You are on page 1of 88

Deutsche Bank

Markets Research
Global

Cross-Discipline

Date
12 September 2013
Jim Reid

Nick Burns, CFA

Long-Term Asset Return Study

Strategist
Strategist
(+44) 20 754-72943 (+44) 20 754-71970
jim.reid@db.com
nick.burns@db.com

A Nominal Problem

Seb Barker
Strategist
(+44) 20 754-71344
sebastian.barker@db.com

________________________________________________________________________________________________________________
Deutsche Bank AG/London
DISCLOSURES AND ANALYST CERTIFICATIONS ARE LOCATED IN APPENDIX 1. MICA(P) 054/04/2013.

Deutsche Bank
Markets Research
Global

Cross-Discipline

Date
12 September 2013
Jim Reid

Nick Burns, CFA

Long-Term Asset Return Study

Strategist
Strategist
(+44) 20 754-72943 (+44) 20 754-71970
jim.reid@db.com
nick.burns@db.com

A Nominal Problem

Seb Barker
Strategist
(+44) 20 754-71344
sebastian.barker@db.com

As we publish this annual report, optimism on global growth is rising and there
is also much discussion on the Feds plan to taper unconventional monetary
policy. However, as we highlight, the 5-year moving average of global nominal
GDP growth is now at its lowest rate since the 1930s. In the US, which is one
of the bright spots globally, nominal GDP growth has been at 3.1%, 3.1% and
3.8% in Q2 13, Q1 13 and Q4 12 respectively. These numbers are lower than
where they were in the prior two quarters (4.8% and 4.5%) when QE infinity
was being formulated and announced. If we had a nominal GDP target we may
now be discussing increasing QE and not tapering.
Any recovery should be seen in this context. Given the structural issues that
we think will continue to hold back growth, unconventional monetary policy
may actually need to increase in the years ahead. However, given the far
superior performance of asset prices relative to economic activity since QE
started, perhaps how monetary policy is distilled through the economy
needs to be improved.
Expanding traditional QE might not be the answer. We think that more debate
is needed on policies that directly target nominal GDP and not just asset prices.
Perhaps the groundwork is currently being laid for this by the blurring of lines
between governments and central banks. In Japan Abenomics is fostering a
deeper partnership between the two. In the UK, the government specifically
headhunted new BoE Governor Mark Carney and altered the BoEs remit, and in
the US President Obama is about to hand-pick Bernankes successor. The ECB
is institutionally an outlier but even Draghi has stepped beyond his inflation
remit with his whatever it takes speech last summer. Globally the next few
years may bring politicians and central bankers closer together and monetary
policy that directly targets growth over financial assets.
Weve previously been of the opinion that the end-game to the 2008 financial
crisis is notably higher inflation at some point in the second half of this decade.
We still think this is likely but only if unorthodox monetary policy continues
over the next few years.
In terms of preferred assets for the long-term investor, while we dont expand
on work done in earlier studies our bias remains for Investment Grade
Dividends i.e. IG-type companies but owning their equities over their debt
on a valuation basis. The cheaper names are in Europe rather than the US. The
overall US equity market looks stretched relative to history. Whilst fixed
income markets also offer little long-term value, we think that central banks
will still be forced to keep yields artificially low for as long as they can. Credit is
a fairly low beta but unexciting asset class at the moment worth the
incremental pick-up over government bonds in a low default environment, but
not one likely to see exciting returns.
Overall we think many global assets have been inflated by QE and central
banks may need to spend the next few years engineering higher nominal GDP
to justify such valuations.

________________________________________________________________________________________________________________
Deutsche Bank AG/London
DISCLOSURES AND ANALYST CERTIFICATIONS ARE LOCATED IN APPENDIX 1. MICA(P) 054/04/2013.

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Table Of Contents

Interesting Stats on a Page ................................................. 3


Executive Summary............................................................. 4
A Nominal Problem ........................................................... 11
Global nominal slowdown led by DM ............................................................... 11
Why is nominal GDP so important? .................................................................. 14

Why is Nominal Growth so Low?...................................... 17


(I) Are demographics lowering potential growth? ............................................ 17
(II) Has growing state involvement slowed growth? ........................................ 20

What Drives Real and Nominal Growth? .......................... 26


Real GDP Growth, Revolution Ending? ............................................................. 26
Nominal GDP Growth, Impossible without Money Creation/Innovation ........... 31

Putting Recent Central Bank Action in Context ................ 39


QE not enough to offset financial crisis is inflation terms ? .............................. 43

The Monetary Playbook..................................................... 45


Is Nominal GDP Targeting the Answer? ........................................................... 45
Are the Helicopters Coming? ............................................................................ 51
NGDPT and Helicopter Money pose Deeper Questions then Any Framework
and Policy Yet Used .......................................................................................... 54
Capitalism on Hold: Japan ............................................................................ 55

100 Years of the FED ......................................................... 59


Mean Reversion................................................................. 61
Assessing the mean reversion model through time .......................................... 61
Mean reversion across asset classes ................................................................ 66
Mean Reversion Assumptions .......................................................................... 70

Historical US Asset Returns .............................................. 72


Historical International Asset Returns ............................... 76
International equity return charts ..................................................................... 76
International 10 year government bond return charts ...................................... 77
International Equity minus Bond return charts ................................................. 78
International return tables ................................................................................ 79

All data in this report is up to the end of August 2013 where possible.

Page 2

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Interesting Stats on a Page

In spite of recent positive growth indicators, any recovery should be seen


in the context of nominal GDP growth that has been trending lower across
the globe. The 5-year moving average (MA) growth rate of our global
nominal GDP series is currently at the lowest its been since the 1930s.

DM growth has been progressively slowing for more than a decade but
EM has until now been the global engine. At the extremes, since the end
of 2004, our EM universe has grown nearly 210% (BRICs 267%) in nominal
GDP terms (converted into dollars) with the Eurozone only 13.5% bigger.
Since 1995, the EM number is 482% (BRICs 691%) with the Eurozone only
54% larger.

However, EM nominal growth is now slowing with the 5-year MA at the


lower end of an albeit healthy 50-year range. China is starting to see a rare
period (for the last 35 years) of sub-10% nominal GDP growth.

Since the end of 2007, we estimate that the global economy has
potentially lost around $41tn of cumulative nominal GDP against the prior
trend (relative to around $64tn size of our global economy sample at end2012). DM and the Eurozone have lost up to $33tn and $13tn respectively
over the same period.

The annual output of the world, DM and Eurozone economies would now
be $13.2tn, $10.5tn and $3.9tn bigger if nominal growth had been 7%, 5%
and 4% respectively since the end of 2007.

Perhaps the $7.5tn expansion of the six major global central banks since
the Lehman default looks less aggressive when seen in this context.

Given the $13tn of lost Eurozone output, its interesting that the ECB
balance sheet has only expanded by about $1tn over the past 5 years. Over
this period combined European bank balance sheets are flat at around
$32tn after having increased by $13.5tn in the prior 5 years.

Its not just the 5-year MA thats weak. H1 13 has seen low nominal
activity. Indeed in the US (one of the brighter spots), nominal GDP growth
has been 3.1%, 3.1% and 3.8% in Q2 13, Q1 13 and Q4 12 - lower than
where they were in the prior 2 quarters (4.8% and 4.5%) when QE infinity
was being formulated and announced. If we had a nominal GDP target we
may be discussing increasing QE at this juncture and not tapering.

Growth is a modern phenomenon, nominal growth even more so. Before


1750 there was hardly any of either and before the 20th century nominal
growth often lagged real growth as real growth led to price falls in a
broadly fixed hard metal currency system. This perhaps helps prove that
money creation/innovation remain key to nominal growth.

In a US 60/40 equity/bond portfolio our mean reversion model suggests


10-year annualised returns of only 2.77% p.a. the fourth-lowest in the
143 years since 1871. The only years with a lower 10-year prediction were
in 1998, 1999 and 2000.

Deutsche Bank AG/London

Page 3

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Executive Summary
As we compile this report, which looks at longer-term themes in financial
markets, there is optimism that recent data is suggesting an imminent
rebound in growth, particularly in the developed world. Also, as has been
well flagged and debated all summer, this month will likely mark the point
where the Fed starts to taper their $85bn per month bond-buying program.
They seem confident that stronger activity is just around the corner. However,
what concerns us is how low nominal GDP growth has been in recent
quarters across the globe and indeed how weak the post financial crisis
nominal recovery has been in spite of seemingly aggressive monetary policy.
So any recovery should be seen in this context. This piece argues that
unconventional monetary policy may actually need to increase over the years
ahead. However, given the far superior performance of asset prices relative
to economic activity since QE started, perhaps how it is distilled through the
global economy needs to be enhanced.
A nominal problem
H1 2013 saw weak nominal activity across the world. Indeed in the US, which
is one of the bright spots globally, nominal GDP growth has been at 3.1%,
3.1% and 3.8% in Q2 13, Q1 13 and Q4 12. These numbers are lower than
where they were in the prior two quarters (4.8% and 4.5%) when QE infinity
was being formulated and announced. If we currently had a nominal GDP
target we may be discussing increasing QE at this juncture and not tapering.
Figure 1: Nominal US GDP Growth At Lowest Since QE1 & Lower Than Start
of QE Infinity
YoY Nominal GDP Growth (LHS)

4.0

5%

Fed Balance Sheet ($tn, RHS)

3.5

4%
3%
2%

-1%

-4%
2007

QE1

-3%

QE2

-2%

2008

2009

2010

2011

QE3

0%

Operation Twist

1%

2012

Operation Twist become outright


purchases

6%

3.0
2.5
2.0
1.5
1.0

0.5
0.0

2013

Source: Deutsche Bank, Bloomberg Finance LP

This report argues that nominal GDP is important as we live in a nominal world.
We receive wages, pay our debts and manage our savings in nominal terms. In
the current environment, we continue to believe that nominal GDP is more
crucial than normal as we have record and climbing levels of global debt which
is virtually all nominal. Asset prices are also tied to nominal activity over the
medium-long run. Its impossible to get revenue growth detaching from
nominal activity over any sustainable period and as such the valuations of
assets like equities will be heavily influenced by nominal GDP.
In this piece we construct a comprehensive nominal global GDP series (split by
regions) back to the late 1920s and find that the 5-year moving average global
nominal growth rate is now at its lowest level since the 1930s. This has been
driven by the developed world. But even in EM, growth in many
countries/regions is flirting with the lower end of the most recent decade range.
Page 4

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

In Figure 2 and Figure 3 we show the data back 50 years (the period where our
data is most comprehensive). In the main report we extend back to 1928
where possible.
Figure 2: Nominal GDP Growth World (left, Log Scale), G7 (middle), DM (right)
1954 1964 1974 1984 1994 2004
100%

10%

World

1%

5yr MA

20%

G7

20%

5yr MA

15%

15%

10%

10%

5%

5%

DM

5yr MA

0%

0%

-5%
1954 1964 1974 1984 1994 2004

-5%
1954 1964 1974 1984 1994 2004

Note: 5yr MA is 5 year moving average


Source: Deutsche Bank, GFD

Figure 3: Nominal GDP Growth Eurozone (left), EM (middle, Log Scale), BRIC (right, Log Scale)
25%

Eurozone

1000%

5yr MA

EM

10000%

5yr MA

20%
15%

BRIC

5yr MA

1000%

100%

10%

100%

5%

10%

10%

0%

-5%
1954 1964 1974 1984 1994 2004

1%
1954 1964 1974 1984 1994 2004

1%
1954 1964 1974 1984 1994 2004

Note: 5yr MA is 5 year moving average


Source: Deutsche Bank, GFD

Why has growth been so weak in this recovery?


This question has vexed the greatest minds in economics and the financial
industry but one would have to say that too much debt has been a hindrance
to many whereas trying to reduce it too quickly (austerity) has been an issue
for others. The problem may actually be that growth was too high in the
leverage bubble of the decade that preceded the financial crisis. As such we
are flat-lining until we catch-down with the appropriate new trend rate of
growth. Perhaps this trend rate of growth has been declining due to
demographics and this is now slowly being exposed post crisis. This is more
true of the developed world but even in EM many countries are either past or
are fast approaching their demographic peak.

0.5

Source: Deutsche Bank, UN Population Database

Deutsche Bank AG/London

Source: Deutsche Bank, UN Population Database

0.4

World
BRIC
EM

0.3
0.2

G7
DM
Eurozone

2050

2045

2035

2025

2015

2005

1995

1985

1975

1965

1955

2045

Europe

2035

EM

2025

50%
2015

-5%
2005

DM

1995

BRIC

1985

G7

1975

World

1965

55%
1955

0%

2040

60%

2030

Europe

0.6

2020

5%

65%

2010

EM

2000

DM

0.7

1990

10%

70%

1980

BRIC

1970

G7

Figure 6: Productivity Ratio (35-54yr


vs. <24yr & >65yr)

1960

World

15%

Figure 5: Working Age / Total


Population

1950

Figure 4: Working Age Population


Growth

Source: Deutsche Bank, UN Population Database

Page 5

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

We also argue that the DM and EM world has interfered with the forces of
creative destruction and capitalism post crisis which has structurally lowered
trend growth even if such policies prevented an even deeper crisis post 2008.
Do we have a divine right to growth?
Economic history tells us that growth is a modern phenomenon, only emerging
on a consistent basis from the middle of the eighteenth century.
Figure 7: Annual Global Real GDP Growth
6%
5%
4%
3%

2%
1%
0%

-1000000
-25000
-8000
-4000
-2000
-1000
-500
-200
14
350
500
700
900
1100
1250
1340
1500
1650
1750
1850
1900
1925
1940
1955
1965
1975
1985
1995
2011

-1%

Year
Source: Deutsche Bank, Delong, World Bank

The economic literature suggests that long-term growth derives from


improvements in technology and in economic organization to deploy that
technology, from capital intensity and also increasing labour input. Over the
long run and on a per-capita basis the first factor, technological improvement,
is king. An influential 2012 paper by Robert Gordon suggests that the easy
growth era could be over and that the growth seen over the last two or three
centuries has been driven by three industrial revolutions which have
dramatically changed the economic landscape. He argues most of the
innovations in the last decade or so have been based on communications and
entertainment which are less growth-enhancing than prior leaps forward.
Indeed he contrasts the invention of running tap water and flushing toilets with
the modern day Facebook era of inventions and argues the former is
considerably more growth-enhancing.
Regardless of the outlook for real GDP, its fair to say that nominal GDP can be
manipulated and that in an economy with no monetary or velocity expansion,
nominal GDP growth will always be zero whatever the value of real GDP
growth. This is based on the MV=PY identity. So if real GDP (Y) increases due
to productivity gains, whilst the money supply (M) and velocity (V) remain
constant then prices (P) will have to fall to offset increased real activity.
Nominal GDP (PY) remains constant.
If one believes this narrative then nominal GDP growth is, always and
everywhere a monetary phenomenon and central banks and financial
institutions have the ability to heavily influence nominal GDP regardless of the
real level of activity. To see the importance of todays monetary system for
nominal GDP growth, we note how during the nineteenth century the real GDP
of both the US and the UK regularly grew at a faster rate than nominal GDP.
This suggests that the authorities didnt have the ability or the willingness to
expand/manipulate the money base enough to keep up with real GDP growth.
As such there was constant downward pressure on prices.

Page 6

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Figure 8: US (left) and UK (right) 5-Year Moving Average of YoY Real and Nominal GDP Growth
25%

Real GDP

25%

Nominal GDP

20%

20%

15%

15%

10%

10%

5%

5%

0%

0%

-5%

-5%

-10%

-10%

-15%
1789

1900
1829

1869

1909

1949

1989

Real GDP

Nominal GDP

1900
-15%
1831 1851 1871 1891 1911 1931 1951 1971 1991 2011

Source: Deutsche Bank, GFD

In fact its useful to remember that inflation is largely a modern day


phenomenon and economic progress prior to the twentieth century was often
met with positive deflation. In the last century, central banks and creditcreating institutions have basically ensured that nominal GDP growth is now
above real GDP growth everywhere. This is an entirely artificial and
manipulated construct but one that encourages us to believe there is a way of
elevating nominal GDP if there was the appetite.
Central banks have been seen to be aggressive post 2009 but have they
actually been too timid? Or perhaps pursuing the wrong target?
If we look at the main six central banks that have actively expanded their
balance sheets post crisis, the aggregated dollar value of their holdings have
doubled to over $14.5tn since the Lehman default. In trying to put this in
perspective, the left-hand chart of Figure 9 adds the annual flow of their
balance sheets to the annual nominal GDP of these countries (all converted to
USD). The right-hand chart of Figure 9 then looks at this on a YoY basis.
Figure 9: Nominal GDP Plus Central Bank Flows of Six Key Global Central Bank Countries Levels ($tn, left) YoY (right)
50

Nom GDP (Annual)

CB Balance Sheet

Nom GDP

Nom GDP + CB BS

15%

45

10%

40

5%

35

0%

30

-5%

25
20
Mar 02

20%

-10%

Mar 04

Mar 06

Mar 08

Mar 10

Mar 12

-15%
Dec 00

Dec 03

Dec 06

Dec 09

Dec 12

Source: Deutsche Bank, Bloomberg Finance LP

This is highly simplistic and ignores multipliers (albeit ones that are currently low)
but puts the monetary expansion seen so far in some context. Although central
banks have generally been seen to have been aggressive over the last 5 years,
the sizes of their interventions are not substantial versus the annual size of their
respective economies. We also calculate that the global and G7 economies have
potentially lost around $41tn and $25tn of cumulative output relative to what
trend growth might have been expected to be from the end of 2007 up to June
2013. Again this puts the scale of recent central bank actions in context.
Deutsche Bank AG/London

Page 7

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Figure 10: Global Annual Increase in CB Balance Sheets ($tn) vs. Annual
Nominal Loss of Output Relative to LT Trend
14

Annual Global Increase in CB Balance Sheets ($tn)

12

Annual Nominal Loss of Output Relative to LT Trend ($tn)

10

8
6
4
2
0
2008

2009

2010

2011

2012

2013 (LTM)

Source: Deutsche Bank, GFD, Bloomberg Finance LP

So theres perhaps a debate to be had that monetary policy needs to expand


still further globally to ignite nominal GDP. However maybe there is an
argument here for broader and better targeted policy, directed more towards
the economy rather than the current situation where asset prices appear to be
the main beneficiary. Contrast Figure 11 below with Figure 1 that showed the
Feds balance sheet and nominal GDP.
Figure 11: Have Asset Prices Benefited Most from QE To Date?
1,800

S&P 500 (LHS)

4.0

Federal Reserve Balance Sheet (RHS, $tn)

1,600

1,200

600

QE2

200

QE1

400

QE3

800

Operation Twist

1,000

Op. Twist -> outright purchases

3.5

1,400

0
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013

3.0
2.5

2.0
1.5
1.0
0.5
0.0

Source: Deutsche Bank, Bloomberg Finance LP

In this note we discuss the pros and cons of nominal GDP (NGDP) targeting
and how helicopter money is potentially the final untried monetary policy left
post-crisis and one that might be directed more towards the economy rather
than financial assets.
We think nominal GDP is crucial in this cycle as the debt burden remains
incredibly high relative to history. The sooner we can start to meaningfully
erode it, the sooner we can reduce its inherent systemic risks and potentially
free up animal spirits. Nominal GDP is also important to revenues and with it
equity prices. One concern is that QE has to date brought forward tomorrows
equity returns today.
Indeed our mean reversion exercise suggests that projected 10-year US equity
returns are back down to an annualized 3.3% over the next 10 years. Backtesting this model, the predicted 10-year annualized return didnt fall below 5%
in any year between 1871 and 1997 (Figure 12). So this shows that we still live
Page 8

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

in a world of elevated equity valuations relative to history using our preferred


long-term valuation techniques. It doesnt mean that positive returns wont be
seen but it perhaps shows the impact of central bank liquidity in bringing
future returns forward. Can the US afford to reverse course aggressively at this
stage? Do they need to try to increase nominal activity to allow markets to
catch-up with the valuations theyve helped engineer? Maybe QE isnt the
most effective way of achieving this which brings us back to the debate about
the possibility of helicopter money in future years.
Figure 12: S&P 500 Mean Reversion Expected 10yr

Figure 13: Mean Reversion Expected 10yr Annualised

Annualised Returns vs. Actual (1958 Method)

Returns vs. Actual for a 60/40 US Equity/Bond Portfolio

30%

Mean Reversion

Actual

25%

20%
15%
10%
5%

0%
-5%
1871 1886 1901 1916 1931 1946 1961 1976 1991 2006
Source: Deutsche Bank, GFD

20%
Mean Reversion
Actual
18%
16%
14%
12%
10%
8%
6%
4%
2%
0%
1871 1886 1901 1916 1931 1946 1961 1976 1991 2006
Source: Deutsche Bank, GFD

Figure 13 expands the exercise to show the annualized 10 year mean reversion
returns of a portfolio weighted 60/40 US equity/bonds. This chart again
highlights what a low return world were potentially in and how careful central
banks might need to be. The trick is to increase nominal activity to allow
equities to grow into their valuations whilst also ensuring that bond yields
dont rise dramatically thus hurting the typical equity/bond portfolio. Before
1997 the model never dipped below a projected 4% p.a. return over 10 years.
Since 1997 the only years the model went slightly back above it were in 2002
and 2008-2009. The realised annualized 10 year returns of this portfolio since
the late 1990s have generally been as low as the model suggested they would
be. The current prediction of 2.77% p.a. return is the fourth-lowest in the over
140 years since 1871. The only years with a lower 10-year prediction were in
1998, 1999 and in 2000.
So the model suggests its going to be very difficult to generate real returns
from this starting point. The long-run average inflation rate of the US since
1871 is 2.4% which if repeated would imply a negative real return from this
starting point.
Such an exercise is not easy to repeat across the globe as the US is one of the
few countries that have long histories of growth, inflation, earnings, PE ratios,
and bond yields without going through huge permanent structural change
(through politics, war etc). However we do argue in the report that European
equities are cheaper than those in the US on a mean reversion basis.

