom via improved understanding of what
RBL is.
But there is a tension between the company achieving its target of first transaction profitability,
which requires a very efficient transaction funnel, and opening the channel, which may allow for
greater understanding of the proposition, but jeopardise those initial transaction economics.
Management demonstrates an understanding of this tension and a long-term mindset that is
willing to consider disrupting short term financial progress in the pursuit of more durable and
dependable value creation:
113 Etsy take rate c. 17%, AirBnB 13%, Uber 22%, Booking 14%, Zillow 14%, GrubHub 20%, eBay 8%, Next plc 39%
(Tollymore holding), Farfetch 30% (Tollymore holding).
114 Gross margin has increased from 34% to 40% FY16-21, and opex/MPR has fallen from 48% to 33% over the same
period.
115 Former Chairman Richard Cawsey: “The opportunity is for Redbubble to become a consumer brand, rather than a
place people find”.

116 This has not declined post lockdowns, an encouraging marker of the quality of customers acquired throughout the
pandemic.
117 These repeat customer weighting and growth numbers are very similar for Etsy. However, Etsy has more
customers (half of the total) who transact more than twice per year.
175
“We’re trying to move the business from single-transaction focus to much more around total
customer order rate and average annual order value. We think that they are the right metrics for
us to be thinking about. At the moment, there is a lot of focus around user conversion and AOV;
that is right, as it’s the first-time funnel. But the next evolution is to move from first-time funnel
to customers’ order rate, annual order value/lifetime value. That is an evolution in the business
that we need to go on.”
Today MPR excluding masks is c. AUD 500mn, EBITDA margin 10% and S&M 12%, implying
$110mn of owner earnings, or a yield of 10%. This is not an appropriate valuation for a special
business with substantial profitable growth potential 118.
Thank you for your partnership.
Mark
118 Note that valuation comparisons with Etsy must be adjusted for each company’s different definitions of revenue .
Using Etsy’s definition RBL, excluding masks, generated c. AUD 190mn of revenues last year, c. 6x EV/sales, vs. Etsy at
15x. Etsy’s expected revenue growth of > 30% for FY21 implies an 11x multiple. It is not appropriate to justify this
difference with respect to profitability and cost structures. Using Etsy’s definition of revenues, RBL’s current EBITDA
margin is similar to Etsy’s at 32%. And management’s medium-term guidance of 13-18% of MPR implies c.45% of
Etsy-definition-revenues.
176
A letter to partners, December 2021: shrinking supply chains part
II
Dear partners,
Shrinking supply chains: part II
Naked Wines plc (WINE.LN)
In our search for value creating supply chain shrinkage, Tollymore made one addition to the
portfolio in 2021. Naked Wines (WINE) is a platform that connects wine drinkers in the US, UK,
and Australia with independent winemakers. Wine drinkers can learn about winemakers, read
30mn reviews of wines, purchase wines and provide feedback on the wines they buy.
In a conventional supply chain, winemakers buy grapes, make wine, bottle and label it, and
market the wine prior to getting paid. It is a working capital-intensive supply chain with
uncertain payback and much effort focused on marketing, promotion, and distribution.
WINE’s model is to agree orders in advance at a fixed fee per bottle, paid upfront, mitigating
winemakers’ business risk associated with demand uncertainty and working capital depletion.
WINE has almost 1mn ‘Angels’ who contribute £25/$40 per month into a piggy bank. In return
Angels get: Angel only prices (up to 33% discount); access to Angel exclusive wines; a free sample
every month when ordering a case; and a no-quibble guarantee – WINE will credit Angels back
for anything they don’t enjoy.
The Angel monthly subscription float is used to back c. 250 independent winemakers to make
wine exclusively for WINE customers. The larger this float becomes the more volume WINE can
commit to purchase and for longer, the more winemakers it can onboard onto the platform, and
the lower the per unit costs of dry goods such as bottles and corks. These efficiencies, along with
customer insight data, are shared with winemakers to offer better value wine to customers.
Problems faced by winemakers are global but different in WINE’s markets. In the UK and
Australia, the principal challenge is consolidated retail. The same problem exists in the US, but
this problem is compounded by the gatekeeper role of the big distributors, who have little
commercial incentive to allow independent producers access to market, impeding their ability to
scale. The consequence for US consumers is the most expensive wine in the world. As such the
US is the market with the greatest consumer surplus, making it particularly ripe for business
model innovation.
In a nutshell WINE offers winemakers higher profits: the same gross profit margins but larger
and more predictable volumes. WINE pre-commits to future production using Angel
prepayments and review data, giving winemakers a secured route to market and peace of mind.
The precommitments solve two problems: (1) they improve winemaker production efficiency
177
(cost per bottle to produce), and (2) the guaranteed audience saves money on marketing and
sales and distribution teams. Winemakers can increase earnings by removing middlemen, have
volume predictability and a relationship with the end consumer.
Solving problems for the supply side is often a cornerstone of value chain symbiosis (VCS) when
creating value for vendors creates value for consumers. WINE is addressing the challenges
winemakers face in the conventional supply chain: access to capital, distribution, and
consumers. It does this by creating value in two ways: (1) A shrinking supply chain, and (2)
offering non-monetary benefits to suppliers (greater certainty) via crowdfunding and flipping
the working capital burden to consumers. WINE allows the winemakers to build their own
brands and fanbases – there is no Naked Wines wine brand. This may shrink the supply chain
less than a model such as HelloFresh, which has one single brand and therefore less supply chain
economic leakage. But it allows for brand affinity and winemakers to develop relationships with
their customers. This makes sense when there is an emotional connection to the product, an
emotional connection that is harder to establish with potato farmers for example. WINE
simplifies the lives of winemakers and allows them to focus on making great wines. WINE also
creates predictable demand; this predictability and peace of mind is a non-monetary stakeholder
incentive that enables value chain shrinkage and VCS. That WINE can attract some of the world’s
best wine producers, such as Jesse Katz, Matt Parish and Daniel Baron, is testament to the
supplier proposition.
There seems to be tremendous potential value in allowing independent vendors, whether they
are artists, fashion brands or winemakers, to participate in a more level playing field. To allow
them to compete on quality and value rather than fight against the legacy constraints of
conventional business models. To not only enable the customer proximity, but to benefit from
the collective demand generation engine of the platform – the ‘marketplace + operating system’
proposition.
There seems to be a ‘right place right time’, or antifragile, element to the relationships that WINE
has strengthened with winemakers during COVID, which has accelerated the challenges of the
traditional supply chain. WINE has been able to lean in and solve a problem for its winemaker
partners in a time of crisis by providing a demand outlet to compensate for the enormous loss of
on trade.
WINE's proposition addresses a disgruntled part of the value chain - the winemaker - and solves
a problem for them by being an operating system. A legacy from the prohibition era, the US
market structure is regulated according to three separate tiers: producers, distributors and
retailers. Alcohol producers are required to sell to wholesale distributors, who then sell to
retailers, who sell to the end consumer. An exception to this mandate is that wineries can sell
directly to consumers, both in person (at a tasting room) or by shipping to the consumer in states
that allow it (45 states). Producers cannot distribute their own wine; for a century consolidated
178
distributors have acted as a “legally required middleman that doesn’t have a single reason to
provide value because your job is secured by the law”119.
Because WINE funds and enables the production of wine, it is classified as a winery and enjoys
the ability to sell and ship directly to consumers. By disintermediating the middlemen of the
conventional supply chain – the distributors and supermarkets – there are more dollars to share
with fewer components, and WINE can facilitate proximity, connection and affiliation between
winemakers and consumers.
The three-tier system makes the value of a shrinking supply chain even higher – there are more
steps to disintermediate and the value extraction by distributors makes for disgruntled value
chain participants. This is an industry ripe for VCS via business model innovation because
customers are not served well by incumbents. This can be objectively supported by the low
portion (16%) of the cost of a bottle of wine that goes on what you can taste – the winemaking,
barrels, and grapes. The remainder is spent on marketing, middlemen and packaging 120. By
cutting out the middlemen and pooling marketing dollars, WINE can sell $10 of wine costs for
$25, vs. $60 retail. VCS is achieved by offering the customer affordable luxury while earning >
50% product gross margins.
WINE is not simply a marketplace but a vertically integrated model that showcases 1.5k exclusive
wines and creates collective demand for independent producers. This is a key business model
difference that is appealing given my evolving preference for operating systems/true platforms
over simply supplier aggregation. This model is harder to achieve but increasingly difficult to
replicate at scale. The WINE proposition takes care of production, marketing/demand
generation and distribution, and flips a working capital cycle from positive to negative. This
allows winemakers to focus on making great wine and benefit from collective bargaining power
of a platform in the commoditised aspects of the value chain such as bottling, shipping, and
marketing. Allowing creative vendors to focus on the aspects of their business they most enjoy is
a characteristic of other Tollymore holdings including Farfetch, Next plc and Redbubble.
The customer proposition is nothing more complicated than cheaper, better wine. By shrinking
the supply chain WINE can reduce the costs that cannot be tasted – marketing, bottling,
distribution. Wineries typically spend more on advertising than production. The lower carbon
footprint associated with collapsing supply chains, vertical integration and the superior
inventory management resulting from data-led customer insights will also potentially appeal to
the sustainability zeitgeist of many consumers. WINE has treated customers honestly.
Management has refused to exploit Veblen economic characteristics in the US market to extract
short term profits, and during COVID the company sent an email to all customers reminding
119 https://news.nakedwines.com/2015/10/27/an-open-letter-to-the-3-tier-system-from-scott-peterson-california-
winemaker/
120 Like Gillette razor blades - this is a weak proposition that has lasted for decades, enabled by inertia and misguided
regulation, and therefore ripe for disruption. Business model rather than technological innovation can result in
counter-positioned incumbents. Can the supermarkets compete without cannibalising their highly lucrative off sales?
179
them that in these hard times the money in their Naked piggy bank belonged to them and they
could withdraw it any time they wished.
Like Redbubble, WINE is turning traditionally adversarial relationships into collaborative ones.
WINE is supporting the winemaker, and this role is appreciated by customers who can play their
part to help independent producers while enjoying objectively better value wine. This is the
essence of win-win-win.
WINE competes with many smaller scale DTC wine clubs. The biggest seem to be Virgin in the
UK and Winc, Plonk Wine Club and 90+ Cellars in the US. But these operators are still 10-15% the
size of WINE. Naked’s wines are exclusive and so cannot be found on these other websites. The
DTC operators compete with conventional retailers such as Costco or Tesco’s, who are
disadvantaged by a longer supply chain and encumbered by the three-tier system in the US. A
large portion of supermarket sales are <$10 commodity wines, where WINE does not compete.
Vivino has the largest SKU range and rating pool of any online DTC operator. But it is a
marketplace on behalf of retailers. It may look similar from a consumer perspective, but the
business model is very different to WINE. While WINE is shrinking the supply chain, taking on a
production role and assuming funding risk on behalf of its winemakers, Vivino is doing the
opposite, by adding a fourth margin after the retailer. Wine.com and Vivino are competing for
knowledgeable wine drinkers with typically higher end wine. This is because wine.com is a 30%
gross margin retail model but is still encumbered by the three-tier model, and Vivino makes
money by lengthening the value chain. It is essentially a price for choice, funded by collective
demand generation dollars. These are not examples of business model disruption. These models
work on high end, excess stock SKUs where, for example, a $100 bottle of wine is discounted to
$50 and there is still enough contribution for everyone to get paid. The inability to monetise the
everyman wine drinker restricts the profitable TAM for these businesses, which are simply
migrating existing discovery and distribution online.
WINE principally leverages a cost advantage facilitated by middleman disintermediation. It