Deutsche Bank AG/London

Page 9

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Conclusions
There are many reasons why nominal growth has disappointed since the crisis.
Deteriorating demographics are likely now becoming increasingly important
after being swamped in the pre-2008 leverage boom. Also not allowing more
creative destruction post the financial crisis is perhaps contributing to weak
growth performance relative to the previous trend. Propping up bubble-era
debt with ultra low interest rates and QE has arguably locked in an inefficient
allocation of resources throughout the developed world. EM have also been
increasingly guilty of such activity post 2008. In an ideal world we would have
liked to see more cleansing of debt over the last 5 years which would have
helped eventually free up animal spirits, encouraged a more efficient resource
allocation and allowed for more new entrepreneurial activity to prosper.
However this would have likely had a dramatically negative short-term impact
on the economy and possibly on social cohesion. Politicians needing to be
elected would also have been unlikely to sign off on such policy. As such we
have to be realistic enough to assume that this path is now unlikely to
materialize. The authorities therefore have two options if growth continues to
be so moribund. They can either continue with the just-in-time management of
the problem that has existed since 2008 or they can start to be more radical
and consider options that look a lot more like helicopter money. Given the
worsening demographic outlook and the still systemically high debt levels such
a bold approach might eventually be needed.
Expanding traditional QE might not be the answer. We think that more debate
is needed on policies that directly target nominal GDP and not just asset prices.
Perhaps the groundwork is currently being laid for this by the blurring of lines
between governments and central banks. In Japan Abenomics is fostering a
deeper partnership between the two. In the UK, the government specifically
headhunted new BoE Governor Mark Carney and altered the BoEs remit and in
the US, President Obama is about to hand-pick Bernankes successor. The ECB
is institutionally an outlier but even Draghi has stepped beyond his inflation
remit with his whatever it takes speech last summer. Globally the next few
years may bring politicians and central bankers closer together and monetary
policy that directly targets growth over financial assets.
Weve previously been of the opinion that the end game to the 2008- financial
crisis is notably higher inflation at some point in the second half of this decade.
While we continue to expect such an outcome, it has always been predicated
on liberal money printing by central banks over the coming years. If Fed
tapering marks the beginning of the end to this policy globally then its unlikely
that inflation will be a big issue in the years ahead. However we think that the
reduction of global central bank liquidity in a high debt, poor demographic,
lower real growth world will eventually expose the globes economic problems
again which will inevitably lead to more monetary activism. So we dont see
the expected imminent US tapering as the end of unorthodox monetary policy.
In terms of preferred assets, while we dont expand on work done in earlier
studies our bias remains for Investment Grade Dividends i.e. IG companies
but owning their equities over their debt on a valuation basis. The cheaper
names are in Europe rather than the US. The overall US equity market looks
stretched relative to history. Whilst fixed income markets also offer little longterm value, we think that central banks will still be forced to keep yields
artificially low for as long as they can. Credit is a fairly low beta but unexciting
asset class at the moment worth the incremental pick-up over Government
bonds in a low default environment, but not one likely to see exciting returns.
Overall we think many global assets have been inflated by QE and central
banks may need to spend the next few years engineering higher nominal GDP
to justify such valuations.

Page 10

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

A Nominal Problem
We live in a nominal world. We receive wages, pay our debts and manage our
savings in nominal terms. While measuring all this in real terms may well be a
better measure of our changing relative financial position, the reality is that
money in day-to-day life is largely measured in absolute terms. In the current
environment, we continue to believe that nominal GDP is more crucial than
normal as we have record and climbing levels of global debt which is
measured in nominal terms.
What worries us today is that global nominal GDP growth has been trending
lower over the last year after what was a very subdued rebound post 2009.
Globally we have failed to return to the trend seen over the last several
decades, in spite of still strong EM growth in recent years. Will current
depressed levels be sustained or will we return to something approaching
previous trends? Also are we set up for this slower growth rate? The answers
will have a major impact on the ability to manage the excessive debt loads
most DM economies are still carrying and also be a huge influence on the
returns of all major asset-classes going forward.

Global nominal slowdown led by DM


Figure 14 and Figure 15 looks at our newly created long-term series of World
Nominal GDP (denominated in USD). Weve shown it on a log scale to visualise
the rate of change and this helps highlight the slowing pace of activity over the
past 5 years. In the left-hand chart we have constructed a data set that
aggregates activity in 39 of the largest 50 economies in the world (back to
1953), with Saudi Arabia the only G20 country missing from the sample. The
total GDP of our sample was $63.6tn as at YE 2012. The IMF suggests total
global GDP was just under $72tn at the same point. So our 39-country sample
covers around 89% of global activity. Figure 15 on the right extends the series
back to 1928 but prior to 1953 gaps appear in the data (only 24 countries are
therefore included). Of the top 20 by size, China (#2), Korea (#15) and
Switzerland (#20) are missing. The data around WWII is also missing for some
countries. Nevertheless the data again shows the recent slow-down in activity
in a wider historic context. The gap between current and prior trend nominal
GDP perhaps holds the key to many of the worlds recent troubles and
potential future problems.
Figure 14: World Nominal GDP Level and LT Trend ($tn,

Figure 15: World Nominal GDP Level and LT Trend ($tn,

Log Scale) since 1953

Log Scale) since 1928

100

World

LT Trend

100

World

LT Trend

10

10

0
1953 1960 1967 1974 1981 1988 1995 2002 2009

0
1928 1938 1948 1958 1968 1978 1988 1998 2008

Source: Deutsche Bank, GFD

Deutsche Bank AG/London

Source: Deutsche Bank, GFD

Page 11

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

In the charts that follow well concentrate on breaking down the global data by
region to highlight the evolving trends. Well mainly use the 1953 data set
given its higher level of completeness.
Developed market (DM) nominal GDP trending down
Most of the slowdown in global activity has been occurring in the developed
world. Figure 16 shows the same data but broken down for the G7 and DM
overall.
Figure 16: G7 (left) and DM (right) Nominal GDP Level and LT Trend ($tn, Log Scale) since 1953
100

G7

LT Trend

100

DM

LT Trend

10

10

0
1953 1960 1967 1974 1981 1988 1995 2002 2009

0
1953 1960 1967 1974 1981 1988 1995 2002 2009

Source: Deutsche Bank, GFD

The contrast is perhaps most sharp between the Eurozone and EM (Figure 17).
Figure 17: Eurozone (left) and EM (right) Nominal GDP Level and LT Trend ($tn, Log Scale) since 1953
100

Eurozone

LT Trend

100

EM

LT Trend

10

10

0
1953 1960 1967 1974 1981 1988 1995 2002 2009

0
1953 1960 1967 1974 1981 1988 1995 2002 2009

Source: Deutsche Bank, GFD

Indeed in dollar terms the Eurozone economy is only 13.5% bigger than it was
at the end of 2004 whereas the collective EM universe is 210% bigger (BRICs
267%). The world economy is 54% larger over the same period. Indeed in the
18 years since Q1 1995, the Eurozone economy is only 60% larger in nominal
dollar terms, as compared to 482% in the EM universe (BRICs 691%).
Looking at these growth rates graphically and in more detail, Figure 18 shows
the downtrend in the G7, DMs and in the Eurozone countries over the past 60
years culminating in the very weak post 2009 recovery. Here the growth
numbers for each country are calculated in local currency and then weighted
by their USD nominal GDP to calculate aggregated growth rates. In each of
these regions (obviously with some countries being a member of more than
one group), the 5-year moving average is the lowest over the 60-year period.
Page 12

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

The recovery since the perilous 2009 lows has been anaemic and in the
Eurozone current nominal growth is still less 1% YoY and close to zero on a
rolling 5-year basis. This will likely pick up from these depressed levels but the
charts put the overall level of recent activity in some historical context.
Figure 18: Nominal GDP Growth since 1953 G7 (left), DM (middle), Eurozone (right)
20%

G7

5yr MA

20%

DM

5yr MA

25%

Eurozone

5yr MA

20%

15%

15%

10%

10%

5%

5%

0%

0%

0%

-5%
1954 1964 1974 1984 1994 2004

-5%
1954 1964 1974 1984 1994 2004

-5%
1954 1964 1974 1984 1994 2004

15%

10%
5%

Note: 5yr MA is 5 year moving average


Source: Deutsche Bank, GFD

Indeed if we stretch the chart for the Eurozone countries back to 1928,
Figure 19 shows that European growth is now at levels not seen since the
1930s, apart from the obvious spike down in 2009. The 5-year moving
average is certainly now well below anything seen for over 7 decades. We
should note that there are data gaps during WWII but it shouldnt change
the overall message.
Before we dismiss this lower growth environment as a purely DM trend,
Figure 20 show that the EM (including the BRICs) world has recently
reverted back closer to the lower trend rate of growth seen post WWII. After
this 25-year period there is evidence to suggest that from the late 1960s to
the late 1990s, EM/BRICs saw a uniquely high level of nominal growth. Over
the past two years there has been a dip down in growth to the lower end of
the range of the last 15 years.

Figure 19: Eurozone Nominal GDP


Growth since 1928
50%
Eurozone
40%
5yr MA
30%
20%
10%
0%
-10%
-20%
1929 1944 1959 1974 1989 2004
Note: 5yr MA is 5 year moving average
Source: Deutsche Bank, GFD

Figure 20: Nominal GDP Growth since 1928 (Log Scale) BRIC (left), EM (middle), World (right)
10000%

BRIC

5yr MA

1000%
100%
10%

1%
0%
1929 1944 1959 1974 1989 2004

1000%

EM

5yr MA

1929 1944 1959 1974 1989 2004


100%

100%
10%

10%

1%

1%

0%
1929 1944 1959 1974 1989 2004

0%

World

5yr MA

Note: 5yr MA is 5 year moving average


Source: Deutsche Bank, GFD

So nominal GDP growth seems to be slowing everywhere. In the graphs


above, that stretch back to 1928, we only include countries with data over
the whole period. So a country like China, where our data only starts in 1953,
is completely excluded. To finish off this section we simply splice together
the growth rate of our 1954- series with the data from 1928-1953. This
allows us to look at the longer history whilst including the more complete
data set of the last 60 years. The conclusion remains the same with DM and
world growth trending lower and having just experienced their lowest 5 year
nominal growth rates since the 1930s (Figure 21 and Figure 22).

Deutsche Bank AG/London

Page 13

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Figure 21: World Nominal GDP Growth Based on 1928-

Figure 22: DM Nominal GDP Growth Based on 1928- and

and 1953- Data

1953- Data

50%

World

40%

5yr MA

DM

5yr MA

40%

30%

30%

20%

20%

10%

10%

0%

0%

-10%

-10%

-20%

-20%
1929 1939 1948 1957 1967 1976 1985 1995 2004

-30%
1929 1939 1948 1957 1967 1976 1985 1995 2004

Note: 5yr MA is 5 year moving average


Source: Deutsche Bank, GFD

Note: 5yr MA is 5 year moving average


Source: Deutsche Bank, GFD

Why is nominal GDP so important?


Figure 23 shows the total amount of G7 debt (public plus private) relative to
nominal GDP in these countries. It also shows the Debt/GDP ratio. Post the
financial crisis, the ratio remains stubbornly high. Without nominal GDP
growth it is very difficult to erode the debt burden at a rate quick enough to
remove the systemic dangers.
Figure 23: G7 Debt and Nominal GDP
160

460%

Total G7 Debt ($tn, LHS)

140

G7 GDP ($tn, LHS)

440%

120

Debt/GDP (RHS)

420%

2012

2011

2010

2009

2008

2007

2006

300%
2005

320%

0
2004

20
2003

340%

2002

360%

40

2001

60

2000

380%

1999

400%

80

1998

100

Source: Deutsche Bank, Haver

Putting the increases in debt in some perspective relative to growth and


central bank expansion, since the end of 2007 the G7 have added around
$18tn of debt, relative to only around $1tn of nominal GDP activity and nearly
$5tn of G7 central bank balance sheet expansion (Fed+BoJ+BoE+ECB). So in
the G7, which is a good proxy for the developed world, debt continues to
increase whilst nominal growth remains extremely low thus ensuring that the
deleveraging process has yet to start. At best were stabilising the ratio at or
around record highs.
In an ultra low interest rate environment (short and long-term rates), its
possible to carry this debt in a low growth environment but with little
deleveraging taking place it creates a fragile environment that leaves these
economies vulnerable to shocks and policy errors.

Page 14

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

If rates were to rise notably from these ultra low levels, this could be just such
a shock. This is why in spite of the recent sell-off, rates are likely to stay lower
for longer as the alternative could be highly destabilising given the extreme
debt burden being carried across large parts of the world.
Nominal GDP is also important for asset returns
Nominal GDP tends to drive many variables that contribute to the long-run
performance of various asset classes. Pension funds are an example of an
industry relying on past nominal GDP performance to justify future return
prospects.
Highlighting one fairly significant example, the Boston Colleges Centre for
Retirement Research who looked at 126 US state plans pension funding status
using the states own calculations found their total funding ratio at the end of
2012 stood at 73%. However between these plans was a large amount of
variation in funding levels driven at least in part by variation in discount (i.e.
expected return) rates from a high of 8.5% in Minnesota to a low of 6.25% in
Vermont. The report goes on to argue that using a baseline assumption that
equity returns will be 7.75% (on the Dow Jones Wilshire 5000 Index) which
would give a 2013 funding level of 78.8%, rising to 83.4% in 2016.
So really theres not much to worry about?
Maybe. Though its probably worth having a quick look how they got to the
rather confident 7.75% figure first. They assumed (all YoY) 3.5% real output
growth plus 2.25% inflation giving nominal growth of 5.75%. Then they argued
that profit growth will match output growth and the p/e ratio would be 17
giving stock price increases of 5.75%. Add on a 2% dividend yield and you get
the 7.75% number (see Figure 24 for the Baseline, Pessimistic and Optimistic
forecast breakdown).
Figure 24: Centre for Research Retirement YoY Equity Return Assumptions
Underlying

Baseline

Pessimistic

Optimistic

Real Output Growth

3.50%

2.00%

4.00%

Inflation

2.25%

1.50%

2.50%

Output Growth

5.75%

3.50%

6.50%

Profit Growth

5.75%

2.00%

8.00%

P/E End 2016

17

14

18

Stock Price Increase

5.75%

-2.50%

9.50%

Dividend Yield

2.00%

2.50%

1.50%

Equity Return

7.75%

0.00%

11.00%

Source: Deutsche Bank, CRR

As weve argued throughout this report, the growth numbers are pretty out of
step with recent trends. Looking at Figure 25 and Figure 26 it becomes clear
that the US economy hasnt sustained (here looking at 5-year averages) a 3.5%
YoY real GDP growth rate nor a 5.75% nominal growth rate since 1999 (i.e.
1995-1999).

Deutsche Bank AG/London

Page 15

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Figure 25: US 5Y Rolling Real GDP Growth vs. Baseline

Figure 26: US 5Y Rolling Nominal GDP Growth vs.

Assumption

Baseline Assumption

15%
10%

RGDP 5Y Rolling YoY Average

25%

NGDP 5Y Rolling YoY Average

RGDP Growth Baseline

20%

NGDP Growth Baseline

5%

15%
10%
5%

0%

-5%
-10%
1805 1830 1855 1880 1905 1930 1955 1980 2005
Source: Deutsche Bank, GFD

0%
-5%

-10%
-15%
1805 1830 1855 1880 1905 1930 1955 1980 2005
Source: Deutsche Bank, GFD

Such optimistic assumptions of pension fund investment returns are again not
just a US public pension problem. A recent KPMG report found that 350 of its
UK clients with defined benefit obligation pensions were expecting a 6.8%
return on equity and a 4.1% return on corporate bonds. An Aon Hewitt report
found that S&P 500 companies were expecting a long-term total return on plan
assets of 7.15% from 2012 onwards.
So a world with high debts and high expectations of future returns needs
nominal growth. The crucial and troubling question is why has it been so low
over the last few years?

Page 16

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Why is Nominal Growth so Low?


This subject has been a major topic of debate post the financial crisis and
there is still no universal agreement on the exact reasons. The factors
discussed have included there being too much debt, broken banking systems,
too much austerity/fiscal retrenchment, maybe too much intervention
interfering with capitalism, pre-2008 artificially high growth and that
demographic factors have peaked.
Of all these factors the one that perhaps is most worrying over the medium- to
longer-term is demographics as its the most clear-cut in terms of hard
numbers and perhaps the hardest to influence. Following on from this we look
at the role growing state involvement in economies in both the developed and
developing world have played in reducing nominal growth rates through a
weakening of creative destruction and market pricing discipline.

(I) Are demographics lowering potential growth?