enjoys growing scale advantages associated with single sided (reviews) and double sided
(marketplace) network effects, as well as traditional scale economies in the form of falling unit
production and unit logistics/shipping costs as WINE pushes more volume through its fulfilment
centres.
The growing Angel float allows WINE to commit to higher levels of production for existing
winemakers and extend its proposition to new winemakers, including the best in the industry.
Greater scale allows WINE to increase commitments without increasing purchasing risk. The
growing Angel base also enriches the review repository, providing better feedback to
winemakers and improving the quality of the wines being produced for this exclusive platform.
Of course, taste is personal, and so WINE’s ability to track individual Angel feedback and use
tools such as Never Miss Out (19% penetration) and Wine Genie (2% penetration) will be
important in curating this ever-expanding wine selection. This curation is vital in mitigating
plateauing utility per SKU of the platform.
180
The richer feedback data also allows for superior inventory management, the efficiencies from
which can be reinvested into a better supplier and customer offer. In short, there are multiple
intangible aspects to WINE’s competitive lead. A lead which seems hard to assail with capital
alone.
A key component of WINE’s winemaker proposition is the vertical integration – the operating
system that allows smaller independent winemakers to participate in a more level playing field.
Yet many of the characteristics of the business model are like the world’s best marketplace
models: (1) supply is fragmented and global vs. consolidated and local; (2) supply is exclusive; (3)
non-monetary incentives governing platform behaviour, such as complexity reduction, supplier
autonomy, scarcity, customer feedback, community, desire to help the underdog, emotional
purchasing, demand certainty, and customer data and proximity; (4) merit vs. advertising based
supply curation; (5) a trusted and well-liked platform; and (6) high (and increasingly passive)
purchase frequency; and (7) low demand promiscuity/incentive to switch at the point of sale.
Reviews also enable a quality advantage as underperforming wines are improved or removed
from the selection. Like Redbubble, supplier success on the platform is a function of merit,
rather than advertising dollars. The challenge for these growing platforms is to curate the
content and ensure relevance for consumers. Customer engagement has been rising strongly;
posts per Angel have increased threefold and repeat customer contribution product has doubled
over the last two years.
The working capital ‘burden’ of the conventional vertically integrated supply chain is borne by
the customer in this case. The opportunity cost of committed capital is a tiny price to pay for
access to exclusive high-quality peer-reviewed wine at 30% discounts. The visibility associated
with future demand, and access to customer data are the non-monetary benefits available to the
winemaker that make VCS possible.
A spectrum of scalability exists where the more bespoke the product or service, the less scalable
the model. Redbubble’s content is very scalable because artists create and upload it once, but it
can be consumed multiple times across multiple products and purchases for many years. Etsy, on
the other hand is less scalable because products are manufactured and distributed by the
creators themselves, and they are consumed once by the customer. Netflix is a hybrid case
because content created once can be consumed by multiple consumers, but likely only once per
consumer. Wines can be consumed repeatedly by the same consumers; the winning formula
needs to be created only once, but the wine needs to be manufactured each time. Hence
production scale is important, and the utility of an operating system proposition is high, as it is
with Redbubble.
Wine consumption per person has historically been non-cyclical and grown c. 2% pa. Like the
low-cost gym businesses in which Tollymore has invested, we are not betting on a structural
change in human behaviour, but rather the appeal and sustainable profitability of a different
mousetrap for consumption. In such bets, the key ingredients for durable, profitable enterprise
181
value growth are a wide proposition gap vs. incumbents and other challengers, a large
addressable market, and unfair advantages that grow with scale.
The value of off premise wine consumption in the US alone is c. $40bn, more than 200x WINE’s
US sales. After deducting what management believes to be non-addressable (infrequent wine
purchases and <$10 price points), the company estimates a US TAM of c. $20bn.
Like many industries, COVID accelerated online purchasing trends for alcohol. It may be the case
that the last 12-18 months of lockdowns have habituated the online purchasing of wine. There is
some evidence to support this: repeat customer sales retention was 80% in 1H22 vs. 79% in 1H20;
and ecommerce trends have not returned to prior levels.
CEO Nick Devlin believes WINE can enjoy a 10% share of the US TAM ($2bn, >10x US revenues).
There is evidence of some progress towards this internal target: market share in the UK, where
WINE has been trading longer, is higher, and in some US states market share is in the double
digits.
It is possible that WINE’s authentic no-stigma approach to connecting wine drinkers to the
winemakers may encourage more people to drink high quality wine, previously put off by the
intimidation, informational asymmetry, and stuffiness of conventional methods of discovery,
selling and distribution. But the principal opportunity to grow is one of DTC market share take
driven by a large and sustainable proposition gap vs. the conventional supply chain.
WINE has focused its efforts at the $15-30 price range and has built distribution facilities that
allow it to ship wine at reasonable cost; shipping is free for minimum orders of six bottles in the
US or £100 in the UK. WINE seems to have a logistics architecture and scale that has suppressed
the cost to ship wine. It costs a business $50-200 to ship a case of wine in the US; WINE pays a
fraction of this. This is an advantage that grows with scale; WINE fulfilment costs per order have
fallen from $40 to $30 over the last year. DTC peers have focused on higher priced wines given
the difficulty in economically shipping heavy glass bottles in temperature-controlled conditions.
Having cracked the harder (the lower end) part of the market, WINE now can produce and
distribute more expensive wines and enjoy the associated superior unit economics.
The unit economics of higher priced wine are likely to be superior because (1) it weighs the same