One of the problems that may have helped to slow down growth has been
demographics a well-worn topic in earlier editions of this note. This is not a
trend that changes overnight but perhaps the rolling bubbles in the decade
prior to the financial crisis helped mask the deteriorating demographics.
Perhaps these are now being exposed.
In previous editions weve tied demographics more to its impact on asset
prices but here we correlate it more with growth. Figure 27-Figure 29 look at
aggregated G7 numbers. For these countries we have actual labour force
participation numbers back to 1970. This allows us to enhance any analysis
that simply looks at raw population numbers. The G7 should act as a very good
proxy for the entire DM universe. The LHS chart of Figure 27 starts by looking
at the overall size of the G7 working age population, the size of the labour
force and the actual total employment numbers. The middle chart shows the 5
year rolling change in the sizes of these groups whilst the RHS chart looks at
the percentage of the labour force and working age population actually
employed.
Figure 27: G7 Working Age, Labour Force and Employment Levels (mn, left), 5yr Change (middle), Employment as a
Percentage of Working Age and Labour Force (right)
Employment

700

600

Working-Age Pop

500
400

300
200
1971

1981

1991

2001

Employment
Labour Force
Working-Age Pop

12%
10%
8%
6%
4%
2%
0%
-2%

Labour Force

2011

0.64

% of WkAgePop Employed

0.62

% of LabFor Employed

0.60
0.58

0.56
0.54
1976

1986

1996

2006

1971

1981

1991

2001

2011

Source: Deutsche Bank, BLS

One of the problems for the G7 (which represents 51% of global growth) is that
the growth rate of the working age population has been slowing but that the
growth of those in the labour force and those actually employed are falling at a
faster rate. This is probably both structural and cyclical. The structural issues
could be due to the population spending longer in education (hopefully a
positive longer-term), earlier voluntary retirement (perhaps a waste of
resources) and maybe increased disincentives to work (too generous benefits
Deutsche Bank AG/London

Page 17

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

or lack of perceived opportunities). The cyclical is clearly more economically


sensitive and this last cycle has been particularly damaging on this front to
employment, especially for the young where there are astonishingly high rates
of unemployment.
An additional and very important structural issue is that the rise in women
working in the population seems to have plateaued after strong increases up to
the turn of the millennium. Figure 28 and Figure 29 demonstrate this by
showing the percentage of each sex working in the G7 relative to the total
population. We then also show this as a rolling 5 year percentage change.
Figure 28: G7 Employment to Population Ratio Weighted

Figure 29: 5 Year Rolling Change of the G7 Employment

by Working Age Population (%)

to Population Ratio

80
75
70
65
60
55
50
45
40
35
30

Women

Men

10%

Total

Women

Men

Total

8%
6%

4%
2%
0%

-2%
-4%
-6%

-8%
1976 1980 1984 1988 1992 1996 2000 2004 2008 2012

1971 1976 1981 1986 1991 1996 2001 2006 2011


Source: Deutsche Bank, BLS

Source: Deutsche Bank, BLS

The percentage of men working has been declining since 1970 but the rise in
women had previously offset this. The percentage of women working in the G7
rose from 40% to 50% from 1970 to 2000 but has flat-lined since. So since
2000, the percentage of the G7 working has actually edged lower.
Unfortunately we dont have labour participation numbers beyond the G7 and a
handful of other mostly DM countries. So in Figure 30-Figure 32 we look at the
overall working age population split by the same regions we looked at in
compiling our GDP numbers. As a reminder this took the top 50 countries by
economic size and spilt them by region. So the numbers cover approximately
89% of global economic activity. The data starts at 1950 and includes the UNs
medium projections out to 2050. The line in each graph shows the 2015 estimate.
Figure 30: Working Age Population

Figure 31: Working Age Population

(bn)

(bn)

Source: Deutsche Bank, UN Population Database

-5%
1970

1990

2010

2030

2050

Source: Deutsche Bank, UN Population Database

2045

2050

0%
2035

2030

Europe

2025

2010

EM

5%

2015

1990

DM

2005

1970

BRIC

10%

1995

0
1950

G7

1985

World

15%

1975

G7
DM
Eurozone

1965

0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1950

1955

World
BRIC
DM
EM

Figure 32: Working Age Growth

Source: Deutsche Bank, UN Population Database

Figure 30 shows the most populous regions (mostly EM) and the overall world
and Figure 31 the smaller regions (the DM world). Figure 32 combines the
regions and shows the data as a percentage change over each 5 year period.
Page 18

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Its quite clear that the working age population in the DM world saw consistent
strong growth in the 3-4 decades up to the end of the 1980s. The growth rate
then took a step down for the next 15 years and has migrated towards zero
growth over the last 5-10 years. Without significant changes in retirement ages
across the globe, by 2015 the G7 and the DM will be facing up to at least 4
decades of steady declines in the total working age population. The Eurozone
sees a steeper decline in the working age population but the deterioration
starts later, nearer the end of this decade.
Figure 33 and Figure 34 then look at the demographics data in terms of
important ratios. Figure 33 looks at those of working age relative to the total
population and Figure 34 looks at those that are expected to be at their
economic peak (35-54 years) relative to those that in theory they may have to
economically support in the population (under 24 and over 65 year olds).
Figure 33: Working Age / Total Population

Figure 34: Productivity Ratio (35-54yr vs. <24yr & >65yr)


0.7

70%

0.6

65%

0.5
60%
0.4
55%

World

G7

BRIC

DM

EM

Europe

50%
1955 1965 1975 1985 1995 2005 2015 2025 2035 2045
Source: Deutsche Bank, UN Population Database

0.3

World

G7

BRIC

DM

EM

Eurozone
0.2
1950 1960 1970 1980 1990 2000 2010 2020 2030 2040 2050
Source: Deutsche Bank, UN Population Database

In terms of the proportion of working age relative to the total population we


again see the ratio for the G7, DM and the Eurozone increase up to the mid1980s. We then see 20 years of stability, followed by a subsequent decline that
continues beyond 2015. In terms of the 35-54 year olds relative to the under 24
year olds and over 65s, for the G7 and the DM world the ratio peaked and
levelled off in the decade after 2000 and then started to decline post 2010, a
decline that will continue over the next few decades. Its a similar story of
decline in Europe albeit from a higher base.
Interestingly the EM world sees the working age / total population ratio rise
from 1965 to 2015 before steadily declining to 2030 and then seeing this
decline accelerate thereafter. In terms of the 35-54 year olds relative to the
under 24 year olds and over 65s, the EM world sees the ratio increase from
1975 out to around 2030 before steadily declining.
So its clear that the developed world has a problem with demographics and
perhaps the recent weak recovery is being influenced by this. The pre-crisis
growth rates could have been artificially elevated by the debt bubble and its
only now that the weakening demographics of the last couple of decades are
becoming more obvious. The end game is likely to be that we will all be
forced to work much longer which will be eventually a great boost to
economic activity. However forcing people to retire later will be political
suicide and as such will only be implemented gradually and possibly only
fully when funding crises arise.

Deutsche Bank AG/London

Page 19

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

(II) Has growing state involvement slowed growth?


The Wealth of Nations: Interaction of the State and private sector
Throughout history the interaction of the state and markets has been a major
driver of economic outcomes and growth. As we have argued nominal GDP
growth has been at historic lows across most of the developed world and has
fallen sharply across many EM nations. Now we pose the question of whether
the step up in state involvement post-GFC (Global Financial Crisis) may be one
of the causes of this.
The interaction between the state and the private sector has been of unique
economic importance from the very birth of civilisation. Whilst the interactions
today may be more complex, convoluted or concealed this basic fact of
economics hasnt changed. Time and again history has highlighted the role of
the state in promoting or holding back economies. From the Glorious
Revolution in 17th century Britain and the subsequent Industrial Revolution,
through the collapse of the USSR at the close of the 20th century, onto the
divergence in economic fortunes between Asia and Africa as the 20th century
ended and the 21st began and into the pre-crisis growth of the US real estate
bubble and the Chinese economic miracle the evolution of the state and its
interaction with the private sector has played a major role.
So too today. Across the developed and developing world since the GFC the
state has intervened ever more heavily in economies. In the developed world
this has chiefly taken the form of fiscal and especially monetary intervention to
support economies experiencing major trauma. In the developing world the
growth of intervention has been more direct, with the state using direct control
over key industries to influence the broader economy.
Developed Markets: Capitalism on Hold
With the onset of the GFC in late 2008 governments and central banks across
the developed world stepped in to support their economies on an
unprecedented scale. Governments blew their fiscal deficits to levels
previously seen only during world wars (Figure 35) and central banks cut
interest rates to all-time lows (Figure 36) and began unorthodox expansionary
policies such as QE. On top of this Developed World Governments took large
stakes in a number of fragile financial institutions to bail them out as they
flirted with bankruptcy in the midst of the GFC. They also increased the power
and scope of regulation.
Figure 35: US Government Surplus/Deficit (% of GDP)
10

Figure 36: Central Bank Rates at the ZLB


25

5
20

Bank of England Base Lending Rate


USA Federal Funds Rate Market Rate

-5

15

-10

-15

Bank of Japan Discount Rate

2009

Europe Central Bank Deposit Rate

10

-20
-25
-30
1929 1939 1949 1959 1969 1979 1989 1999 2009

Source: Deutsche Bank, FRED

Page 20

5
0
1694

1744

1794

1844

1894

1944

1994

Source: Deutsche Bank, GFD

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

At a corporate level, the net result of all this intervention was a sharp falling off
in bankruptcies in most core countries. The Moodys global default rate after
spiking in 2009 at 5.9% for all rated and 13.2% for high-yield, fell away sharply
through 2010/11 (Figure 37 and Figure 38).
Figure 37: Moodys Global Default Rate, by Rating
18%

Inv Grade

Spec Grade

All rated

16%
14%
12%
10%
8%
6%
4%

2%
0%
1920

1930

1940

1950

1960

1970

1980

1990

2000

2010

Source: Deutsche Bank, Moodys // Note: pre-1980s all rated and speculative grade calculations should be viewed with caution due to
the un-developed nature of the HY market before the 1980s.

Figure 38: Creative Destruction Cut Short?


Default Cycle
Cycle

Start Year

Length (Years)

All Rated Default Rate

Peak Year Trough Year Total Length Start to Peak Peak to Trough

Prior

Peak Trough

Cumulative Default (Cohort)


All

HY

1928-1942

1928

1933

1942

14

0.36%

8.49%

0.46%

32%

50%

1969-1971

1969

1970

1971

0.00%

2.63%

0.29%

3%

10%

1981-1994

1981

1990

1994

13

0.16%

3.75%

0.66%

13%

40%

1996-2005

1996

2001

2005

0.59%

4.33%

0.73%

12%

34%

2007-2011

2007

2009

2011

0.40%

5.93%

0.86%

10%

22%

Source: Deutsche Bank, Moodys

What is clear from Figure 37 is that whilst the initial force of the GFC and Great
Recession did cause default rates to shoot up to levels not seen since the Great
Depression, these elevated levels collapsed unusually swiftly. In the final
column of Figure 38 we have calculated a simple estimate of the cumulative
default pain felt trough-to-trough through each of the past 5 default spikes (see
final column). The chart gives a rough empiric estimate of how for all-rated the
default intensity of the post-2007 default cycle was just 83% of the far less
economically devastating dotcom and telecoms bust of the early 2000s (65%
of HY defaults) and only 31% of the size of that during the Great Depression
(44% of HY defaults).
The huge amount of government intervention during the GFC successfully
helped reduce the number and proportion of companies going bust. But as
Joseph Schumpeter wrote over half a century ago, whilst one half of
capitalisms success is creation, the other is its brother destruction. The
fundamental impulse that sets and keeps the capitalist engine in motion
comes from the new goods, new methods of production or
transportation, new markets, new forms of industrial organization
that incessantly revolutionizes the economic structure from within,
incessantly destroying the old one, incessantly creating a new one. This
process of Creative Destruction is the essential fact about capitalism. It is
what capitalism consists in and what every capitalist concern has got to live
in. So is there any evidence in a slowdown in creation in the developed
worlds capitalist economies after government intervention cut short the
destruction cycle and put capitalism on hold?
Deutsche Bank AG/London

Page 21

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

One of the most direct ways to get an insight into this is to see whether
productivity growth has fallen in the aftermath of the GFC and crisis response.
Taking a quick look at the US, its clear that since late 2010 productivity
growth in the US has been very low compared to its historic mean and in
recent quarters has been falling even further (see Figure 39 and Figure 40).
This has been the case across much of the rest of the developed world too
(Figure 41).
Figure 39: US Productivity Growth 1947-Present

Figure 40: US Productivity Growth 2000-Present

4.5%

3Y Average YoY Labour Productivity Growth

8%

4.0%

Post-1950 Average

7%

3.5%

6%

3.0%

5%

2.5%

4%

2.0%

3%

1.5%

2%

1.0%

1%

0.5%

0%

0.0%
1950

1960

1970

1980

1990

2000

Source: Deutsche Bank, FRED

2010

-1%
2000

YoY Growth of Labour Productivity

2002

2004

2006

2008

2010

2012

Source: Deutsche Bank, FRED

Figure 41: Other Developed World Output Per Employed Person YoY Growth Rate
4%

EA17

2%
0%
-2%

-4%
-6%
1996 1999 2002 2005 2008 2011

10%

Japan

8%

5%

4%

0%

0%

-5%

-4%

-10%
1981 1986 1991 1996 2001 2006 2011

-8%
1976

UK

1983

1990

1997

2004

2011

Source: Deutsche Bank, Haver

From this it seems reasonable to argue that governments and central banks
interventions during the GFC to help their economies weather the crisis may
well, for better or worse, have put capitalism on hold across much of the
developed world. This may help explain why real growth post-GFC has been so
low compared to other post-recession recoveries.
Emerging Markets: Beijing Consensus
State intervention in the GFC and its aftermath was not limited to the
developed world. Indeed since the GFC there has been a notable turning of
much of the developing world away from the kind of free market policies of
the Washington Consensus which seemed so indomitable in the late 20th and
early 21st century towards a model of State Capitalism. As the Developed
World fell into the global financial crisis much of the Developing World turned
to China, which had continued to power ahead, as a new model for economic
success based on a far greater involvement of the state in the economy. The
result was an increase in both direct and indirect state influence in economies
through control of firms and their business decisions (such as the pricing of
crucial commodities like electricity and credit) and via fiscal and monetary
macro policy.

Page 22

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

First its important to point out that state control over the EM market is high.
Using calculations by our EM equity strategist John-Paul Smith and his team
(Figure 42), the level of state ownership of listed GEM equities is 30%, ranging
from lows of 0% in Chile, Mexico, Egypt and Peru to highs in Poland (86%),
China (78%) and Russia (55%).
Figure 42: Areas of Significant State Control over the Corporate Sector
GEM
country

Total weighting in MSCI


EM top 500 (%)

Weighting of statecontrolled subset in MSCI


EM top 500 (%)

% of listed market under


state control

Poland

1.1

0.9

85.8%

China

16.4

12.8

78.1%

Russia

5.8

3.2

55.0%

Indonesia

2.7

1.4

52.7%

Czech

0.3

0.1

49.5%

Malaysia

2.9

1.3

44.4%

Thailand

2.4

1.1

43.8%

Colombia

1.0

0.4

40.8%

Hungary

0.2

0.1

40.8%

Brazil

11.8

3.6

30.4%

Taiwan

9.2

1.1

12.3%

Philippines

0.8

0.1

10.0%

Turkey

1.7

0.2

9.5%

India

5.5

0.4

8.0%

South Korea

14.2

0.6

3.9%

South Africa

6.8

0.1

0.9%

Chile

1.9

0.0

0.0%

Egypt

0.2

0.0

0.0%

Mexico

5.1

0.0

0.0%

Peru

0.6

0.0

0.0%

Total GEM

90.6

27.4

30.3%

Source: Deutsche Bank The Month in GEM Equities June 2013 Chart Book, MSCI, Thomson Reuters, Company reports

Moreover these numbers may understate the true impact these governments
have over their nations companies. Indeed whilst states direct control over EM
corporate sectors hasnt necessarily increased greatly in recent years, its indirect
influence has. In China for example state-owned banks are directed in where and
at what interest rate to allocate capital. State-owned utilities provide power under
the same kinds of state guidance. The result is that state control over key
industries (which in general are the types of industries states have a greater
hand in anyway) has led to increased indirect control and influence over the entire
spectrum of the nations public and private businesses. Whilst China is taking
steps to liberalise parts of its financial system, these changes will likely be difficult
(on both political and economic fronts) and lengthy to implement.
At a very basic level when the state overrules the market (through, for example,
direct control of businesses) this leads to the mispricing of the goods, services
and resources. This in turn leads to their misallocation across the economy
which ultimately results in inefficient economic outcomes. So for example
when Chinas state-controlled banks are told to lend cheaply to the states
favoured companies, this leads to too cheap capital in these industries and too
much investment, at the expense of companies/industries not favoured by the
government and consumers. The result is inefficient investment and overinvestment to the extent that in China investment as a % of GDP has rocketed
since 2008 from 41% to almost 50% (see Figure 43). Sustained periods of
overinvestment ultimately lead to overproduction and overcapacity, an
economically inefficient outcome.
Deutsche Bank AG/London

Page 23

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Figure 43: China Investment as % GDP


50%

48%
46%
44%
42%
40%
38%
36%
34%
32%

30%
1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012
Source: Deutsche Bank, Haver

So at a macro level, how would we expect these micro inefficiencies to show


up? Well first in slowing productivity growth. As Figure 44 shows this has
indeed been the case across a number of important EM nations. And ultimately
lower labour productivity growth will result in the type of falling real GDP
growth we have already seen is occurring in EM.

Figure 44: Annual Growth of Labour Productivity


15%

Brazil

China

15%

Russia

10%

10%

5%

5%

0%

0%

-5%

-5%

-10%
2000

2002

Source: Deutsche Bank, Haver

2004

2006

2008

2010

2012

Czech

Poland

Russia

-10%
1996 1998 2000 2002 2004 2006 2008 2010 2012
Source: Deutsche Bank, Haver

Going forward the concern is that the type of inefficiencies which come from
sustained state influence and control over the corporate sector can be very
hard to leave behind or grow out of and indeed as the recent series of minicrises events in China, India and a host of other EM nations shows inefficient
resource allocations are always susceptible to periods of panic and fears of
collapse. All of this is not to say that the age of strong EM growth is
necessarily over. However it strikes us as reasonable to assume that the rate of
sustainable growth across EM is now structurally lower then has been the case
for at least the past decade.

Page 24

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

World sustainable growth rate has fallen on all fronts


On top of reasons for lower nominal growth weve discussed elsewhere in this
piece, the elevated level of state involvement in the developed and emerging
worlds corporate sectors (both direct and indirect) post-GFC has likely reduced
the sustainable rate of world real growth. Whilst over the long-run these issues
of developed world capitalism on hold and emerging world state-led
resource misallocation may be wrung out of the system, either through crises
or policy change, the short- and medium-term outlook is likely to be one of
continued sub standard growth.
Now weve looked at two factors working to undermine post-crisis GDP
growth. Next we look more deeply into what growth is and what drives it. This
is split into real and nominal drivers.

Deutsche Bank AG/London

Page 25

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

What Drives Real and Nominal


Growth?
Real GDP Growth, Revolution Ending?
In the previous chapter we looked at what has been a clear global trend
towards slower growth over the past few years. However in the grand scope of
human history we continue to live in a golden age of growth, even if it shines
somewhat less brightly than it used to. Even in its current straits the world
economys real output per person is growing 5 times faster each year than at
the height of the first Industrial Revolution, 13 times faster than the
Renaissance economy of DaVinci and Medici and infinitely faster than the
world economy at the apex of the great Roman, Parthian and Han empires of
100AD (see Figure 46 and Figure 47).
Figure 45: Epochal Economic Events
Point

Event

Point

Event

The Black Death

1st Industrial Revolution Ends

The Beginning of the Renaissance

2nd Industrial Revolution Begins

America Discovered

2nd Industrial Revolution Ends

The Beginning of the Enlightenment

10

Green Agricultural Revolution Begins

British Agricultural Revolution Begins

11

3rd Industrial Revolution Begins

1st Industrial Revolution Begins

12

3rd Industrial Revolution Ends

Source: Deutsche Bank

Figure 46: Annual Global Real GDP Growth


6%
5%

11

4%
9

3%

12

10

2%
1%

0%

6
3 4

-1000000
-25000
-8000
-4000
-2000
-1000
-500
-200
14
350
500
700
900
1100
1250
1340
1500
1650
1750
1850
1900
1925
1940
1955
1965
1975
1985
1995
2011

-1%

Year
Source: Deutsche Bank, Delong, World Bank

Figure 47: Annual Global Real GDP per Capita Growth


5%
4%
3%

11

2%

1%

1 2

0%

12

10

5
3 4

-1000000
-25000
-8000
-4000
-2000
-1000
-500
-200
14
350
500
700
900
1100
1250
1340
1500
1650
1750
1850
1900
1925
1940
1955
1965
1975
1985
1995
2011

-1%

Year
Source: Deutsche Bank, Delong, World Bank

Page 26

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Do we take growth for granted today after 250 years of continuous growth? The
reality as Figure 46 and Figure 47 show is that this has been a unique period in
human history. In the centuries before 1750 there are a handful of examples of
economies where growth flourished and then collapsed. Examples include the
great North Italian City-States in the first half of the 2nd millennium, Portugal and
Spain in the 16th century and Holland in the 17th. All grew but then faltered.
The uniqueness of post-1750 growth raises an important question will
continuous growth continue? Do we have a divine right to growth? Or have we
simply gone through a Golden Age which is fading as it did in Middle Ages
Venice, 16th century Lisbon and 17th century Amsterdam?
This question is extremely contentious given that it challenges one of the
fundamental foundations of modern economics and society. In light of this we
will first overview the academic literature on what drives long-run economic
growth before diving into Robert Gordons controversial 2012 thesis, Is US
economic growth over?. We will then critique this question by bringing
economic history and long-term data to bear on the debate and journey back
to the very roots of economic growth.
What determines long-run real economic growth?
Robert Solows Standard Growth Model states that long-run per capita
economic growth it is determined by the efficiency of labour and capital
intensity of the economy.
The efficiency of labour refers to the level of a nations technology and how it
is deployed to increase the output each worker can produce for a given level of
capital. Capital intensity refers to how much capital (i.e. machines, buildings,
infrastructure etc) has been set aside for use to increase the output of workers
for a given level of technology.
Therefore at a very basic level long-run real economic growth is driven by
improvements in technology, improvements in economic organization to
deploy that technology and increases in capital intensity. Importantly one of
Solows (1957) core findings was that technological improvement was the
dominating factor in economic growth. Specifically his study of US growth
from 1909-1949 found that 87.5% of the increase in output per hour worked
was attributable to technological progress and only 12.5% to capital increases.
Later work (for example Galor (2005)) added to the model by arguing for the
importance of the growth of human capital, though its importance is largely
derived from its role in the creation and application of technology.
A host of models have expanded, bolted-on, modified, augmented,
embellished and magnified this basic thesis. Nevertheless the models central
conclusions have remained constant long-run economic growth is chiefly
determined by applied advancements in technology. So maybe instead of
asking whether world growth will continue at the same pace as it has for two
and a half centuries previously, the better question is to ask whether
technology will continue to advance at a similar pace?
Gordon and the End of Growth
Robert Gordon published his controversial paper on this subject in 2012 and
his conclusion was that the easy growth era could well be over. His paper
comes in two parts the first (and most significant) part argues that innovation
(i.e. technological advancement) is faltering in the US and the second part that
six growth headwinds (bad demographics, growing inequality, globalization
driven factor price equalisation, worsening education standards, environmental
regulation and high debt) will suppress US growth even further.
Deutsche Bank AG/London

Page 27

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Focusing on the first part of his thesis, he argues that modern day growth has
been driven by three industrial revolutions, each of which has been
implemented faster than its predecessor. The first industrial revolution ran
from 1750-1830 and involved the technological advancements of steam
engines, cotton spinning and railroads. The second industrial revolution ran
from 1870-1900 and included the invention of electricity, the internal
combustion engine and running water with indoor plumbing. Both of these
revolutions took around 100 years for their full effects to be felt (for example
the invention of air conditioning and home appliances from 1950-70 were byproducts of the second industrial revolution). The third industrial revolution
was, Gordon argues, the most short-lived and involved the computer and
internet revolution, which began around 1960 and peaked in the dotcom era of
the late 1990s. Since then technological advancements have mainly been
focused on communication and entertainment devices which in Gordons view
do not fundamentally change the amount each worker is able to produce.
To sum up his argument innovation isnt what it used to be and thats going
to be reflected in lower growth via lower labour productivity growth. Data from
the G7 does show a steady decline in productivity from the 1960s to the 2010s
(see Figure 48).
Figure 48: Average Annual G7 Productivity Growth by Decade
5.0%

1960s

1970s

1980s

1990s

2000s

2010s

4.5%
4.0%
3.5%
3.0%

2.5%
2.0%
1.5%

1.0%

0%

0.5%
0.0%
US (1960)

UK (1970)

Italy (1980)

Germany
(1970)

France
(1970)

Canada
(1980)

Japan (1970)

Note: Start date for each countrys data in brackets.


Source: Deutsche Bank, Haver

Looking back further, real GDP growth in the US through the latter half of the
2000s and the 2010s has been at the lowest levels since the cyclically scarred
decades of the Great Depression and the First World War (see Figure 49).
It is hard to argue with the evidence that real GDP/capita growth has been
exceptionally low so far in the 21st century. Indeed since 1800 the USAs
average annual per capita growth rate has only been lower than the 2000s
0.8% in the decades of the 1800s, 1860s and 1930s. Whilst this alone is
insufficient to argue that world growth as we knew it has ended, it is valid to
ask whether it could ever ascertain the higher levels we came to expect in the
20th century. In order to answer this question its important to get to the very
roots of what has driven economic growth since 1750.

Page 28

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Figure 49: US Real GDP per Capita Average Annual Growth Rate by Decade
5%
4%
3%
2%
1%

0%
-1%

2010s

2000s

1990s

1980s

1970s

1960s

1950s

1940s

1930s

1920s

1910s

1900s

1890s

1880s

1870s

1860s

1850s

1840s

1830s

1820s

1810s

1800s

-2%

Source: Deutsche Bank, GFD

The roots of economic growth


As discussed above, long-term economic growth (both in general and
specifically in per capita terms) post-1750 has been driven chiefly by
technological advances and supplemented with capital accumulation. But this
doesnt really uncover the roots of growth, i.e. where it came from in the first
place. As Nobel Prize-winning economic historian Douglas North and coauthor Robert Thomas wrote in 1973, the factors we have listed (innovation,
economies of scale, education, capital accumulation, etc.) are not causes of
growth; they are growth. We have seen that technological advancement
drives economic growth. But what drove continuous technological
advancement post-1750 after being absent for so many centuries?
There have been three broad academic approaches to answering the question
of where continuous technological advancement (and so growth) came from.
First Unified Growth Theory highlights the linkages between population growth,
real wages and human capital. Second is New Institutional Economics which
focuses on the interaction of institutions, markets and technology. Finally
within the Schumpeterian Growth framework is a theory which looks at the
special role of General Purpose Technologies in sustaining growth beyond
the one-off advances seen in earlier ages.
The Unified Growth Theory developed by Galor (2005) has been used by a
number of economists (see DeLong 2002) to argue that the one-off inventions
of the centuries preceding 1750 allowed for steady (but still exceptionally slow)
population growth (Figure 50). Between the year 0 and 1750 the world
population grew by 163% (which works out at an annual rate of 0.06%). As the
population grew so too did the rate of invention (two heads are better than
one) which then increased the level of real wages as technology began to
increase faster than the population could grow, boosting labour productivity
and the returns to acquiring human capital so as to use these new
technologies, one example would be learning to read following the invention
and spread of the printing press after 1450, which in turn raise output per
worker further.

Figure 50: World Population (bn)


8
6

4
2
0
0

400

800

1200

1600

2000

Source: Deutsche Bank, UN

As people grew richer economies then went through a demographic transition,


as first death rates and later birth rates fell, further boosting output per worker
and settling us into modern economic growth. Thus Galor and Weil (1998)
argued that it wasnt a single shock which created continuous economic
growth but rather long periods of gradual change which in the mid-18th began
to translate into sustained economic growth. Britnell and Campbell (1995) for
example pointed to a long-period of commercialization (which can be viewed
through the lens of urbanization levels) of the British economy in the centuries
before the Industrial Revolution.
Deutsche Bank AG/London

Page 29

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Whilst New Institutional Economics does not directly disagree with the above
theory, it does argue that the Industrial Revolution beginning in 1750 was a
unique and pivotal event in bringing about continuing economic growth, and
that this revolution was the result of significant developments in the
institutions of Northern Europe and Great Britain in particular. This school of
development, laid out in North (1990), Acemoglu, Johnson and Robinson
(2005) and Grief (2006), defines institutions as the rules of the game which
determine the incentives and constraints facing economic agents. A particular
stress is put on the development of property rights, without which agents have
little incentive to invent or innovate for fear that the all-powerful Monarch will
simply steal any resulting profits.
Within this school North and Weingast (1989) argued that the stand-out event
in the build up to the Industrial Revolution was Britains Glorious Revolution of
1688 which created the necessary institutional environment for what was to
come. It irrecoverably proved the power of Parliament and the Law (including
private ownership of property) over the Monarch. It was this institutional
environment which would foster the Industrial Revolution and sustain it well
beyond the life of previous technological advances. There is one final arrow in
the institutional quiver. Institutional developments such as Francis Bacons
Enlightenment scientific empirical method and the development of financial
markets to allow for credit and investment meant that the 18th and 19th
centuries saw not only a revolution in industry but also a revolution in the
process of invention. The initial breakthroughs of the first industrial revolution
were continually adapted and improved which eventually allowed for the
second and third industrial revolutions to continue the transformation of the
world economy.
Finally the General Purpose Technologies theory from the Schumpeterian
framework doesnt attempt to explain where the first and most important
innovation of the Industrial Revolution, the steam engine, came through but
rather why this innovation was special in allowing for continuous
technological progress afterwards. Previous inventions had always run into
diminishing returns and their growth effects soon petered out. The steam
engine on the other hand was an invention which (a) could be improved
dramatically from its early forms, (b) could be applied across an entire range of
industries from textiles to transport and (c) freed the world from the age-old
energy and power constraint of burning wood and straining oxen (Wrigley
(2004)). These three factors marked the steam engine out from the
breakthroughs of previous ages. Some mixture of the three of these also mark
out the other great General Purpose Technologies of world history electricity, the internal combustion engine and computing to name a key few.
So why does all this matter today?
Chiefly because it provides a guide to what drives long-run economic growth
and so provides us with a guide to analyse growth going forward. Economic
theory tells us that technological advance is the key to long-run growth.
Economic history tells us that there are 3 fundamental determinants of
technological progress (1) Population size and human capital, (2) Institutional
Developments and (3) the invention in particular of General Purpose
Technology.
Looking into the future, on the one hand there are more people on the planet
than ever before who are on aggregate investing more than ever before in
human capital. One worry on this front might be the DM and some EM
demographic trends we have already highlighted ageing and (over longer
horizons) shrinking populations. The picture for institutional development and
new ground-breaking technology is more clouded. Do we understand the
problems inherent in our current institutional set-ups? And if so do we have
Page 30

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

the political will to resolve them? Also is there a technology with the
revolutionary power of the steam engine or the internet on the horizon? The
answers to these questions are beyond the scope of this report but it is
interesting reading Gordons assertion that the invention of the flushing toilet
was infinitely more important than the advance of social media inventions
such as Facebook. In his influential paper he offers the hypothetical choice
between option A and option B. With option A you are allowed to keep 2002
electronic technology, including your Windows 98 laptop accessing Amazon,
and you can keep running water and indoor toilets; but you cant use anything
invented since 2002. Option B is that you get everything invented in the past
decade right up to Facebook, Twitter, and the iPad, but you have to give up
running water and indoor toilets. You have to haul the water into your dwelling
and carry out the waste. Even at 3am on a rainy night, your only toilet option is
a wet and perhaps muddy walk to the outhouse. Which option do you
choose? Its certainly food for thought and perhaps puts recent technological
advances into some kind of growth perspective.
Furthermore it is important to note that from the long-run growth perspective
laid out above, the prospects for the world economy as a whole may be brighter
than for the worlds leading economies given that huge parts of the world are
still yet to have developed the institutions which allow for human and physical
capital accumulation, or for the use of all of the modern worlds technological
breakthroughs. Many of these countries also have superior demographics to the
developed world. The rise of China in the past few decades in the wake of Deng
Xiaopings Reform and Opening Up reforms and the dynamism this has
brought to the world economy is evidence of the possible revolutionary power of
institutional development. If everyone on the planet had the same PPP USD
income (the current world average of $12,700) as the average American
($50,700), the world economy would be 4 times the size.
So we would be very hesitant to give up on the concept of global growth but it
is worth remembering that measurable economic growth has occurred for only
a very small percentage of the planets history.

Nominal GDP Growth, Impossible without Money


Creation/Innovation
MV = PY Nominal GDP growth requires MV growth
As weve discussed throughout this piece, real and nominal GDP growth are
not one and the same. In the previous chapter we discussed why and how
technological development drives real economic growth. In this chapter we
will look at what else we need to consider when analysing the basis for
nominal GDP growth. And rather interestingly, the answer is monetary
development.
We get to this answer through a basic economic identity, MV=PY. That is, the
money supply (M) multiplied by the velocity of money (V, a measure of the rate
at which money in circulation is spent on goods and services) equals the price
level (P) multiplied by the level of real GDP (Y). PY is also known as nominal
GDP. Intuitively this makes sense the monetary value of the economy
(nominal GDP) can only be worth the total amount of money in the economy
multiplied by how many times each unit of money is used.
Real GDP growth is possible without a corresponding growth in MV. Without
monetary growth/innovation, any increases in real GDP would simply be offset
in nominal terms by productivity-driven deflation the value of money
relative to goods would increase as technological advances increased the
supply of goods whilst the money supply remained constant (i.e. money would
Deutsche Bank AG/London

Page 31

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

become relatively more scarce in comparison to goods). Indeed this pattern of


real GDP growth and price level deflation was common before the twentieth
century, a period dominated by the hard metal currency system of the
Classical Gold Standard.
The US provides us with compelling evidence that without monetary
expansion, nominal GDP growth can and will lag real GDP growth. Figure 51
shows us that prior to the 20th century the US saw very strong real GDP but
lower nominal GDP. Indeed between 1800-1900, real GDP increased by 6368%
(4.26% p.a.). However over the same period nominal GDP growth only
increased by 3826% (3.74% p.a.). Whilst our long-term CPI index rose 12%
over this hundred year period, producer prices fell 40% showing the
deflationary aspect of 19th century real growth. So there is evidence to suggest
that in a world without an organized central bank and without the ability to
create money as can occur today, real GDP growth has deflationary tendencies.
Clearly in the US there was likely some monetary expansion in the 19th
century (through new gold supply and more importantly increasing credit in
the form of banknote issuance) but nowhere near enough to keep nominal
activity growth above real.
Figure 51: 5 Year Moving Average US Real and Nominal

Figure 52: US YoY CPI and 5 Year Moving Average

GDP Growth
25%

Real GDP

25%

Nominal GDP

20%

20%

15%

15%

10%

10%

5%

5%

0%

0%

-5%

-5%

-10%
-15%
1789

Source: Deutsche Bank, GFD

1869

5yr MA

-10%

1900
1829

CPI

1909

1949

1989

-15%
1789

1900
1829

1869

1909

1949

1989

Source: Deutsche Bank, GFD

In the UK the comparison between nominal and real GDP growth in the 19th
century are slightly different, however the big picture remains the same. From
1830-1900 (our data doesnt start until 1830 for the UK) real GDP growth was
314% whilst nominal GDP growth was only slightly higher at 369% , an annual
difference of just 0.2%, extremely small in comparison to the 20th century
nominal v real divergence (see Figure 53). Whilst over this entire 70 year period
RPI prices fell -7% and wholesale prices fell -28%. This restates the point that
prior to a world of free money creation/innovation, real growth exhibited
deflationary pressures.

Page 32

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Figure 53: 5 Year Moving Average UK Real and Nominal

Figure 54: UK YoY RPI and 5 Year Moving Average

GDP Growth
25%

Real GDP

25%

Nominal GDP

20%

20%

15%

15%

10%

10%

5%

5%

0%

0%

-5%

-5%

-10%

-10%

1900
-15%
1831 1851 1871 1891 1911 1931 1951 1971 1991 2011

CPI

-15%
1831

Source: Deutsche Bank, GFD

5yr MA

1900

1861

1891

1921

1951

1981

2011

Source: Deutsche Bank, GFD

Indeed looking across the globe, it is clear that even leaving aside swings in
velocity, money supply growth has a very close relationship with NGDP growth
(Figure 55). As can be seen from the chart below the relationship between M2
growth and NGDP growth over time has held up extremely well across
countries over the past century. Figure 55 looks at the average annual NGDP
and M2 Growth rate for 18 different nations using all available data whilst the
smaller chart within Figure 55 excludes the extreme point of Argentina. The
result is almost a 1-for-1 relationship and in general we have a very confident
estimate that a 10% annual growth rate of M2 would be associated with
around a 9.9% nominal GDP growth rate.
Data

Figure 55: By Nation: NGDP Annual Growth (y) v M2 Annual Growth (x)
160%

M2 Growth NGDP Growth


(Ann, av)
(Ann, av)

y = 1.01x - 0.02
R = 0.99

Argentina (1941-2009)

145%

146%

Turkey (1987-2011)

56%

52%

120%

Russia (1997-2012)

34%

25%

100%

Korea (1961-2012)

25%

18%

China (2010-2012)

17%

16%

Hong Kong (1970-2012)

17%

12%

Spain (1963-1998)

14%

14%

Greece (1981-2012)

14%

13%

France (1921-1998)

12%

13%

Italy (1949-2012)

11%

10%

UK (1987-2012)

11%

6%

Japan (1958-2012)

10%

7%

Germany (1970-1998)

8%

7%

Sweden (1962-2012)

8%

8%

Norway (1914-2012)

7%

8%

USA (1949-2012)

7%

7%

New Zealand (1989-2012)

7%

5%

Denmark (1993-2012)

5%

4%

140%

80%

60%

60%

40%

40%

20%

y = 0.87x - 0.00
R = 0.96

0%

20%

0%

20%

40%

60%

0%
0%

20%

40%

60%

80%

100%

120%

140%

160%

Source: Deutsche Bank, GFD

This goes to show the potential power central banks have to influence nominal
GDP growth through money supply growth. Lets break this chart down into a
few of the major component nations with longer time series to see how well
the NGDP v M2 growth relationship holds over time (Figure 56-Figure 64):

Deutsche Bank AG/London

Source: Deutsche Bank, GFD

Page 33

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Figure 56: US M2 v NGDP Growth


20%

M2

NGDP

15%

Figure 57: Japan M2 v NGDP

Figure 58: Norway M2 v NGDP

Growth

Growth

30%

M2

60%

NGDP

20%

10%

0%

0%

0%
-5%
1949 1959 1969 1979 1989 1999 2009
Source: Deutsche Bank, GFD

-20%

-10%
1958 1968 1978 1988 1998 2008
Source: Deutsche Bank, GFD

Figure 59: France M2 v NGDP

-40%
1914

1934

1954

1974

1994

Source: Deutsche Bank, GFD

Figure 60: Italy M2 v NGDP Growth

Growth
40%

NGDP

20%

10%

5%

M2

40%

Figure 61: Germany M2 v NGDP


Growth

M2

NGDP

40%

M2

30%

NGDP

20%

15%

10%

-20%

10%

0%

-40%
1921 1936 1951 1966 1981 1996
Source: Deutsche Bank, GFD

5%

-10%
1949 1959 1969 1979 1989 1999 2009
Source: Deutsche Bank, GFD

Figure 62: Korea M2 v NGDP Growth

NGDP

20%

20%

0%

M2

25%

30%

0%
1970 1975 1980 1985 1990 1995
Source: Deutsche Bank, GFD

Figure 63: HK M2 v NGDP Growth

Figure 64: Argentina M2 v NGDP


Growth

80%

M2

NGDP

60%
40%
20%

0%
-20%
1961 1971 1981 1991 2001 2011
Source: Deutsche Bank, GFD

100%

M2

80%

NGDP

750%

60%

550%

40%

350%

20%

150%

0%
-20%
1970

M2

950%

NGDP

1980

1990

2000

2010

Source: Deutsche Bank, GFD

-50%
1941

1956

1971

1986

2001

Source: Deutsche Bank, GFD

It seems fair to conclude that with the exception of data issues surrounding
major wars, M2 and NGDP generally have closely interrelated rates of growth
across a whole spectrum of countries and time periods. For nominal GDP
growth to occur, the monetary system needs to grow and innovate to
accommodate real GDP progress. So how can the monetary system (made up
of M and V) grow? Given that M and V have in fact got very separate drivers
we analyse each independently here.
How can M grow?
The money supply (M) is equal to the monetary base (currency plus bank
reserves) multiplied by the money multiplier (the ratio of currency + deposits
to the monetary base). This multiplier reflects the rate at which banks increase
the central banks monetary base in the economy. Therefore money supply
growth can be driven either by an increase in the monetary base via the central
bank creating money or via the banking system increasing the amount of
money available by greater deposit creating (i.e. lending) activity for a given
unit of central bank money. One way in which nominal GDP growth can be
Page 34

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

achieved is through central bank activity or higher lending in the banking


system. Looking at US monthly data from 1959 to 2013 in Figure 65 it can be
seen that historically the banking sector worked to increase the central bank's
balance sheet (monetary base) by a multiple of around 9, however today this
ratio has fallen below 4.
Figure 65: The Components of US Money Supply Growth
12,000

Monetary Base ($bn, LHS)

M2 Money Supply ($bn, LHS)

Money Multiplier (RHS)

10,000

14
12
10

8,000

8
6,000
6
4,000

2,000
0
1959

2
0
1964

1969

1974

1979

1984

1989

1994

1999

2004

2009

Source: Deutsche Bank, FRED

Through history there have been three important determinants of money supply
growth which have worked through these two channels of monetary base and
money multiplier growth. The first is the role of exchange rate policy in
determining changes in the monetary base. The second and third are financial
innovation and financial instability and its effects on the money multiplier.
To quickly cover these three points. First, if a nation fixes its exchange rate (say
to Gold) then the central bank has extremely limited ability to increase or
decrease the monetary base as its monetary policy is determined solely by the
needs to fix the exchange rate in world FX markets. Second, financial innovation
is a major positive driver of the money multiplier as it determines, among other
things, the amount of leverage the banking system runs. Finally financial
instability plays a major negative role in the money multiplier as periods of
financial stress (for example bank runs) can significantly reduce and reverse the
amount of deposit creation and lending done by the banking system.
How can V grow?
In economic terminology velocity (V) is the inverse of the desire to hold money
(money being cash and cash-like assets). However fundamentally what
velocity reflects is the rate at which money in circulation is spent on goods and
services. It is a window on the animal spirits of the economys households
and businesses.
As such velocity growth can be extremely volatile often swinging away for
lengthy periods of time from its long-term mean, despite the fact that it does
display mean reversion (see Figure 66) within a given financial system or era.
Thus the question to be asked is not what causes velocity to grow, but rather
what causes it to diverge from its long term mean. The answer is the variability
of animal spirits.