as lower priced wine, and (2) the value proposition is even greater. One of the challenges that
WINE has faced in the US is the Veblen nature of the product; low price is often seen as
synonymous with low quality. A bottle of wine in the US costs more than virtually anywhere in
the world. WINE needs to educate without patronising wine drinkers in this market that they are
the victims of an archaic distribution system. And that the more expensive a wine becomes, the
higher the percentage of the cost is allocated to non-wine costs (marketing and three sets of
margin dilution).
182
Innovators have benefited from incumbents’ ignorance of low-cost challengers in other
industries121, possibly a result of counter positioning. Perhaps WINE can exploit it here too.
Management’s choice not to exploit the Veblen characteristics of this market in the pursuit of
ephemeral profit extraction supports long duration capital allocation and a business owner
mindset.
The cost of WINE’s growth is low. New customer cohorts deliver a cash contribution of 2.5-3x
CAC within five years, and the payback has been increasing over time. Payback on ad spend was
unusually effective during COVID induced lockdowns as wine consumption shifted to the off
trade and customers sought out online channels while stuck at home. The larger the customer
base the cheaper the additional customers due to the power of referral; c. 60% of WINE’s website
traffic is direct, and three quarters of search traffic is organic.
Management has a clear focus on growth profitability, monitoring and reporting cohort
paybacks in some detail. Broadly, as the company has grown, so too has the return on CAC. In
turn, the amount invested into customer growth has increased, begetting larger scale and so the
cycle repeats. We can observe a similar phenomenon with Trupanion’s customer acquisition
flywheel; short sellers seem to have persistently misinterpreted increasing CAC as something
outside of the company’s control.
Wine Genie is a nascent subscription product that could be very important in improving the
profitability of WINE’s future growth. Subscribers to Wine Genie automatically receive cases of
wine selected by the company at a cadence of the customer’s choosing. Wine Genie customers
rate the wine they receive, and WINE adjusts the algorithm to deliver wine in future that
increasingly matches the customer’s tastes. This allows the customer to eliminate the time
required to review and choose wines, to experiment and drink wines that are increasingly well
curated. Wine Genie could be very important because it makes the renewal decision passive by
outsourcing and personalising wine consumption, potentially increasing customer satisfaction,
lowering CRC and churn. The data flywheel turns faster as Wine Genie customers are more
incentivised to provide feedback. It could also improve working capital turns and lower
inventory and shipping costs as WINE is choosing which wine is shipped. Finally, Wine Genie
may increase the TAM to less confident/educated and more time poor customers, without
competing for the attention of those, usually older, consumers that know what they like and
want control over their choices. The challenge for WINE is determining which customer to
target with each proposition. The reward for getting this right will be better cohort economics,
reinvested in faster business growth.
CEO Nick Devlin is a clear thinker and speaks plainly despite his strategy consulting background.
He seems to share a first principles mindset with founder Rowan Gormley who has said that he
never thought about culture, but just “built a group of people with passion and a shared goal”.
121 Steel mini mills, transistor radios, PCs, supermarkets, airlines, online encyclopaedias, video streaming, gyms
183
Mr. Devlin demonstrates an interest in VCS to defend against replication:
“I’m always more interested in ways to make production more efficient and share that value back
with consumers and build a higher retention business that is harder to replicate and has a deeper
competitive moat, than taking the short-term opportunity to put it into a higher contribution
margin”
There is a clear focus on long term profit maximisation. Mr. Devlin outlines the two paths to high
long term operating profitability: (1) increase contribution margins by sharing fewer scale
efficiencies with consumers, and (2) hold contribution margins but reinvest into better
consumer value, increasing retention and lowering the cost to replenish lost custom. The
company is pursuing the second path because it is much harder to compete with.
Standstill EBIT, that which would be reported if investment in new customers was reduced to
the level needed to just replenish the current customer base, is c. £40mn on a rolling 12-month
basis. This is a 10% yield to the current market cap, adjusted for net cash that does not seem to be
required to fund business activities122. These owner earnings are being invested at IRRs of 30%+.
This could imply a mid-20s equity IRR assuming no rerating of existing owner earnings.
Standstill EBIT assumes a repeat contribution margin of 29% currently reported 123, sales
retention of 82% and a year one payback of 70%. Assuming WINE continues to share improving
scale economies with the value chain, preserving rather than growing contribution margins and
year one payback, and that it can increase retention by 1ppt per year, then 20% repeat sales
growth would flow through to owner earnings, or business value, growth some 50% higher at
>30% pa.
Despite the ostensibly very different business models, there are several business characteristics
and valuation merits in common with one of Tollymore’s other holdings, Trupanion. Both are
recurring revenue business models with customer retention c.80%; both are examples of
business model innovation disrupting incumbent distribution and working capital cycles; both
are taking a third of industry volume growth; owner earnings margins are 12-15%; and most of
these owner earnings are redirected into the business at 30%+ IRRs. It may be helpful to make
these observations given that Darryl Rawlings, Trupanion founder and CEO, recently joined
WINE’s board as Chairman. Darryl has created tremendous owner earnings’ growth at TRUP and
has presided over a multi-fold increase in the company’s quoted equity value in the three years
since we acquired shares. Having run this disruptive DTC pet insurance company for 20 years in
the US, he likely has valuable insights to share regarding shareholder communication, internal
stakeholder incentives and motivation, and possibly the redirection of WINE’s primary listing to
the US, where TRUP trades at an owner earnings multiple almost three times higher than WINE.
122 A capital return has been considered by the board.