Deutsche Bank AG/London

Page 35

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Figure 66: History of US Velocity


2.3

US M2 Velocity

Post-1900 Velocity Average

2.1
1.9
1.7
1.5
1.3

2012

2008

2004

2000

1996

1992

1988

1984

1980

1976

1972

1968

1964

1960

1956

1952

1948

1944

1940

1936

1932

1928

1924

1920

1916

1912

1908

1904

1900

1.1

Source: Deutsche Bank, GFD, FRED, BEA, Robert Gordon

Here event analysis may help shed some light. During two of the great
economic events of the 20th century, velocity seems to have reflected and
amplified the crises of the periods. In the German Hyperinflation of June 1921
to January 1924 velocity began to speed up as inflation expectations increased,
driven by what was seen to be politically impossible demands made on the
German budget. During the US Great Depression the huge declines in
consumer and businesses confidence in the face of mass unemployment can
be seen in the extremely and persistently low level of velocity. Velocity also
moved during the recovery from the Great Depression as the US war machine
swung into action in the early 1940s.
Figure 67: Hyperinflation V
1.E+14

German Banknote Velocity (log scale)

1.E+12
1.E+10
1.E+08
1.E+06

1.E+04
1.E+02
1.E+00
1851 1871 1891 1911 1931 1951 1971 1991 2011
Source: Deutsche Bank, GFD

Figure 68: Depression V


1.7

US M2 Velocity

1.6
1.5
1.4
1.3
1.2
1.1
1921

1924

1927

1930

1933

1936

1939

1942

Source: Deutsche Bank, GFD, BEA, Robert Gordon

Moving to the present day, velocity has also been incredibly low since the
onset of the GFC (and interestingly, incredibly high in the high-debt, highleverage boom of the preceding decade). Running a quick comparison its
possible to see evidence that one negative driver of animal spirits (financial
stress) played an important role in determining velocity growth during the GFC
(see Figure 69). Here the measure of financial stress is the St Louis Feds
Financial Stress Index which measures a number of interest rate and yield
spreads series as well as a number of other indicators, each of the variables
measures provides a gauge of financial stress and the average value of the
series is designed to be zero. Any positive increase represents an increase in
stress and vice versa.

Page 36

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Figure 69: Previous period change in financial stress (absolute, x) vs. this
Period Velocity Change (%, y) (2007-Present)
2%

R = 18%

1%
0%
-1%
-2%
-3%

-4%
-5%
-6%

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

Source: Deutsche Bank, FRED

Summary of how MV and thus Nominal GDP can grow


Throughout history nominal GDP growth has been dependent upon monetary
development to support increases in real GDP. Such development has been felt
through the channels of central bank and exchange rate policy, banking
system innovation and stability and economic animal spirits.
As can be seen in Figure 70, there have been important and incremental
advances in the worlds monetary system throughout history. Indeed from
this timeline two things stand out. First that major monetary developments
seem to have coincided with the real economic rise of various powers
through history be it the Qin dynasty of 200s BC China, the North Italian
City States of the Middle Ages, the Spanish Empire of the 16th century, the
commercial hub of 17th century Amsterdam or the Industrial Revolution of
18th century Britain.
The second point of interest is the clear long-term accumulation of financial
sophistication in the build-up to the industrial revolution and the subsequent
initiation of nominal GDP growth. It certainly seems fair to argue that at the
dawn of the Industrial Revolution there were the financial building blocks in
place, such as a central bank, bond market and fractional reserve banking for
MV to expand and so allow nominal GDP growth. Nevertheless this financial
architecture was not put to use for long periods of time (perhaps mainly due to
a gold standard type monetary policy) meaning real growth generally
outstripped nominal growth.

Deutsche Bank AG/London

Page 37

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Figure 70: Financial Innovations and Monetary Developments Through History


Year

Monetary Development

600 BC

Early evidence of coinage in Ephesus (Modern Turkey)

221 BC

China introduces standardised coinage

1255

First historical mention of bills of exchange in Sienna and Florence - early form of credit

1397

Founding of the Medici Bank - further developed double entry book-keeping and dealt in foreign exchange via bills of exchange

1519-1521

Spain conquers the precious metal rich Aztec Empire

1602

The joint-stock Dutch East India Company formed, shares were tradable

1609

Amsterdam Exchange Bank set up - created system of cheques and direct debits, had around a 100% ratio of deposits to reserves

1668

Swedish Riksbank set up - beginning of credit creation and fractional reserve banking with a >100% lending to reserve ratio

1694

Bank of England set up - purpose was to help the British government with war finance, had partial monopoly on bank note
creation, first British government bonds issued

1744

First insurance fund, The Scottish Ministers Widows Fund, formed

1858

Join-stock banking restrictions lifted in UK

1870s-1913

Period of Classical Gold Standard

1913

Federal Reserve System set up in the US

1919-1926

Floating exchange rate system with moves back towards the Gold Standard

1925-1931

Gold Standard restored in a altered form of the Gold Exchange Standard

1939

Introduction of deposit insurance in the US

1945-1971

Bretton Woods System - USD pegged to gold, rest of world tied to USD

1970

US Department of Housing and Development creates first modern residential mortgage-backed security

1971

US stops conversion to Gold

1973

Worlds major currencies begin to float against one another

1973

Black-Scholes model published

1978

401(k) plan begins

1982

First stock index futures introduced

1987

First CDO issued

1994

First modern credit default swap developed in the wake of the Exxon Valdez tanker spill

1999

Eurozone established

1999

Repeal of 1933 Glass-Steagall Act which prohibited commercial banks from participating in investment banking activities

2006

ABX sub-prime mortgage backed credit derivate index on home equity loans as assets launched

2010

Dodd-Frank Act passed in the US, codifying the new regulatory framework

2010-11

New Basel III Regulatory Standard Agreed

Source: Deutsche Bank

MV and Nominal GDP growth going forward


For nominal GDP growth to expand at its historical long-term average rate, the
monetary system will again have to develop and expand to allow for this.
Persistent growth of nominal GDP today, as always through history, will
require higher M and/or higher V. Historically sustained growth of M and V has
required a healthy, stable and innovative financial system as well as a certain
level of consumer and business confidence. Whether these building blocks will
fall into place is another question for todays world economy. We have
increased regulation and capital requirements for banks and an uncertain
economic outlook which seems to have played a part in reining in animal
spirits whilst demographics may continue to chip away at growth.
Following on from examining what drives nominal GDP growth through time,
we now look at recent global central bank activity in the context of the last 5
years of growth.

Page 38

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Putting Recent Central Bank Action in


Context
In the last section we discussed how real growth without MV growth should
lead to downward pressure on prices at the expense of nominal GDP growth.
This indicates that monetary manipulation is needed to create higher nominal
activity. In attempting to place this theory in the context of current activity, its
fair to say that since the financial crisis, global central banks have generally
been deemed to have been highly accommodative, especially in terms of
balance sheet expansion. So why havent we seen much higher nominal GDP
as a result? To help answer this, Figure 71 looks at the growth of the balance
sheets of six of the key global central banks, indexed at 100 in May 2006.
Figure 72 then looks at the aggregated dollar value of these six central banks
balance sheets.
Figure 71: Big 6 Central Bank Balance Sheets Rebased to

Figure 72: Total of Big 6 Central Bank Balance Sheets

100 May06

($tn)

700

Fed

ECB

BoE

16

600

SNB

BoJ

PBOC

14

500

12

400

10

300

200

100
0
Dec 98

Dec 01

Dec 04

Dec 07

Dec 10

Source: Deutsche Bank, Bloomberg Finance LP

4
May 06

May 08

May 10

May 12

Source: Deutsche Bank, Bloomberg Finance LP

On an aggregated basis the dollar value of these six central banks balance
sheets has doubled to over $14.5tn since the Lehman default. In trying to put
this in perspective the left hand chart of Figure 73 adds on the flow of this
balance sheet size to the annual nominal GDP of these countries (all converted
to USD) and then on the right hand chart looks at this on a YoY growth basis.
Figure 73: Nominal GDP + Central Bank Flows of Six Key Global Central Bank Countries Levels ($tn, left) YoY (right)
50

Nom GDP (Annual)

CB Balance Sheet

Nom GDP

Nom GDP + CB BS

15%

45

10%

40

5%

35

0%

30

-5%

25
20
Mar 02

20%

-10%

Mar 04

Mar 06

Mar 08

Mar 10

Mar 12

-15%
Dec 00

Dec 03

Dec 06

Dec 09

Dec 12

Source: Deutsche Bank, Bloomberg Finance LP

Deutsche Bank AG/London

Page 39

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Although central banks have generally been seen to have been aggressive over
the last 5 years, the sizes of their interventions are not substantial versus the
annual size of their respective economies. Figure 74-Figure 77 breaks down
this data by these six central banks/economies individually. This continues to
be all in dollar terms and the percentage changes are YoY.
Figure 74: US (left), Eurozone (middle) and UK (right) Nominal GDP Plus Central Bank Flows Levels ($tn)
18

Nom GDP (Annual)

15

Nom GDP (Annual)

3.0

16

CB Balance Sheet

13

CB Balance Sheet

2.5

Nom GDP (Annual)

CB Balance Sheet

14

11

12

10

1.5

5
Sep 99 Sep 02 Sep 05 Sep 08 Sep 11

1.0
Sep 99 Sep 02 Sep 05 Sep 08 Sep 11

8
Dec 98 Dec 01 Dec 04 Dec 07 Dec 10

2.0

Source: Deutsche Bank, Bloomberg Finance LP

Figure 75: US (left), Eurozone (middle) and UK (right) Nominal GDP Plus Central Bank Flows YoY Growth
30%

10%

20%

5%

10%

0%
-5%

Nom GDP

Nom GDP + CB BS
-10%
Sep 00 Sep 03 Sep 06 Sep 09 Sep 12

0%
Nom GDP

-10%

Nom GDP + CB BS
-20%
Sep 00 Sep 03 Sep 06 Sep 09 Sep 12

30%
20%
10%
0%
-10%
-20%
Nom GDP
-30%
Nom GDP + CB BS
-40%
Sep 00 Sep 03 Sep 06 Sep 09 Sep 12

Source: Deutsche Bank, Bloomberg Finance LP

Figure 76: Japan (left), China (middle) and Switzerland (right) Nominal GDP Plus Central Bank Flows Levels ($tn)
7.0

Nom GDP (Annual)

11

6.0

CB Balance Sheet

5.0

Nom GDP (Annual)


CB Balance Sheet

7
5

4.0

3.0
Sep 99 Sep 02 Sep 05 Sep 08 Sep 11

1
Mar 02

Mar 05

Mar 08

Mar 11

0.9
Nom GDP (Annual)
0.8
CB Balance Sheet
0.7
0.6
0.5
0.4
0.3
0.2
Sep 99 Sep 02 Sep 05 Sep 08 Sep 11

Source: Deutsche Bank, Bloomberg Finance LP

Figure 77: Japan (left), China (middle) and Switzerland (right) Nominal GDP Plus Central Bank Flows YoY Growth
Nom GDP
40%
Nom GDP + CB BS
30%
20%
10%
0%
-10%
-20%
-30%
Sep 00 Sep 03 Sep 06 Sep 09 Sep 12

50%

Nom GDP

40%

Nom GDP + CB BS

30%

20%
10%
0%

-10%
Mar 03

Mar 06

Mar 09

Mar 12

Nom GDP
50%
Nom GDP + CB BS
40%
30%
20%
10%
0%
-10%
-20%
Sep 00 Sep 03 Sep 06 Sep 09 Sep 12

Source: Deutsche Bank, Bloomberg Finance LP

Page 40

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Although this exercise is highly simplistic and ignores any multiplier, the
graphs above confirm that balance sheet expansion has not been of a size that
is radical relative to the loss of output seen since the financial crisis began.
We can look at this a slightly different way. Figure 14-Figure 17 showed
how nominal GDP growth has been slowing relative to its long-term trend
since the financial crisis struck in 2007. So, how much nominal output have
we lost since. In Figure 78 we calculate this annually based on two different
realistic growth rates that markets may have thought as possible long-term
targets on the eve of the financial crisis. These are still lower than the longterm trend (included in the table) and not too far off what economists would
feel comfortable predicting today. For the world we use 6% and 5%, for the
G7 and DM we use 5% and 4% and for the Eurozone we use 4% and 3%.
For each region we then aggregate these numbers and work out a
cumulative loss of output relative to expectations from the start of 2008 to
the end of 2013.
Figure 78: Actual Nominal GDP vs. Projected Level for a Given Trend Growth Rate by Region
Nominal GDP Growth Trend

World

1954-2007

1990-2007

1999-2007

7.8%

6.0%

6.6%

Dec 07 Dec 08 Dec 09 Dec 10 Dec 11 Dec 12 Jun 13* Cumulative


Actual

52.7

52.2

53.8

58.6

61.4

64.1

63.2

@ Chosen Growth (7.0%)

52.7

56.4

60.4

64.6

69.1

74.0

76.5

4.2

6.5

6.0

7.7

9.8

13.3

55.9

59.2

62.8

66.6

70.6

72.7

3.6

5.4

4.2

5.2

6.4

9.4

Difference
@ Chosen Growth (6.0%)

52.7

Difference
G7

7.8%

5.0%

4.6%

Actual

30.7

30.6

30.2

31.5

32.3

32.7

32.1

@ Chosen Growth (5.0%)

30.7

32.2

33.8

35.5

37.3

39.2

40.1

1.7

3.6

4.0

5.0

6.5

8.1

30.7

31.9

33.2

34.5

35.9

37.3

38.1

1.3

3.0

3.0

3.6

4.6

6.0

Actual

40.9

40.0

40.3

42.2

43.1

44.1

43.0

@ Chosen Growth (5.0%)

40.9

43.0

45.1

47.4

49.7

52.2

53.5

2.9

4.8

5.2

6.6

8.2

10.5

42.6

44.3

46.0

47.9

49.8

50.8

2.5

4.0

3.9

4.8

5.7

7.8

Difference
@ Chosen Growth (4.0%)
Difference
DM

8.0%

5.3%

5.4%

Difference
@ Chosen Growth (4.0%)

40.9

Difference
Eurozone

9.0%

5.6%

6.2%

Actual

12.9

12.7

12.5

12.0

12.0

12.3

12.1

@ Chosen Growth (4.0%)

12.9

13.5

14.0

14.6

15.1

15.7

16.1

0.8

1.5

2.5

3.2

3.5

3.9

13.3

13.7

14.1

14.6

15.0

15.2

0.6

1.2

2.1

2.6

2.7

3.1

Difference
@ Chosen Growth (3.0%)

12.9

Difference

40.8

29.5

24.8

18.6

33.0

24.7

13.4

10.8

Note: * - June 2013 levels are calculated on a LTM basis. The cumulative total includes just 50% of the June 2013 difference.
Source: Deutsche Bank, GFD

In the 5 years from the end of 2007 to the middle of 2013, the world
economy has potentially lost up to $41tn worth of cumulative
income/activity relative to a reasonable pre-crisis expectation of growth. To
put this in context, our annual global GDP figure at the end of 2012
(covering 89% of the world) was $64.1tn. Its fair to say that the DM makes
up the lions share of this potential loss of output. The DM, G7 and Eurozone
have potentially lost up to $33tn, $25tn and $13tn of cumulative output in
the 5 years up to mid-2013. This is relative to end 2012 GDP of $44.1tn,
$32.7tn and $12.3tn respectively. For LTM ending June 2013 alone, output
for the World, the DM, G7 and Eurozone would now be $13.3tn, $10.5tn,
$8.1tn and $3.9tn higher if annual growth had been 7%, 5%, 5% and 4%
respectively over this six year period.

Deutsche Bank AG/London

Page 41

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

So the $7.5tn expansion since the Lehman default in the balance sheets of
the six central banks discussed earlier should be seen in this context. Figure
79 looks at this in terms of the annual lost potential global output since 2008
against the annual change in the size of the six key global central banks.
Both are in dollar terms.
Figure 79: Global Annual Increase in CB Balance Sheets vs. Annual Nominal
Loss of Output Relative to LT Trend
14

Annual Global Increase in CB Balance Sheets ($tn)

12

Annual Nominal Loss of Output Relative to LT Trend ($tn)

10

8
6
4
2
0
2008

2009

2010

2011

2012

2013 (LTM)

Source: Deutsche Bank, GFD, Bloomberg Finance LP

We can graphically see that the global increase in balance sheet expansion has
been small relative to the loss of output compared to what could be regarded
as a reasonable pre-crisis trend rate of growth. 2008 saw a huge initial
response which could have been why growth bounced back reasonably
strongly after the trough in 2009. However subsequent years have not seen as
large a liquidity injection in dollar terms and recent activity has been light in
terms of the potential lost output.
Clearly this is a simplistic exercise aimed at putting the liquidity numbers in
context. The discussion ignores multipliers (probably quite low still though) and
ignores the potential huge negative impact had central banks been less
aggressive. It also assumes that the old trend was a reasonable assumption
about what was sustainable. The reality is however that growth pre-crisis may
have been artificially too high for a period of time due to excessive leverage
driving activity beyond its natural level.
Talking of leverage, when you put the central bank action discussed above in
context to the debt accumulation in recent years then its worth pointing out
that G7 total debt increased from $79.75 trillion at the end of 2002 to $130
trillion at the end of 2008 and around $142 trillion at the end of 2012. So with
GDP fairly flat in this period, perhaps central bank balance sheet expansion has
not been as large as the raw numbers suggest. Clearly the debt has shifted
more from the private to the public sector and has generally been financed at
lower yields but balance sheets havent grown at a faster pace than overall
debt accumulation over the past 5 years.
Central Bank balance sheet growth small relative to lost banking system
growth
As Figure 80 shows, the size of the European Banking system balance sheet
has been oscillating around 32tn in the 5 years since September 2008. To put
this stagnation in context, in the proceeding 5 years it increased by around
13.5tn, a rate of around 10% growth per year.