123 This is mid to high 30s in the US, the fastest growing market, currently half of revenues.
184
How are these owner earnings being directed? For every £1 WINE spends on acquiring new
customers, it makes £0.70 - £1 in contribution within a year. This contribution attrits at 15-20% pa,
delivering a five-year payback of 2-3x. Even at the recent trough five-year payback of 1.7x, the
IRRs on new customer acquisition are likely to be very high, at least 25%. If we assume two thirds
payback over the first 18 months (1H21 year 1 payback = 67%), the IRR is closer to 50%.
We are not relying on recreating the wheel here, nor predicting lasting shifts in consumer
preferences. This is a bet that people will continue to drink wine, and prefer cheaper, higher
quality wine. Finally, we are making a bet that WINE can do this very profitably and the way they
do it is hard to copy.
What is so defensible or special about WINE that prevents the value of the shortened supply
chain accruing to consumers and/or producers at the expense of owners? There are many,
individually small, characteristics of WINE’s model that may compound to form a tough to
replicate business, including: (1) Marketplace attributes such as single- and double-sided
network effects; fragmented global supply and demand; exclusivity/scarcity; non-monetary
incentives such as feedback, sense of community, ability to support independent vendors,
curation, trust, high frequency, emotional and passive purchase patterns; high retention due to
low incentive to switch at the point of sale. And (2) Operating system attributes such as
complexity reduction; vendor recognition and a level playing field; supplier autonomy; supplier
business risk reduction and greater security.
Thank you for your partnership; I hope you have an amazing 2022.
Mark
185
Appendix: podcast interview - Cultivating Independent Thought
Capital Employed: Hello. Welcome to the Capital Employed podcast. In this episode, I had the
pleasure of talking to Mark Walker, the managing partner at Tollymore Partners. Since launching
the funding 2016 until the end of 2020, Tollymore has generated a return of 28% per annum net
of all fees and expenses, which is a great performance. So, I was keen to get Mark onto the show
to learn more about his investment style and philosophy. In this episode, Mark also talks about
two companies that he feels have great long-term potential. I really enjoyed listening to him and I
think you will too. Hi Mark. Thanks for coming onto the podcast.
Mark Walker: Thanks for the invitation to talk with you. My pleasure.
Capital Employed: Can you give a brief overview of Tollymore Partners and what the investment
style and philosophy is?
Mark Walker: So Tollymore manages capital for its principals and a small group of partners who,
I'd say, have demonstrated some ability to think somewhat unconventionally. We are trying to
make decisions in the interest of long-term results, which means that I think that most of those
results should really come from the internal learnings power growth of the companies we hold.
What it means is that we don't expect good results to come from any IQ advantage, but from very
simple synergies between the components of an investment firm. So, the investment horizon,
temperament, working environment, incentives, and investment partners. That last element, the
investment partners is really crucial, I think. I think that flourishing relationships drive enduring
contentment. And so, we're trying to discover wine-win in investing; that's really Tollymore's
reason for being.
Capital Employed: And I believe you're quite concentrated as well in your approach. Is it
between 12 and 15 holdings?
Mark Walker: Yeah, that's right. There are no hard rules around this, but the remit is 10 to 15
securities. We've gravitated towards the lower end of that range.
Capital Employed: And what type of businesses do you like to invest in? What characteristics do
you normally look for?
Mark Walker: We look at all kinds of businesses. In a nutshell, all we're trying to do really is to
find really good, really cheap businesses. I'll come back to that idea of cheapness in a minute, but
our edge doesn't come from being analytically better than anybody else. It just comes from
something simpler than that. It's a willingness to look at everything, a willingness to just do some
digging into the characteristics of the businesses as you refer to it that might be conventionally
regarded as low quality, and there's no predetermined ways I'm trying to form a positive
conclusion on those characteristics. What's important is that each potential investment
186
candidate is assessed at a face value and using available facts. And we're just trying to extract
insights from first principles as much as possible.
And by doing that, I'm trying to avoid analytical errors that are often made by using mental
shortcuts. I think that the more we shorten the description of an investment process, the more
memorable it is for sure, but I don't think that that means it's more repeatable. And the second
point about cheapness, I'd say that we're value investors because our goal is to pay prices that
allow us an equity IRR at least as good as the company's own intrinsic value growth. But I think
that labels, like ‘value investor’, they restrict the freedom and the honesty with which we can try
to understand the investment credentials of companies. And I think what we've seen in recent
years is that those labels really give an excuse to underperformers investing according to style
factors, which is the unpopularity of their particular style. And that's led to a failure to own up to
bad investment results.
So if there's a single observation that has informed what value investing means to me, it's that the
cost of growth for really great companies has just been systematically overestimated throughout
time. Visa and MasterCard are the often-cited examples of this, businesses that have been
consistently under-priced. You could've paid eighty or a hundred times after tax profits a decade
ago and still earn a market return. That's because of this heuristic that high growth warrants a
high multiple and doesn't consider the cost of growth, which is the proportion of prior year
earnings needed to achieve that growth. And it underscores why for us business quality
standards are just non-negotiable and those standards should be applied to all investment
opportunities, whether you're classically regarding yourself as value or growth or GARP. I think
an interesting example, a more modern example is that if you think about from first principles,
whether the economic characteristics of modern SaaS companies are desirable. I think that it's
clear that most market participants believe these business models to be vastly superior to
software of the past.
And that observation alone already kind of tips the odds against you of achieving a decent result.
But if you dig into to the business model and understand that this transition that's taken place, to
vendor hosted subscription models, certainly seems to have better unit economics than software
of the past in the medium term and the web-based customer adoption models that have taken
out the customer acquisition friction, that's what's ignited much more rapid growth again in the
medium term. But if you then consider the very long term, if our only advantage here is a time
arbitrage one, this very factor that's ignited the rapid growth in enterprise software businesses,
which is the removal of this friction, has that also removed the software moat, the switching
costs? Now I use this example partly from a place of bitterness, right? Because we haven't owned
SaaS companies. Tollymore has not owned pure play SaaS companies. It's not owned big tech.
It's not owned healthcare or consumer staples. And that's not because of an anticipation of the
market's pendulum swing and what the value investor might describe as a mean reversion, but
it's really just an output of requiring individual stocks to jump very high quality and valuation
hurdles. On the one hand, our inability to think very creatively about when prior base rates just
187
aren't relevant anymore, that's cost our investment partners because we've not owned these
sectors that have undergone substantial re-ratings in recent times.
But on the other hand, price absolutely matters. Microsoft took 15 years to reach the 2000 highs
despite very strong and steady business progress. And price matters because it's one of the very
small number of determinants of IRR. Rather than asking the question will SaaS company
valuations mean revert, which might be implied by prior cycles. We try and ask, is it really
obvious to me that by buying this multiple of company X, that my IRR will be really quite
remarkable? And I think it's those types of microeconomic discussions and thoughts that we find
interesting, and we think are worth doing if we have a chance of creating value.
Capital Employed: What are your valuation parameters? Is there just a certain PE multiple or do
you use discounted cash flow? How do you come to your valuation?
Mark Walker: We try and think about valuation from multiple angles and triangulate those
angles. We're typically not building detailed financial models and anticipating a future
predetermined path or outcome for a company. A central evaluation framework that we do
typically rely on is one of Bruce Greenwald's EPV valuation framework, where we're just trying to
think about what are the owner earnings of this business? So, what is the cash the company
could put in the pockets of owners today if it invested what it needed to, to preserve unit output
and competitive positioning. And what is that as a yield to the cost of ownership of the
company?
And then the second step thinks about what the company is going to do, what it should do and
what it's going to do with those owner earnings and what kind of incremental returns are
available and for how long? When we get really excited, is when we think that the unit
economics are very different to the aggregate economics of the business. The aggregate
economics that are obviously presented in financial statements and are appearing on various
screens. And so, when we think that they're not reflecting the economic reality or the economic
potential of the company, that is often the ignition for our work.
Capital Employed: Can you talk us through two companies in your fund that you feel have been a
bit underappreciated by the market and you feel have good long-term potential?
Mark Walker: Sure. So I'm speaking from the UK today. Tollymore's based in London. And there
are two UK businesses that we purchased in the last 12 months that I think have a very
underappreciated and outsized potential for extraordinary equity IRRs, and they're Next PLC and
Farfetch. Next is sort of typically, I think, perceived in the UK as this very credible, safe, single
fashion brand operating distributing products predominantly through brick-and-mortar outlets
across the UK. But I think that that reality has slowly been changing and that change has been
accelerating over the last year or 18 months. The reality is that Next is actually a multi geography,
multi brand, omni channel retail operating system. And so, some of these changes, as I said, I
188
think the changes initially started happening around five years ago when Next opened up it's
offering to other brands via a vertically integrated wholesale model.
And then more recently since 2019 via a marketplace model. And that's been extended then to an
operating system for third party brands in 2020. In the provision of these services for brands and
suppliers, the Next store network, which is obviously typically seen as a low quality way to
operate a business is actually a very crucial component of the online value chain. And that's for a
couple of reasons. One is that most orders are collected or just over half of orders are collected in
the stores. And around 80% of online returns are returned to the store. Secondly, the delivery is
around two thirds of fulfilment cost; typically, the last mile is the most expensive component.
And so, by saving that and reinvesting that into the consumer proposition, Next is in this, I think,
unique position in the UK apparel industry. Aspects of the competitive positioning are
consistent with what we have called over the years value chain symbiosis, which is basically,
there's a very strong customer and supplier proposition.
The consumer proposition comes from two factors of online retailers: one is product choice,
much greater SKU variety. And the other is convenience. The former I think is much more
important than the latter, which is good for Next because I think the former is much more
difficult to replicate than the latter. And then the supplier proposition is one of providing a multi
brand context in which to sell the supplier's products. 80% of shopping has historically taken
place in multi-brand, whether that's offline or online multi brand environments. And that's the
limitation of pure play operating system models such as Shopify. And then the desire for direct-
to-consumer relationships or customer proximity is the limitation of pure play marketplace
models, the most obvious being Amazon. And so, by providing this sort of hybrid solution, Next
allow brands to avoid the fixed cost step ups associated with growth and allows them to focus on
the customer and the product, which is what they enjoy doing.
In return for around 39% commission, those brands get access to 8 million customers growing
25% a year. They have the ability to control price and representation of their products. They have
access to a next day delivery network and returns to Next stores and a consumer lending service
as well administered by Next. I think that a lot of Next progress has been driven by some
reasonable managerial foresight, but also they've enjoyed having the right model at the right
time, which is something we've observed in the number of our holdings. COVID has really
shaken up the UK apparel industry and there's been a lot of restructuring and there's the stores
that have fallen into administration difficulties. There's a high cross shop between Next and
those stores. So that's how we determine what the owner earning of this business might be like
and how sustainable they are by virtue of this competitive positioning.
And then the question is, well, what are the profitable avenues for redeploying these owner
earnings? I think the growth opportunity is attractive through the lens of a UK apparel market,
which is a hundred times larger than Next multi-brand store. It's really the quality of the supplier
and consumer proposition, I think, that should lead to continued penetration online. And then
189
you have adjacencies into more SKUs and more brands, but also more verticals such as home and