Page 42

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Figure 80: Size and Growth of the Eurozone Banking System


40

Eurozone Bank Balance Sheet (tn,LHS)

Annual % Change (RHS)

16%

14%

35

12%
10%

30

8%

25

6%
4%

20

2%
0%

15

-2%

10
Sep 97

-4%
Sep 99

Sep 01

Sep 03

Sep 05

Sep 07

Sep 09

Sep 11

Source: Deutsche Bank, ECB

In the 5 years between September 2008 and now, the ECB has expanded its
balance sheet by less than 1tn. While the numbers are by no means directly
comparable, the point we are trying to make is that one can continue to argue
that the size of central bank activity so far has been dwarfed by the scale of
the financial crisis.
Importantly even this small expansion overstates the true impact of the
banks balance sheet growth as the rate at which this money is being used
in the economy (the money velocity) has fallen sharply (more on this later,
see Figure 97).

QE not enough to offset financial crisis in inflation terms?


So nominal growth has been subdued since the crisis with aggregate debt
levels notably higher. Central bank balance sheet expansion has been sizeable
but not of a level that competes with the loss of output, the drop in velocity, or
the accumulation of debt. This is perhaps one reason why the perceived
explosion in central bank balance sheets has had little impact on inflation
since the initial response to the crisis.
Figure 81: World YoY CPI (Log Scale)
1000%

World

Figure 82: EM YoY CPI (Log Scale)


10000%

EM

Figure 83: BRIC YoY CPI (Log Scale)


10000%

100%

1000%

1000%

10%

100%

100%

1%

10%

10%

0%
1954 1964 1974 1984 1994 2004
Source: Deutsche Bank, GFD

Deutsche Bank AG/London

1%
1955 1965 1975 1985 1995 2005
Source: Deutsche Bank, GFD

BRIC

1%
1955 1965 1975 1985 1995 2005
Source: Deutsche Bank, GFD

Page 43

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Figure 84: DM YoY CPI


20%

DM

Figure 85: G7 YoY CPI


20%

Figure 86: Eurozone YoY CPI

G7

Eurozone

20%

15%

15%

15%

10%

10%

10%

5%

5%

5%

0%

0%

0%

-5%
1955 1965 1975 1985 1995 2005

-5%
1954 1964 1974 1984 1994 2004

-5%
1955 1965 1975 1985 1995 2005

Source: Deutsche Bank, GFD

Source: Deutsche Bank, GFD

Source: Deutsche Bank, GFD

Figure 81 to Figure 86 show that across the globe, inflation has generally been
trending back down since its post-crisis rebound in 2011. While low inflation is
often a cause to celebrate economically, in this environment of low real growth
and high debt it helps show how delicately balanced the global economy is.
Even China is seeing nominal activity at the lower end of its range of its now 3
and a half decade transition into a global superpower (Figure 87). Indeed it
looks set to extend a rare period of sub-10% activity over the next few quarters.

Figure 87: China Nominal GDP


40%

Nominal GDP Growth

30%
20%
10%

On top of this below capacity demand, it could also be argued that the world
has experienced continued overcapacity supply from China. As we argued in
greater depth earlier, state influence over businesses and in particular banks in
China has led to sustained underpricing of capital for key industries. The result
has been massive investment and the creation of major over-capacity across a
host of Chinese industries. The result has been serious producer price deflation
since early 2012 (see Figure 88). This overcapacity has been exported around
the world, in dollar terms Chinas goods trade surplus is again approaching its
2008 all-time high, see Figure 89). The net result is that China has arguably
been exporting deflation around the globe.
Figure 88: China Producer Price Inflation (YoY, %)
15%

0%
1978 1984 1990 1996 2001 2007
Source: Deutsche Bank, GFD

Figure 89: China Balance of Trade in Goods ($m)


400,000

350,000
10%
5%
0%
-5%

300,000
250,000

200,000
150,000
100,000

50,000
0

-10%
1997 1999 2001 2003 2005 2007 2009 2011 2013
Source: Deutsche Bank, Haver

-50,000
1982 1986 1990 1994 1998 2002 2006 2010
Source: Deutsche Bank, Haver

It seems reasonable to argue that below capacity levels of demand in the


worlds economies as well as continued exports of overcapacity supply from
China have suppressed inflationary pressure significantly in the post-crisis
worlds economies, leaving room for money growth to generate noninflationary demand growth. Central bank asset growth in the developed world
has indeed been huge relative to history since the crisis started but perhaps
not relative to the lost output, the loss of velocity (animal spirits), the impaired
financial system and the natural deteriorating impacts of demographics.
Perhaps monetary policy needs to be even larger or perhaps better channelled?

Page 44

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

The Monetary Playbook


Is Nominal GDP Targeting the Answer?
Monetary policy and human history Tales of Change
Monetary policy has evolved through the economic history of the world and
economists and historians have found it at the root of crises and in the green
tips of recovery. John Laws creation of bank notes in the build-up to the
Mississippi Bubble of 1718-1720 has been blamed for laying the groundwork
for the French Revolution later that century. Red-hot money printing by the
Reichsbank has been singled out for causing the German Hyperinflation of
1921-24 and sowing the seeds of WWII. Monetary policy was there when the
US fell into the Great Depression and when it began its long recovery from it:
President Roosevelts Second Fireside Chat on 7 May 1933 detailed how We
do not seek to let them [borrowers] get such a cheap dollar that they will be
able to pay back a great deal less than they borrowed. In other words, we seek
to correct a wrong and not to create another wrong in the opposite direction.
That is why powers are being given to the Administration to provide, if
necessary, for an enlargement of credit, in order to correct the existing wrong.
These powers will be used when, as, and if it may be necessary to accomplish
the purpose.
Roosevelt made this speech to signal a shift in US monetary policy towards a
policy to return the US to the pre-Depression price level. With it the President
of the United States threw away a decade long adherence to gold-standard,
low-inflation money and accepted the need for the US economy and price level
to catch up with its pre-crisis level.
A question which has been posed often in the past few years is whether or not
such a paradigm shift in monetary policy is required today in the wake of the
worst financial crisis and recession since those dark days of the 1930s. As
discussed throughout this piece, nominal activity continues to be extremely
low globally in spite of 4-5 years of zero-interest rate policies (ZIRP) and QE. So
do we need to fundamentally change the way monetary policy is enacted? And
if so, how? We would argue there should be more debate around the idea of a
Nominal GDP Target (NGDPT). This could be the next big theme if nominal
activity remains as stubbornly low as it has been so far post-crisis.
What is NGDPT?
Nominal GDP is the dollar value of everything produced in an economy in a
given year and has two components total output (real GDP) multiplied by the
price level. The current monetary policy framework across much of the world
seeks to keep the price level increasing at (give or take a percentage point) 2%
a year1. The objective is to ensure low and stable inflation which is thought to
be conducive to stable economic growth. It is a clear and simple framework
whilst at the same time allowing for policy flexibility in the face of economic
slowdown as rising unemployment should reduce inflationary pressures and so
give the central bank room to manoeuvre.
During the pre-2007 Great Moderation period, this monetary policy framework
was perceived to be working, at least on the surface. Inflation and
unemployment were low and stable and real GDP expanded at a steady rate.

Different nations central banks have marginally different inflation targets and relative unemployment vs. inflation rate
stresses; however this framework is a broadly fair reflection of the current legislated central bank framework.

Deutsche Bank AG/London

Page 45

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

The result was steady nominal GDP expansion. However with nominal activity
now dramatically lagging behind all long-term trends, especially in the
developed world, do central banks need to have a big rethink?
There has been some debate about possibly targeting the level of NGDP and
perhaps such a policy should get more airtime. Such a policy was first mooted
in the late 1970s and by the late 1980s was offered as a possible successor to
the money targeting of that decade. A NGDPT would embody two major
changes from current policy. First the central bank would act to stabilise
nominal GDP, rather than inflation, at some constantly increasing level. Second
it would target the level of nominal GDP rather than its rate of change.
How would NGDPT work?
The special feature of NGDPT is this second distinction. Currently if a central
bank aiming to hit a 2% annual inflation target were to undershoot and achieve
only a 1% rate then when the next year came around, the central bank would
have to enact monetary policy still with the aim of hitting a 2% inflation rate.
Its 1% miss the previous year is forgotten. With a level target if the central
banks objective is to hit a level of NGDP 2% higher at the end of the year then
at the start, and it achieved only a 1% increase, then in the next year it has to
make up for lost ground and put in place expansionary policies to grow the
nominal economy by an extra 1% on top of the +2% it would have been
expected to hit anyway.

This last and rather extreme figure isnt far away from where a nominal GDP
targeting Fed would have found itself in 1933 (see Figure 91, LHS). If the Fed
had been told to achieve a level of nominal GDP consistent with a 5%-a-year
growth rate (the 1790-1929 average) after 1929 then by 1933, after 3 years
of Depression, the Fed would have had to have generated 135% growth in
1934 to get back on target. As it turned out, the US economy managed to
grow at an average of 13.5% a year over the next 10 years and was back on
target by 1944.

Figure 90: Nominal GDP with (1) 1%


and (2) 2% Per Period Growth Rate

Nominal GDP

This demand to correct for past mistakes can have big implications down the
road. Lets continue with the above example of the central bank who
undershoots by 1%. After 5 years (see Figure 90) the central bank would have
to try to generate 7% nominal growth in the next year. After 10 years it would
need 13% nominal growth. After 100 years the hapless undershooter would
need to almost treble (x2.7 or +170%) the size of the nominal economy.

800
700
600
500
400
300
200
100

170%
1
0 10 20 30 40 50 60 70 80 90 100
Period

Source: Deutsche Bank

Figure 91: US NGDP ($bn) vs. Target through Great Depression and Great Recession
230
210

US Post-1929 NGDP with 17901929 Growth Average

190

US Actual NGDP

19,000
18,000

US Post-2007 NGDP with 19902007 Growth Average

17,000

US Actual NGDP

170

16,000

150

15,000

130

14,000

110

13,000

90

12,000

70

11,000

50
1919 1922 1925 1928 1931 1934 1937 1940 1943

10,000
2000

2002

2004

2006

2008

2010

2012

Source: Deutsche Bank, GFD

Page 46

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Fast forward to the end of 2012 and assuming the central bank was targeting
a level of NGDP consistent with an increase post-2007 at its NGDP 19902007 average growth rate of 4.7% (see Figure 91, RHS) then the central bank
would need to ensure a 2013 growth rate of 18%. Assuming a more spaced
out catch up rate of reducing the gap by 2% a year then the US economy
would be back on track by 2019 (see Figure 92), requiring an average growth
rate of 6.7% a year.
Figure 92: US NGDP Catch-Up ($bn)
26,000

US Post-2007 NGDP with 1990-2007 Growth Average

24,000

US Actual NGDP with NGDPT Catch-up

22,000
20,000
18,000
16,000
14,000

12,000
10,000
2000

2002

2004

2006

2008

2010

2012

2014

2016

2018

Source: Deutsche Bank, GFD

The key difference between a nominal GDP target and an inflation target is that
central banks would, after a period of economic slowdown, be ready to accept
a higher inflation level and/or (ideally) above-trend real GDP growth for a time
to get the economy back on track. Inflation picking up to 3%, 4% or even 5% a
year would no longer be viewed as a failure of the central bank. Indeed it
would likely be a central aim of its policy as it seeks to eliminate the nominal
GDP gap. For this reason adopting a Nominal GDP target would mark a
fundamental change in monetary policy, far beyond what has so far been seen.
Would it be a change for the better or for the worse?
The Pros for NGDPT
Weve already covered one of the proposed advantages of NGDPT over
inflation targeting increased flexibility of monetary policy in the face of
(recently rediscovered) economic volatility. So in an economy which has just
experienced a severe demand slowdown, monetary policy would be allowed to
be far more expansionary then would be possible under a fixed 2% inflation
target. Using the US economy as an example, assuming a targeted increase in
NGDP of 4.7% a year (the 1990-2007 average) then at the end of 2008 the Fed
would have had a end-of-2009 targeted nominal GDP level 10.6% higher than
at the end of 2008 (targeted increase 2008 growth of 4.7% + targeted 2009
growth of 4.7% + 1.2% 2008 actual decline). Given realized real GDP growth of
-0.1% in 2009, this would have given the Fed 10.7% of inflation flexibility
through 2009. The theory is that this inflation space would have given central
banks more swinging room when it comes to cycle fighting policy responses.
On the other side, if inflation had picked up due to a sharp negative supply
shock (say an oil price spike) then the central bank would have more room to
accommodate this non-core price rise as the impact of the supply spike
would fall on both real GDP and inflation, thus the change in nominal GDP
would demand less dramatic inflation-busting action.
The increased flexibility of monetary policy with a NGDPT and in particular
NGDP targetings allowance for a period of higher inflation rates in the face of
economic weakness is actually even more powerful then it may initially appear
Deutsche Bank AG/London

Page 47

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

to be, especially after severe economic shocks which have stuck economies
interest rates at the zero lower bound (ZLB). By allowing for a higher inflation
rate in the face of a weak economy, NGDPT allows for lower real interest rates
(that is the nominal interest rate minus the inflation rate) which are crucial for
investment and other borrowing decisions. As economist Robert Hall noted in
a paper presented at Jackson Hole in August 2013, with nominal interest rates
stuck at the zero lower bound and inflation low and stable the real interest rate
is also constrained at a lower bound equal to minus the inflation rate. In his
calculations this has resulted in a far too high real interest rate. In the United
States today, with a policy rate of about 10 basis points and an inflation rate
around 180 basis points, the safe short real interest rate is minus 170 basis
points, well above the level of around minus 400 basis points that would
generate output demand equal to normal levels of output supply. So in his
analysis, Hall argues that if the Fed were able to allow and generate an
inflation rate of 4.1% (and by committing to such a rate implicitly in a NGDPT
the Fed would find it far easier to achieve it) then the US economy would
return to full capacity (assuming rates held at 0.1%).
A second hypothesized advantage of NGDPT over inflation targeting flows
from the reasoning laid out above. Not only would a NGDPT allow for more
expansionary policy in busts, it would demand contractionary policy in booms.
From the end of 1996 to Q1 2000 US nominal GDP growth averaged 6.1%,
notwithstanding the Asian and LTCM crises. Perhaps a NGDP target closer to
5% would have demanded more aggressive contractionary policy than actually
seen. So there is a case to be made that a NGDP target might help central
banks fight asset bubbles which are generally positively related to a booming
nominal economy.
The third argued for advantage is that a NGDPT would ensure greater nominal
GDP growth stability and stronger real GDP growth via a stronger
communications channel. Christina Romer (a leading advocate of NGDPT and
former Chair of President Obamas Council of Economic Advisers) argues that
a NGDPT would have the same impact on moving NGDP back to target that
Fed Chairman Volckers monetary targets had in getting US inflation under
control in the 1980s. By committing to return nominal GDP back to target
consumers would regain confidence in their economic future, driving up
consumption, and businesses would note that the central bank has promised
to ensure that the markets for their products continue to grow, and so invest
more. There would be a second-round impact as people begin to expect higher
inflation (given the nominal GDP chasm) and so lowering real interest rates
(nominal rates expected inflation rate) and boosting borrowing and
investment.
Scott Sumner, an economics professor and NGDPT blog-supporter-in-chief,
adds a fourth advantage. He argues that an NGDPT is actually simpler than
current monetary policy frameworks (especially in the US) as it combines the
two current objectives of (1) the need to control inflation and (2) the need to
support the real economy into one objective of a nominal GDP level. Simplicity
is crucial for any central bank framework to be understood by the public at
large and so have the type of expectation effects discussed above.
A fifth major advantage already alluded to and perhaps most relevant today is
that a nominal GDP target would help reduce debt burdens. As we have shown
elsewhere in this report, developed economies have either barely reduced or
increased their debt burdens since the onset of the GFC (as a percent of
nominal GDP). A nominal GDP target could potentially have eased this burden
by increased the nominal incomes which are used to pay these nominal debts.

Page 48

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

One final and fundamental advantage of NGDPT is that it should allow for and
promote more radical and direct monetary policy options to be considered.
These policy options are discussed in the next section (Are the Helicopters
Coming? and particularly, Route One Policy Helicopter Money vs. QE).
Despite these proposed advantages, there are many who argue that NGDPT
would be ineffective or downright dangerous.
Disadvantage of NGDPT
One major disadvantage, as was illustrated by our example of the serial
underperforming central bank 173%-required-NGDP-boost discussed above, is
that a NGDPT could allow for dangerously high levels of inflation and an
unanchoring of inflation expectations in a deeply depressed economy. If the
US economys real GDP had grown by its 1800-1929 average of 4.2% each
year from 1934 then the Fed would have needed to have generated 9.3%
inflation a year for 10 years to get the country back to its post-1929 trend
target by 1944. What actually happened was that real US GDP grew at an
average rate of 10.1% a year from 1934-1944 due to the ultra-low starting
point of the Depression economy and growing US war production, meaning
inflation only had to average 3.4% to get the US back on track by 1944. Might
the Fed have over-reacted had they had a specific target? Adding todays
numbers to this concern is informative. As of the end of 2012 (as weve
already discussed above) US NGDP could be argued to be around 13% below
its pre-2007 track. Given average US real GDP growth of 2% a year (20092012) then to get NGDP back on track by 2019 (again using the earlier
example) would require an annual inflation rate of 4.7% a year. Whether this
level of inflation is dangerous in and of itself is debatable, however it seems
fair to argue that once NGDP was back on target in 2019 even if the central
bank announced it would then reduce inflation so as to keep to its target (now
without any catch-up required) there is a danger that inflation might stick at
its 4.7% rate. If this were the case (and assuming a steady 2009-2012 average
2% real GDP expansion) then a decade after the NGDP target had been hit the
central bank would be faced with the task of deflating nominal GDP by 15%, a
decline similar to that seen in 1931 (-16%) and far in excess of the -1.2% seen
in 2008.
Figure 93: NGDP with a sticky high inflation rate post catch-up ($bn)
55,000
50,000
45,000

US Post-2007 NGDP with 1990-2007 Growth Average


"Actual" Nominal GDP Path

40,000
35,000
30,000
25,000
20,000
15,000

10,000
2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020 2022 2024 2026 2028 2030
Source: Deutsche Bank

This again is a rather extreme example and assumes that monetary policy
changes wont shift real GDP growth. Nevertheless these numbers reflect real
concerns. After three decades fighting to be credible inflation tamers, could
central banks really risk losing that credibility?
The second major disadvantage flows straight from the first even a
successful NGDPT in the face of a fluctuating economy will likely demand
Deutsche Bank AG/London

Page 49

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

much greater volatility in inflation than under an inflation targeting system to


offset swings in real GDP. Such volatility in inflation has its own costs and so
the advantages of reducing the volatility of nominal GDP via NGDPT have to be
weighed vs. the costs of higher volatility in inflation. Such costs include great
uncertainty over real interest rates when signing long-term contracts and the
eponymous menu cost problem to name a few. There is also a question as to
whether central banks being obliged to boost NGDP to catch up on lost growth
could lead to asset bubbles.
A third disadvantage is that targeting NGDP is not as simple as targeting the
rate of inflation so is harder to communicate to the public and thus weaker at
controlling inflation expectations.
A fourth argument against might be that it is very difficult to pick the correct
NGDP target. Is it based on a previous trend or is it based on perception of
whats realistic going forward. In a world of declining productivity and weaker
demographics, policymakers might have to be realistic above the potential
growth rate. However who makes such a judgment as to the appropriate rate
could become highly politicised.
The final issue is how nominal GDP targeting would be implemented and how
it would work alongside fiscal objectives? Given that even the worlds
independent central banks receive their mandates on their objectives and
institutional framework from the government, getting a nominal GDP target
monetary policy system up and running would require running the gauntlet of
national politics. Such issues are magnified at least 17-fold when we look at
the ECB and changing Eurozone monetary policy. Given the Eurozones torrid
performance since the global financial crisis, would a nominal GDP target
demand too expansionary a policy for Germanys tastes? Given Germanys
long and successful history of inflation targeting, would they truly be willing to
give this up? Furthermore they may have issue with the second political
economy issue of nominal GDP targeting. Would it encourage governments to
be weak in the face of difficult economic decision-making? Would they try to
rely on central bankers to do the difficult heavy lifting, ignoring much needed
structural and/or budgetary reforms?
These are the major disadvantages of NGDPT. One can also add that monetary
policy has long lags before it affects the economy and expectations can shift
rapidly, ruining careful central bank planning. However such drawbacks are
not unique to NGDPT, indeed the current global policy bias towards inflation
targeting has similar problems. We should also ask whether central banks
alone could truly hit nominal GDP level targets? Would they have the armoury?
This brings us onto our second point of discussion. There are two parts to a
central bank. The first is its policy objective, however it is highly unlikely that
simply announcing a NGDP target and a schedule for hitting it would, alone, be
enough for nominal GDP to actually hit the targeted levels. If central banks
objective framework was to be changed to NGDP targeting then the central
bank would then have to decide how to enact it. Specifically it would have to
introduce policies which would make hitting the new target credible, a
particularly tough task in the current environment given the large post-2007
NGDP gap.
So what policies are left in the monetary policy playbook?