beauty. So, I talked about this heuristic, this error that high multiples are typically warranted for
high growth companies without really understanding the cost of growth. Well, the cost of this
growth for Next is very low. As they increasingly direct capital investments to warehouses versus
retail stores, well, those have very high asset turns, five or six times asset turns and they're
generating 20% after tax margins. And the incremental returns on their online marketing efforts
overseas are in the order of two to 400%.
What are you paying for this business? Well, I think the owner earnings are in the order of a
billion pounds. So, you're paying around 10 times for that, which I think assumes no value
creation. And I think that everything I've touched upon to me suggests that there's plenty of
evidence to the contrary.
The second business, Farfetch, in many ways has sort of similarities to Next through a different
vertical and it's a much more global business. It's a luxury online marketplace for brands and
retailers and consumers. Like Next, it also offers what they call Farfetch Platform Solutions, a
white label operating system. So, they're effectively providing the Amazon and Shopify of luxury.
And the consumer proposition is very similar. It's that most products are sold in multi-brand
environments. The marketplace model is superior to the vertically integrated peers that Farfetch
has because of greater SKU variety, which facilitates this rapid iteration. Understanding very
quickly which products are working, at which price points and how to present those products.
And it also facilitates a kind of newness or freshness, which is so crucial in the luxury industry.
And again, the supplier proposition is one of luxury brands being very preoccupied with brand
integrity, wanting control over pricing and visual representation. And superior unit economics
for brands who earn much higher gross margins by partnering with Farfetch. Competitive
positioning is facilitated by global two-sided network effects, which are strong by virtue of very
fragmented supply. Luxury inventory is distributed by a very fragmented network of boutique
sellers. They don't really have any competition with this business model. They're the only scaled
luxury marketplace.
And also like Next, I think they've been in the right place in the right time, which is that they've
demonstrated anti-fragile qualities. The two years where Farfetch has really thrived in my
opinion are 2008 to 2020. Periods of distress as their partners really leaned into those
relationships. The times when the boutiques were either forced to shut their doors or were
facing a material drop off in demand, they provided this outlet, this ability to keep the lights on.
So that's something that's compounded the goodwill in the relationships they have with those
brands over the last decade plus. The growth opportunity is characterised by very large $400
billion global luxury market, which might be a $30 to $40 billion online commission pool, versus
a couple billion of revenues for Farfetch. And they have a very clear framework for value
accretive investment, very trackable definitions of what a lifetime value and an acquisition cost
might look like.
190
I think that there's been a lot more optionality for exponential value growth coming out of China
with a partnership that they've struck with Alibaba and Richemont. China's been a very, very
difficult market for many Western businesses to crack and Farfetch has been no exception.
They've tried different partnerships, which haven't really worked too well. But now this new
partnership really gives them access to around eight hundred million consumers on Alibaba's
Tmall. And so just having a very small modicum of success in penetrating that could be really
transformational to their ability to succeed in China. Again, through this owner earnings
valuation lens, I think it's a business that generates four to 500 million of owner earnings, and
that's being reinvested in really quite extraordinary incremental returns. Even if you make very
modest assumptions about the useful economic life of customers and order frequency and
contribution margins, et cetera. So, I think it's a business that can earn an equity IRR of 30, 40%
plus for quite a long time.
Capital Employed: Thanks Mark for sharing those two companies. Are you a keen reader? And if
so, have you read any book recently that you really enjoyed.
Mark Walker: Yeah. I think that the book I've most enjoyed reading this year is the School of Life
by Alain de Botton. The book really discusses the gaps that might exist in our formal education
journeys. My son has just started school this year and so it's something I'm kind of thinking about
more and more. But efforts, I think, in education are typically focused on becoming successful in
technical areas, right? Science and maths rather than emotional areas. This book certainly
contends there's insufficient effort focus directed to being successful in relationships or
managing anxiety or how to be kind. And so, what de Botton is arguing really is this has sort of
led to this somewhat lopsided evolution where we've had this very rapid and vast technological
progress, but that's been managed by humans who emotionally have evolved very little.
And I think that this book has some interesting ideas about how to let go of this myth of a perfect
life. If you're trying to be emotionally mature, right? No one is normal and actually the
recognition, if you can recognise and accept the flaws in ourselves and others, that can be quite a
healthy motivator for empathy and kindness. Ostensibly, I guess a non-investing book, but
emotional intelligence and self-awareness I think are superpowers in investing. And so, they're
worth striving for. I think the common thread that I find with good books, the ones that I enjoy,
which are probably the minority of the ones that I read, is not that they give me this feeling that I
have the world figured out, that they're not there to make me sound smarter, but actually, they
do the opposite, right? They reiterate how little I know. They entrench, hopefully, this humility
that I think is required for successful investing.
Capital Employed: Thanks so much, Mark, for coming onto the show. It's been a pleasure to
listen to you.
Mark Walker: It's been really nice to chat
191
Disclaimer
The contents of this document are communicated by, and the property of, Tollymore Investment
Partners LLP. Tollymore Investment Partners LLP is an appointed representative of Eschler Asset
Management LLP which is authorised and regulated by the Financial Conduct Authority (“FCA”).
The information and opinions contained in this document are subject to updating and
verification and may be subject to amendment. No representation, warranty, or undertaking,
express or limited, is given as to the accuracy or completeness of the information or opinions
contained in this document by Tollymore Investment Partners LLP or its directors. No liability is
accepted by such persons for the accuracy or completeness of any information or opinions. As
such, no reliance may be placed for any purpose on the information and opinions contained in
this document.
The information contained in this document is strictly confidential. The value of investments
and any income generated may go down as well as up and is not guaranteed. Past performance is
not necessarily a guide to future performance.
192

Tollymore Letters To Partners Dec 2021

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Tollymore Letters To Partners Dec 2021

Uploaded by

Copyright:

Available Formats

Letters to partners

A letter to partners, December 2018

Tollymore investment results ................................................................................................................................ 2

USD Tollymore (gross) Tollymore (net) MSCI ACWI

Cumulative 235% 194% 109%

So the cornerstones of my investment philosophy are thinking long-term, maximising the

Portfolio management: The portfolio is concentrated, typically holding 10-15 securities. If I

Investment case study

• Expansion of the global travel industry.

4He refers to this as “optimising the multiyear financial outcome”

• You want smooth returns.

Philosophy: why quality matters

Process: competitive advantage

While an appreciation of frameworks such as Porter’s Five Forces is useful in understanding

Execution: making better decisions

Case study: Gym Group plc

TripAdvisor Inc: an update

Thank you for entrusting me with your capital. Yours sincerely.

Balance sheet risk

Several developments challenged this thesis:

In addition to making mistakes relating to disaggregating business quality from favourable

Thank you for entrusting me with your capital.

As an emerging investment management firm Tollymore has the advantage of organically

ITE Group (ITE.LN)

What should a demanding investor be willing to pay for:

• A simple, founder-led business.

Wishing everyone a healthy and happy year ahead,

My professional experience has provided a broad overview of the institutional investment

24 Examples may include primary research and idea funnels.

Narrative fallacy is the backward-looking mental drive to attribute a cause-and-effect chain

The forest and the trees: profiting from time arbitrage

services to attractions owners for free.

And on the FY17 conference call:

Thank you for entrusting me with your capital.

With my best wishes,

A history of business models

The economic characteristics of platforms

The challenges in building and protecting valuable platforms

One of the principal risks associated with linear

Persistently low scoring passengers are now also being banned:

Implications for owners

Appraising addressable markets when the future ≠ the past

We initially acquired our ownership of Sotheby’s in May 2016

When we initially invested, we identified some opportunities to improve return on operating

49 E.g. incentive pool as a percentage of EBITDA.

owning the item in advance, in doing so placing a higher value on it.

Service digitisation: The digital delivery of services could make the

55 Increasingly through an online consignment platform.

David Foster Wallace

There is evidence to suggest that the method of wealth

The role of relationships in leading a good life

“An important dimension of successful investment management organisations

Charles Ellis, The Characteristics of Successful Investment Firms

There is a natural alignment between an investor with a long-term, possibly perpetual,

59 Harvard Study of Adult Development

• 1% management fee and 10% of any investment returns above a 5% hurdle

Narrative fallacy and the Murray Gell-Mann Amnesia effect

We purchased an equity interest in Grubhub (GRUB)

The restaurant proposition is the generation of higher

“…battleground in the escalating competition for dominance…”

GRUB has an opportunity to redeploy discretionary profits into value-accretive activities:

Thank you for your partnership.