Page 50

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Are the Helicopters Coming?


What monetary policy has done so far
The old unorthodoxy is now orthodoxy. Central banks have sent interest rates
to lows never seen in centuries of existence (see Figure 94). Once they hit
these zero-lower bounds central banks embarked on programmes of
quantitative easing, expanding their balance sheets (see Figure 95) by
multiples of their pre-GFC levels and via purchases of assets previously seen as
beyond the spectrum of viable central bank assets (for example MBS and longterm government bonds). As we write the US and Japanese central banks
continue to expand their balance sheets by enormous quantities, even if the
former seems close to paring some of this back.
So on many historical measures monetary policy is already extremely and
aggressively accommodative. However the worlds economies are not running
at what is deemed to be a sufficiently high nominal rate. And as weve
discussed above, if central banks objective frameworks are changed to allow
them to be more accommodative in the face of continued economic
disappointment, greater policy activism will be demanded of them.
Throughout the modern history of monetary economics one policy has been
put forward as a monetary super drug (or deadly poison depending on your
view). That is helicopter money. It has long been seen as being too powerful
to control and thus beyond the scope of contemplation. However in the past
decade such policy has slowly emerged from the shadow of heterodoxy.
So what is helicopter money? What are the dangers? And could it really be put
into action, possibly as the policy to make central bank nominal GDP targets
credible.

Figure 94: Rates at the Zero Lower


Bound
25
20

15

Bank of Japan Discount Rate


Bank of England Base Lending Rate
USA Federal Funds Rate Market Rate
Europe Central Bank Deposit Rate

10
5
0
1694 1744 1794 1844 1894 1944 1994
Source: Deutsche Bank, FRED, GFD

Figure 95: Balance Sheet Expansions


5
4

ECB ($tn, Aug'13 ER)


BoJ ($tn, Aug'13 ER)
BoE ($tn, Aug'13 ER)
Fed ($tn, Aug'13 ER)

2
1

0
1999 2001 2003 2005 2007 2009 2011 2013
Source: Deutsche Bank, FRED, GFD

Helicopter Money
To explain what helicopter money is we first turn to two heavy-weights of the
monetary policy world Milton Friedman, a giant of 20th century economic
and monetary thinking, and Fed Chairman Ben Bernanke, one of the most
powerful and influential central bankers in world history.
Friedman proffered helicopter money as a once and for all change in the
nominal quantity of money. He gave the policy its helicopter moniker
through a now famous example where he asked what the impact would be if
the government sent out helicopters which dropped a $1000 in bills from the
sky. Given that the drop doesnt change economic agents desires to hold cash,
each agent will try and spend their excess real cash holdings. Given that no
one wants to accept more cash at the current price level (each agent is trying
to reduce their cash holdings), prices are bid up until a new equilibrium is
found. The result is a jump in prices and nominal GDP.
In the early 2000s Ben Bernanke discussed how such a helicopter drop might
be enacted in reality, specifically in reference to the zero-lower-bound
deflation-ridden Lost Decade Japanese economy. Bernanke argued that the
most effective policy for the Japanese economy was helicopter money, or
what he called a money financed tax cut and suggested channels through
which it could increase prices and real GDP. He argued in a 2002 speech
(Deflation: Making Sure It Doesnt Happen Here) that fiscal and monetary
(central bank) authorities should co-ordinate through a broad-based tax cut
accommodated by a programme of open market purchases to alleviate any
tendency for interest rates to increase stimulating consumption and prices.
Even if it didnt help consumption it would boost asset prices as households
rebalanced portfolios.
Deutsche Bank AG/London

Page 51

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Also relevant to today is comments made by Bernanke in 2003 on the likely


effectiveness of helicopter money in Japan when he stresses that with such a
policy, The health of the banking sector is irrelevant to this means of
transmitting the expansionary effect of monetary policy, addressing the
concern of BOJ officials about broken channels of monetary transmission.
The message is clear if done on an adequate scale helicopter money has the
power to raise prices and nominal GDP even in the face of the severe
economic headwinds and dislocations (such as a badly damaged banking
system) seen post-2008. It flows naturally from Friedmans famous statement
that inflation is always and everywhere a monetary phenomenon. If a central
bank really wants inflation-driven nominal GDP growth, it can have it. Indeed
the combination of NGDPT and Helicopter Money appears to be a potent
potential policy package for todays moribund economies.
Route One Policy Helicopter Money vs. QE
A major benefit of a policy of helicopter money over current quantitative easing
policies is its greater potential directness. Quantitative Easing is reliant upon
second-round effects to affect the real economy. When the Fed purchases
government securities it hopes that its purchases will (a) force investors and
financial institutions to rebalance their portfolios towards riskier assets (such
as loans to businesses or corporate bonds) by driving treasury yields lower and
(b) it will raise financial asset prices increasing wealth. It is hoped that once the
Fed has lowered borrowing costs and raised financial asset prices these first
round effects will feed into second round effects of greater consumer spending
and business investment, only then boosting economic activity. From this two
points stand out to us about QE. First its economic effects are secondary to its
financial asset effects. Second it is reliant upon the financial and more
specifically the banking system to act as the transmission mechanism for its
economic impact. If banks, when the Fed purchases government securities off
of them and credits them with reserves, simply hold onto these reserves
instead of lending them out much of the impact of QE will never survive to the
desired second round economic effects.
This is exactly what has happened. As the Fed has expanded its asset
purchase programme, banks have held onto their reserves. In a much talked
about paper presented by economist Robert Hall at the August 2013 Jackson
Hole get together, These countries [US and other advanced economies] have
been in liquidity traps, where monetary policies that normally expand the
economy by enlarging the monetary base are ineffectual. Reserves have
become near-perfect substitutes for government debt, so open-market policies
of funding purchases of debt with reserves have essentially no effect. This
followed from Michael Woodfords 2012 paper, also presented at a Jackson
Hole meet in 2012, that, once the interest-rate lower bound is reached, bank
reserves and other very short-term riskless claims should become essentially
perfect substitutes, so that increases in reserves that come about through
central-bank purchases of riskless short-term assets should have no eect.
What these papers are saying is that at the zero-lower bound (ZLB) QEs main
effect will be to increase banks reserve holdings and so will have little actual
economic effect. All of the impact of QE will be lost in the first-round financial
effects. Figure 96 and Figure 97 show that this has been the case. Fed asset
purchases have increased the monetary base, however most of the increase
has remained stuck in excess reserves. Figure 96 shows that the rate at which
the Feds (and other major central banks) monetary base has been turned into
actual money supply through bank lending activity has collapsed (lower money
multiplier) and Figure 97 shows how any increases in the money supply which
have been achieved have had an incredibly dampened effect on actual nominal
activity (lower money velocity).
Page 52

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Figure 96: Money Multiplier (June 2006=100)


120

EA

Japan

UK

Figure 97: Money Velocity (June 2006=100)


US

100
80

60
40

20
Jun 06 Jun 07 Jun 08 Jun 09 Jun 10 Jun 11 Jun 12 Jun 13
Source: Deutsche Bank, Haver

105

EA

Japan

UK

US

100

95
90
85

80
75
70
65
Jun 06 Jun 07 Jun 08 Jun 09 Jun 10 Jun 11 Jun 12 Jun 13
Source: Deutsche Bank, Haver

From these charts it seems fair to argue that much of the impact of QE has
indeed been lost in the financial and banking system, distorted as it is by postcrisis balance sheet rebuilding, adaptation to new regulations and rates at the
zero lower bound and so never making it to the actual economy.
For helicopter money on the other hand, as weve already highlighted via Ben
Bernankes own words, the health of the banking sector is irrelevant to this
means of transmitting the expansionary effect of monetary policy. The reason
is simple helicopter money bypasses the banking system and puts money
straight into consumers and businesses pockets. Where the first round
effects of quantitative easing hit the financial system and then through the
financial system the second round effects reach consumers and businesses,
helicopter money first hits consumers/businesses and then through them the
financial system.
Helicopter money achieves this direct impact by directly increasing the cash of
consumers and businesses through (say) a money-financed tax cut.
Importantly this money has very high economic power as it is very likely it
will be spent (on consumption or investment) because the central bank has
purchased permanently the debt created to finance the tax cut meaning no
current or future debt liability has been incurred and so higher taxes in the
future shouldnt be expected. As Bernanke stated in 2003, after a helicopter
money policy, essentially, monetary and fiscal policies together have
increased the nominal wealth of the household sector, which will increase
nominal spending and hence prices. All of this is achieved without any
involvement of the banking sector, which is irrelevant.
All the data we have points to the developed worlds financial and banking
system unable and/or unwilling to put their grown central bank reserves to
work in the real economy. All unconventional monetary policy to date has
fallen foul of this fact. Helicopter money wont.
Indeed to our eyes this debate gets to the heart of what central banks
fundamentally can and cannot do, chiefly that they seem to have the ability to
control only one economic variable at a time. During the 1970s central banks
successfully supported high nominal growth at the cost of runaway inflation.
In the 1980s they successfully strangled inflation at the cost of sharp falls in
real economic activity. In the Great Moderation of the 1990s and early 2000s
they kept a lid on inflation and inflation expectations at the expense of a series
of asset bubbles. Post-2009 central banks have successfully avoided deflation
and kept inflation around their targeted levels, but allowed continuing slack
and unemployment in their economies. Maybe helicopter money and a
Deutsche Bank AG/London

Page 53

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

combined nominal GDP target might allow for stable, slack eliminating,
nominal GDP growth. Of course there is a question of what it might leave
uncontrolled
A double-edged sword?
In spite of the theoretical power of helicopter money it has and continues to
face strong opposition. Indeed, fast-forward 9 years from 2002s Professor
Bernanke to 2011s Chairman Bernanke and we see the quote, monetary
policy can be a powerful tool, but it is not a panacea for the problems currently
faced by the US economy. This reasoning can be seen as coming straight out
of a 2003 Fed Policy Paper presented to the FOMC (which by then included
Bernanke among its numbers) by Vincent Reinhart which concluded on
money-financed tax cuts and other extreme policy measures that, You can
see why I put this list last. These options would change how we are viewed in
financial markets, involve credit judgments of a form we are not used to,
perhaps smack of desperation, and pulls us into a tighter relationship with
other parts of the government.
The message is simple helicopter money is a step too far for central bank
policy. It risks creating unintended consequences (change how we are viewed
in financial markets) and asset market mispricing (involve credit judgments
of a form we are not used to) as well as possibility abandoning the
independence of the Fed (a tighter relationship with other parts of the
government).
As with all economic decisions, there is a trade-off. And as with all trade-offs,
priorities and preferences change. There is an argument that helicopter money
could put the worlds economies back on a stronger nominal GDP growth
track, boost spending, increase confidence and reduce debt burdens. But it
could well do so at the risk of financial market mispricing (i.e. asset bubbles)
and letting the inflation genie out of its bottle after three decades ensuring it
stayed in it.

NGDPT and Helicopter Money pose Deeper Questions then


Any Framework and Policy Yet Used
In a world which, to our eyes, continues to be weighed down by uneroded debt
burdens; nominal GDP targeting and helicopter money could be the logical next
step in todays monetary-stimulus heavy economies. More than that, it could be
successful in a way that its less radical predecessors have not been.
However the decision to go down the path of helicopter money would pose
deeper questions and possibly far greater risks then any policy enacted so far.
It is still our view that aggressive expansionary monetary policy post-2008 put
capitalism on hold, saving the economy from undergoing the type of creative
destruction debt liquidation and businesses failure Schumpeter pointed out as
being at the very core of capitalism. In turn this has prevented the type of
rejuvenation that might have been expected post-crisis. Maybe the scale of
the GFC meant such activism was a necessary and unavoidable response as
the alternative would likely have been a socially divisive depression.
With so much previously unforeseen unorthodox monetary policy conducted
since 2008, its impossible to rule out NGDP targeting or even helicopter
money. Perhaps the closest thing we have to this at the moment is in Japan.
Perhaps this will be a test case for such policy that might herald its global
adoption or consignment to the economic graveyard.

Page 54

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Capitalism on Hold: Japan


Abenomics and Helicopter Money
The Lost Decades of Japan could be seen as evidence that QE and greater
monetary policy activism in the face of structural issues does not generate
significant nominal or real growth. This would support our earlier remarks on
the possibility that aggressive monetary policy can put capitalism on hold
and so create the conditions for continued stagnation.
Of course the counter to this would be that to date policy has not been
aggressive enough. Abenomics, with its combination of aggressive monetary
policy expansion and stimulus spending, is a very close cousin to Helicopter
Money and if the BoJ decided never to sell the assets it purchased during the
programme it could be argued they may be one and the same.
So how does Abenomics compare to Japans last two decades of fiscal and
monetary activism? Whilst on the fiscal side Abenomics doesnt represent a
major policy shift with its extra government public works spending worth
10.3tr (around just 2% of GDP), on the monetary side Abenomics is truly
radical in its scale. The magnitude of Abenomics monetary policy elements is
unique in comparison to both Japans long history of monetary easing
programmes and to the rest of the worlds financial crisis and aftermath
responses.
First looking at Abenomics in contrast to Japans asset purchase programmes
of previous years its clear the sheer scale of purchases is on an entirely new
level. When the BoJ first launched QE in Q1 2001 it expanded its balance sheet
by a total of 40.5 trillion yen, with the balance sheet peaking in Q4 2005. At the
time this asset expansion, averaging 675bn yen a month (or around $6bn at
the March 1st 2001 USDYEN exchange rate of 117) was unprecedented in
terms of scale. Abenomics monetary policies are in an entirely different
stratosphere, amounting to balance sheet expansion of around 5.4 trillion yen a
month or around $55bn a month (using current exchange rate of 97.49). So
simply in terms of monthly yen purchases, Abenomics is bigger by a factor of
8. Figure 98 shows how at the BoJs own forecast rate of purchases,
Abenomics monetary expansion dominates anything seen before in Japan. On
top of this is the fact that the BoJ is more than doubling the average maturity
of the assets it is purchasing, meaning every yen it spends today has much
greater economic traction than in its previous shorter term purchases. The
relative scale of Abenomics purchases in contrast to prior expansion policies
remains even when scaled for nominal GDP (Figure 99, for forward nominal
GDP predictions we use DBs official forecasts out the 2014 and continue the
2014 rate for 2015).

Deutsche Bank AG/London

Page 55

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Figure 98: BoJ Assets (JPYtn)

Figure 99: BoJ Assets / Japan Nominal GDP

400

80%

350

70%

300

60%

250

50%

200

40%

150

30%

100

20%

50

10%

0
1985

1989

1993

1997

2001

2005

2009

2013

Source: Deutsche Bank, Haver

0%
1985

1989

1993

1997

2001

2005

2009

2013

Source: Deutsche Bank, Haver

Even when Abenomics is contrasted against other central banks modern day
aggressive expansionary policies, it still stands out for its sheer scale. Even
against the Feds QE Infinity, especially when forecasting how the BoJ and Fed
balance sheets will develop going forward (here we take a conservative
approach and assume the Fed purchases $85bn a month through 2013, half
that amount a month in 2014 and nothing in 2015). Figure 100 converts the
ECB, Fed, BoE and BoJ balance sheets into dollars at current exchange rates.
Here it is clear that the Fed and BoJ are in a league of their own when it comes
to expansion. However when we scale each central banks balance sheets by
their respective nations nominal GDP (see Figure 101) it again becomes clear
that the BoJs Abenomics dwarf even the Feds blown up balance sheet. Come
Q4 2015 (using the assumptions laid out above) the Feds balance sheet stands
at 24% of nominal GDP, whilst the BoJs stands at 69%. Whilst we might
expect the BoJs balance sheet to be larger relative to GDP then the Feds due
to the much greater importance of bank finance as a % of total finance to
Japans economy then Americas, the sheer scale of the divergence and the
fact that the rate of expansion of the BoJs balance sheet far exceeds that of
the Feds again underlines the fact that Abenomics is something new and
radical in the world of monetary policy.
Figure 100: Developed World Central Bank Balance
Sheets
5.0
4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
1999 2001

Figure 101: Central Bank Balance Sheets / Nominal GDP

ECB ($tn, Aug'13 ER)

80%

BoJ ($tn, Aug'13 ER)

70%

BoE ($tn, Aug'13 ER)


Fed ($tn, Aug'13 ER)

ECB BS/NGDP

BoJ BS/NGDP

BoE/NGDP

Fed/NGDP

60%
50%

40%
30%
20%
10%

2003 2005 2007 2009 2011 2013 2015

Source: Deutsche Bank, Haver

0%
1999 2001 2003 2005 2007 2009 2011 2013 2015
Source: Deutsche Bank, Haver

Another reason why Abenomics is more radical than previous Japanese policy
expansions is how it has been communicated and the context in which it has
been made. When the BoJ first began Quantitative Easing back in March 2001,
the BoJ almost simultaneously released a paper which concluded that QE
Page 56

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

would likely be ineffective. Over the next decade the BoJ persistently sought to
exit both from its QE purchases and its ZIRP policy. This desire to end
expansionary monetary policy even in the face of a weak economy was
exemplified by Governor Hayami who argued that if monetary policy was too
loose for too long then a strong economy and higher inflation would allow
zombie firms to survive longer and so delay structural reform. Therefore all
Japanese monetary policy pre-2013 needs to be viewed through this prism of
such seemingly half-hearted commitments.
Abenomics on the other hand has been launched by a new Japanese
government, under a newly returned Prime Minister Abe and a new BoJ
governor Kuroda chosen for his expansionist tendencies. Furthermore the fact
that so far Abe and Kuroda have been positive about their policies and havent
over-stressed the possible inflation risks has created a context within which
the new policy can have the greatest potential impact. This communication
effect is important. Fiscal expansion is more likely to improve consumption if
consumers dont believe it will be succeeded with higher taxes down the road.
Monetary expansion is more likely to improve borrowing, investment and
spending if households and businesses believe low rates will be around to stay.
Indeed as we highlighted earlier, it is possible for Abenomics (and other QE
programmes in general) to become helicopter money further down the road. As
the UKs Lord Turner pointed out in an important speech and paper last year, the
difference between QE and Outright Monetary Financing (i.e. helicopter money),
resides only in the expectations that exist as to future policy. As can be seen
below in Figure 102 and Figure 103 it is inarguable that government deficits and
debt creation have gone hand-in-hand with greater BoJ purchases much as a
policy of helicopter money would call for. So what is the difference? What Turner
means by the above quote is that at present the only thing separating QE from
helicopter money is the belief that eventually central banks will dispose of their
purchased assets on the market. If central banks decided not to sell those assets
and to continue sending interest payments on these assets back to the Treasury
(as they already do), or they roll them over into continuous non-interest bearing
debt then QE (after the fact) would be helicopter money, a direct monetary
financing of governments fiscal deficits.
Figure 102: Deficits v BoJ Bond Buying (JPYtn)

Figure 103: Total Debt vs. Total Monetary Base (JPYtn)

120

BoJ Assets: Government Securities (LHS)

-120

1,400

National Government Debt (LHS)

140

100

Gen Gov Budget Deficit (RHS, Inv.)

-100

1,200

Monetary Base (RHS)

120

80

-80

60

-60

40

-40

20

-20

0
1988

1992

1996

Source: Deutsche Bank, Haver

2000

2004

2008

0
2012

1,000

100

800

80

600

60

400

40

200

20

0
1988

1992

1996

2000

2004

2008

2012

Source: Deutsche Bank, Haver

Perhaps a reason that QE has arguably not had a big impact on growth to date
is that most people believe (and indeed Ben Bernanke keeps telling them) that
central banks will eventually reduce and then sell down their government debt
purchases. Therefore there is an expectation that at some later date interest
rates will rise and government debt will remain at highly elevated levels. If it
came to be expected that QE purchases would never be sold down this could
Deutsche Bank AG/London

Page 57

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

(theoretically) increase consumption and investment directly no longer relying


for much of its impact as current policy does on higher asset prices working
their way through the impaired banking system into lower lending rates and
then into growth as highlighted in the earlier helicopter money section.
Monetary policy and growth
It is an open question whether pushing monetary policy aggression to a whole
new stratosphere can generate sustainable real growth. The evidence from
Japanese monetary policy in the 1990s and 2000s and US/UK/EA policy post2008 certainly points to aggressive monetary policy putting capitalism on hold
and embedding deep-set structural problems, and not generating growth. If
dropping interest rates to zero was Unorthodox Policy #1 and QE was
Unorthodox Policy #2 then it seems very possible Helicopter Money will be
Unorthodox Policy #3. Whether this new level of expansionism, with all the
hopes and theoretic power it is supposed to hold, can generate growth of the
red-hot rather than lukewarm kind remains to be seen.
However in so much as it could potentially raise nominal GDP, it may become
an increasingly more attractive policy option around a global economy
(especially DM) economy that faces many natural and structural growth
concerns in the year ahead. Forcing the nominal economy to grow into the
problems of the bubble era could be the most realistic policy choice over the
remainder of the decade.
Before we move to looking at historical returns across asset classes and our
annual mean reversion exercise, we end the wordy part of the report with a
short two-page chapter marking the 100th birthday of the Fed that will arrive
this December. Given that we think central banks continue to hold the key for
economies and asset prices, it is interesting to see the impact that the most
powerful central bank in the world has had since its inception.

Page 58

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

100 Years of the Fed


In December of this year the Federal Reserve will be 100 years old after having
been created on 23 December 1913. Their creation stemmed from a desire to
end a series of damaging bank runs that culminated in the extreme 1907
banking crisis. Although other central banks existed prior to the Fed, the Fed
soon became the dominant global central bank given that their emergence
coincided with the US becoming the largest and most powerful economy in
the world.
As the anniversary approaches we thought it would be interesting to look at
what impact their creation had on the economy and on asset prices by looking
at the world pre and post the Fed. It also serves as a useful reminder of their
ability to impact the financial world which remains especially relevant today.
Nominal growth and inflation higher, real growth lower
Figure 104 looks at the 100 years before and after the Feds creation in terms
of nominal and real GDP and then in Figure 105 we look at the two periods
average inflation.
Figure 104: Average Annual GDP
7%

Nominal

Figure 105: Average Annual CPI


3.5%

Real

6%

3.0%

5%

2.5%

4%

2.0%

3%

1.5%

2%

1.0%

1%

0.5%

0%

CPI

0.0%
1814-1913

Source: Deutsche Bank, GFD

1914-2013

1814-1913

1914-2013

Source: Deutsche Bank, GFD

Whats striking is that annual real GDP growth was notably higher pre-Fed but
that nominal GDP was a full 2% p.a. lower. The Fed appear to have had a fairly
large influence on the price level which is supported by the fact that CPI has
been 3% p.a. higher in the Fed era than it was in the 100 years before. The
Fed Era coincided with the US moving from a rapidly growing economy to
one that was both more mature and the largest in the world. This helps explain
the lower real growth post the Fed but their influence has been firmly felt in
the nominal growth numbers.
The Fed have supported riskier assets more than bonds
In terms of asset prices, given what weve seen above its no surprise to learn
that nominal returns have seen a bigger improvement than real returns in the
post Fed era. Figure 106-Figure 108 look at the two periods for equities,
treasuries and gold in both real and nominal terms.

Deutsche Bank AG/London

Page 59

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Figure 106: Annualised Equity Total

Figure 107: Annualised Treasury

Figure 108: Annualised Gold

Return (S&P 500)

Total Return (10yr)

Performance

12%

Nominal

Real

6%

Nominal

Real

5%

10%

5%

4%

8%

4%

3%

6%

3%

2%

4%

2%

1%

2%

1%

0%

0%

0%

-1%

1814-1913
Source: Deutsche Bank, GFD

1914-2013

1814-1913

1914-2013

Source: Deutsche Bank, GFD

Nominal

1814-1913

Real

1914-2013

Source: Deutsche Bank, GFD

Interestingly the real returns on equities have been fairly consistent over the
two periods but nominal returns in the modern era have been nearly 3% p.a.
higher. In terms of bonds, real returns dropped by more than 2.5% p.a. to
under 2% in the post-Fed era.
The returns on gold are interesting as prior to the Fed, the price of gold was
largely fixed hence the near zero real and nominal returns. However in the
world post-Fed weve seen a nominal return of above 4% p.a. and real returns
at around 1% p.a.. Virtually all of this growth has occurred in the periods
where the dollar has been free floating rather than on some kind of metalbased currency system.
This analysis fits in with what we discussed in an earlier chapter where we
argued that without money creation (by central or commercial banks) or a
change in money velocity, it is impossible to grow nominal GDP. In an
economy with a fixed money supply, all real GDP growth will be deflationary.
While there was some money creation in the pre-Fed era we can see evidence
that nominal GDP struggled to outpace real GDP. Such an environment may
have contributed to the banking crises that the US more frequently
experienced as recessions quickly saw negative nominal GDP. Debt and
negative nominal GDP growth are a lethal combination, which weve had to
reacquaint ourselves with over the last 5 years.
So the Fed appears to have made a significant difference to nominal GDP
and inflation in their 100 year existence. This has had a positive impact on
nominal asset price returns without impacting real returns that significantly.
The clear exception is bonds where the Feds creation seemingly changed
their return outlook.
Looking forward, this section can be used as evidence that while central banks
conduct policy in a fiat currency world they have the ability to manipulate the
nominal economy and nominal asset prices. Given the current high levels of
debt and lacklustre global nominal growth we think they may have more heavy
lifting to do. The early years of the Feds second century may start more
actively than the market currently expects with its eye fixed on tapering and
the ending of QE.

Page 60

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Mean Reversion
Assessing the mean reversion model through time
We now move to the part of the report that looks at historic asset prices across
the globe and how current valuations compare to long-term trends. One of the
original motivations for first compiling this report back in 2005 was the belief
that traditional developed world asset classes exhibited a rhythm of returns
through time that were subject to clear mean reversion tendencies. In every
edition of this report weve updated what we consider to be the potential
future returns of various asset classes based on them mean reverting over
different time horizons.
In this years study we again update our analysis but enhance it by looking
back through history to assess what our mean reversion models would have
suggested for future returns at different points in history. Clearly the analysis is
purely hypothetical, with the major weakness being that we use assumptions
for modelling past returns based on what we know now. So our 1900 forecasts
are based on what we now think to be the long-run factors that we need to
mean revert. Nevertheless it does show where valuations are today relative to
where theyve been in the past.
In Figure 109 we show the results of this through time for the S&P 500 and US
Treasuries (10yr). We also show what 10 year returns are likely to be from this
starting point based on our mean reversion methodology (**).
Figure 109: Mean Reversion Expected 10 Year Annualised Returns vs. Actual for the S&P 500 (left, based on 1958
method) and 10yr Treasuries (right)
30%

Mean Reversion

Actual

25%

20%

18%

12%

10%

10%

8%

0%
-5%
1871 1886 1901 1916 1931 1946 1961 1976 1991 2006

Actual

14%

15%

5%

Mean Reversion

16%

6%

4%
2%
0%
1871 1886 1901 1916 1931 1946 1961 1976 1991 2006

Source: Deutsche Bank, GFD

For the S&P 500 the mean reversion model has been a good predictor of
annualized returns over the next 10 years. Interestingly the only year where
future decade returns were expected to be negative was 1999, just before the
equity collapse of 2000. After the Lehman collapse in 2008, the predicted
returns increased to just under 7% p.a. which was the highest since the late
1990s but actually still below the long-term average. The current prediction is
back down to an annualized return of 3.24% over the next 10 years. The
predicted 10 year annualized return was only once below 5% in any year
between 1871 and 1996. So this shows that we still live in a world of elevated
equity valuations relative to history. It doesnt mean that positive returns wont
be seen but it perhaps shows the impact of central bank liquidity in bringing
future returns forward.

Deutsche Bank AG/London

Page 61

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

In terms of bonds, future 10 year annualized returns seem a bit more


mechanical and track the starting yield much more closely than equities where
valuations can swing more violently. As such actual returns track the mean
reversion model quite closely. At the moment, expected nominal returns are
2.06% annualized over the next 10 years. In 2011 and 2012, this number
dipped below 2% for the first time since the early-mid 1950s. The still low
prediction indicates the obvious limitation of returns in the asset class going
forward.
In Figure 110 we combine the two charts and show the annualized 10 year
mean reversion returns of a portfolio weighted 60/40 equity/bonds.
Figure 110: Mean Reversion Expected 10 Year Annualised Returns vs. Actual
for a 60/40 US Equity/Bond Portfolio
20%

Mean Reversion

Actual

18%
16%
14%
12%
10%
8%
6%
4%
2%

0%
1871

1886

1901

1916

1931

1946

1961

1976

1991

2006

Source: Deutsche Bank, GFD

This chart again highlights what a low return world were potentially in. Before
1997 the model never dipped below a 4% p.a. annualized return. Since 1997 the
only years the model went slightly back above it were in 2002 and 2008-2009.
The actual annualised 10 years returns of this portfolio since the late 1990s has
generally been as low as the model suggested it would be. The current
prediction of 2.77% p.a. return is the 4th lowest in the 143 years since 1871. The
only years with a lower 10 year prediction were in 1998, 1999 and in 2000.
So the model suggests its going to be very difficult to generate real returns
from this starting point. The long-run average inflation rate of the US since
1900 is 3.2% which if repeated would imply a negative real return from this
starting point for a US based 60/40 equity/bond portfolio.
Such an exercise is not easy to repeat across the globe as the US is one of the
few countries that has long histories of growth, inflation, earnings, PE ratios,
and bond yields without going through huge permanent structural change
(politics, war etc).
We can repeat the exercise for the UK, using full data back around 80-90 years.
Figure 111 and Figure 112 show the results.

Page 62

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Figure 111: Mean Reversion Expected 10 Year Annualised Returns vs. Actual for the FTSE All Share (left) and 10yr Gilts
(right)
40%

Mean Reversion

18%

Actual

35%

16%

30%

14%

25%

12%

20%

10%

15%

8%

10%

6%

5%

4%

0%

2%

-5%
1927

1942

1957

1972

1987

Mean Reversion

0%
1958

2002

1973

1988

Actual

2003

Source: Deutsche Bank, GFD

The results are fairly similar in the UK market to that seen for the US. The
projected returns are perhaps slightly higher, largely reflecting lower current
valuations in the UK equity market and also a higher historic rate of inflation
that the UK mean reversion model uses and assumes to be the future trend.
Figure 112: Mean Reversion Expected 10 Year Annualised Returns vs. Actual
for a 60/40 UK Equity/Bond Portfolio
30%

Mean Reversion

Actual

25%

20%
15%
10%
5%

0%
1958

1963

1968

1973

1978

1983

1988

1993

1998

2003

2008

2013

Source: Deutsche Bank

As already mentioned doing the same analysis for equities in Euroland


countries is more difficult owing to a lack of historic earnings data. However
we can look at the bond data. Owing to the fact that core European bond
yields are also very low in an historical context its no surprise to see that
mean reversion expected returns going forward are likely to be some of the
weakest through time. In Figure 113 we start with German and French
government bonds. For Germany we track the data back to the end of WWII
and we can see that returns going forward are expected to be lower than
anything weve seen since 1946 and are actually very unlikely to be much
above 0.5% p.a. over the next 10 years. Looking now at France where we can
extend the data back further (here we look at the data since 1900), we can see
that expected returns are just below 1.7%, which is a slight improvement from
what mean reversion would have suggested at the end of 2012. Additionally
with the exception of the 10 year periods starting in 1941 and 1942 as well as
the early years of the 20th century this would be the lowest 10 year annualised
returns we would have ever seen. So certainly the picture for core European
bonds is not too different from what we have shown for the US and UK.
Deutsche Bank AG/London

Page 63

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Figure 113: Mean Reversion Expected 10 Year Annualised Returns vs. Actual for 10yr German (left) and French (right)
Government Bonds
12%

Mean Reversion

20%

Actual

10%

Mean Reversion

Actual

15%

8%
6%

10%

4%

5%

2%

0%

0%
-2%
1946

1956

1966

1976

1986

1996

2006

-5%
1900

1920

1940

1960

1980

2000

Source: Deutsche Bank, GFD

Focusing now of the periphery, in Figure 114 we provide the same analysis for
Italy and Spain. However the conclusion is clearly going to be slightly different.
Although the mean reversion expected returns for the next 10 years are not
likely to be particularly high by historical standards they are broadly in line with
what we have seen over the last 10 years. Obviously the current yield starting
point for peripheral government bonds (above 4%) is notably higher than
current 10 year Bund yields (hovering around 2%).
Figure 114: Mean Reversion Expected 10 Year Annualised Returns vs. Actual for 10yr Italian (left) and Spanish (right)
Government Bonds
25%

Mean Reversion

Actual

20%
15%
10%
5%
0%
1900

1920

1940

1960

1980

2000

20%
18%
16%
14%
12%
10%
8%
6%
4%
2%
0%
1946

Mean Reversion

1956

1966

1976

Actual

1986

1996

2006

Source: Deutsche Bank, GFD

Despite the potentially better returns expected for peripherals the general
conclusion for mean reversion expected Eurozone bond returns is that they are
likely to be low by historical standards, similar to what we have already shown
for 10 year Treasuries and Gilts.
European equities appear cheaper than US but earnings will be important
To get an assessment of potential value in Euroland equities we cant use the
same mean reversion exercise as we have a lack of historic European earnings
data. However we can update our analysis from last years study (A Journey
into the unknown, 03 Sep 2012) where we looked at current PEs and ERPs
for the various countrys equity markets and examined where they stood
relative to history. In Figure 115 we aggregate all the results with each
countrys data starting from the first available point (which we detail in the
table). We show the current PE, the average, median, low and high. We then
Page 64

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

repeat this for the rolling 5yr and 10yr Shiller PE ratio which looks at average
earnings over these periods rather than spot. We do the same calculations for
the ERP and then rank the current points for each indicator relative to each
countrys own history. A reading closer to 0% indicates that a market is cheap
relative to its own history and a reading closer to 100% shows that its
expensive. To make the table easier to read weve added a colour coded heat
map. For those observations in the top 10% of cheapness weve shaded them
darker blue and for those between 10-25% weve used a slightly lighter blue.
At the other end of the scale those shaded darker grey are in the bottom 10%
on a valuation basis and those in the bottom 25% are shaded a slightly lighter
grey. Please see last years study for further details on the methodology.
Figure 115: Current PE Ratios and ERP Relative to Average, Median, High and Low
France
Spot PE

Current

Italy
(EUR Yield)

Spain

Spain
(EUR Yield)

UK

US

13.10

19.25

15.80

15.66

17.62

75.8%

63.0%

78.6%

70.6%

Average

13.31

15.12

17.36

14.49

12.14

16.01

Median

12.75

14.33

16.40

13.39

11.81

14.96

Low

5.79

8.43

5.79

6.94

3.96

5.21

High

28.20

26.82

43.95

25.34

28.64

122.41

Jan 73

Jan 73

Jan 86

Mar 87

Mar 27

Mar 1871

14.72

15.21

10.73

9.07

13.77

24.87

61.8%

45.6%

25.0%

13.2%

66.2%

91.6%

Average

14.71

16.89

17.55

16.44

12.43

15.96

Median

13.09

15.91

16.28

14.61

11.58

15.74

Low

5.48

8.10

6.54

5.62

4.33

4.26

High

34.34

34.51

41.43

33.16

26.56

37.04

Jan 78

Jan 78

Jan 91

Mar 92

Mar 32

Mar 1876

13.24

14.20

8.94

9.68

13.81

22.93

25.8%

20.4%

12.3%

8.6%

63.0%

87.0%

Average

17.12

19.38

19.57

19.54

12.64

16.62

Median

16.32

18.59

19.20

19.99

11.75

16.29

Low

6.76

9.09

6.72

6.90

4.38

4.69

High

37.64

40.98

44.12

32.53

28.63

45.52

Mar 37

Mar 1881

Current

Start Date
Current
Current Rank

Start Date
Current

Jan 83

Jan 83

Jan 96

Mar 97

4.43

4.06

7.73

5.99

1.98

3.55

3.67

4.68

6.25

4.95

47.3%

52.4%

8.1%

24.5%

48.7%

37.2%

51.8%

31.6%

47.4%

41.1%

Average

5.20

5.23

3.55

3.60

2.16

2.30

4.02

3.86

7.87

5.28

Median

4.27

4.25

3.56

3.60

1.79

1.76

3.83

3.41

6.05

4.15

-0.34

-0.34

-2.01

-2.01

-4.18

-4.18

-3.18

-3.18

-1.80

-16.52

Current Rank

Low
High
Start Date
ERP (Shiller PE (5yr))

Italy

34.0%

Current Rank

ERP (Spot PE)

Germany
(EUR Yield)

17.51

Start Date

Shiller PE (10yr)

Germany

85.2%

Current Rank

Shiller PE (5yr)

France
(EUR Yield)

Current

15.38

15.38

11.41

10.18

14.09

14.57

11.98

12.21

31.41

33.44

Jan 73

Jan 73

Jan 73

Jan 73

Jan 86

Jan 86

Mar 87

Mar 87

Mar 27

Mar 1871

5.51

5.15

6.67

4.93

6.10

7.67

8.37

9.37

7.12

3.29

36.6%

39.4%

16.7%

36.3%

23.3%

20.0%

9.4%

9.8%

46.1%

54.9%

Average

4.31

4.35

3.10

3.15

3.32

3.49

4.47

4.27

7.97

5.35

Median

3.56

3.58

3.22

3.39

2.22

2.24

4.52

3.81

6.66

3.92

-2.21

-2.21

-3.37

-3.37

-3.08

-3.08

-0.69

-0.69

-1.92

-6.96

Current Rank

Low
High
Start Date
ERP (Shiller PE (10yr)) Current

14.56

14.56

9.99

9.67

12.77

14.35

13.13

16.86

30.29

41.28

Jan 78

Jan 78

Jan 78

Jan 78

Jan 91

Jan 91

Mar 92

Mar 92

Mar 32

Mar 1876

6.27

5.91

7.14

5.39

7.96

9.53

7.67

8.68

7.10

3.63

15.3%

18.3%

6.6%

14.8%

15.7%

11.9%

8.2%

11.7%

48.1%

50.3%

Average

2.69

2.73

2.11

2.18

3.79

4.00

3.92

3.67

7.82

5.32

Median

1.97

2.09

1.69

1.85

2.96

2.91

4.21

3.28

6.82

3.69

-1.97

-1.97

-3.24

-3.24

-1.17

-1.48

-0.30

-0.65

-1.81

-7.02

Current Rank

Low
High
Start Date

9.78

9.40

9.63

8.41

12.34

13.92

9.82

13.54

30.58

38.50

Jan 83

Jan 83

Jan 83

Jan 83

Jan 96

Jan 96

Mar 97

Mar 97

Mar 37

Mar 1881

Note: Data to .... Blue shading indicates current point amongst 25% of cheapest valuations with darker blue shading indicating amongst 10% cheapest valuations. Grey shading indicates current point amongst 25% of
richest valuations with darker grey shading indicating amongst 10% richest valuations.
Source: Deutsche Bank, Bloomberg Finance LP, Datastream, GFD

Deutsche Bank AG/London

Page 65

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

As already highlighted from our mean reversion exercise US and UK are not
cheap from this starting point, and theres certainly nothing in this analysis
that completely contradicts that view. While UK equities may sit within the
central 50 percentile across all the various measures (except spot PE), US
equities look rich based on the Shiller PE valuations and with the exception of
the Spot PE, ERP valuations are above the 50th percentile.
Moving to Europe and the picture is arguably more attractive for equities. With
the exception of French and to a lesser extent Italian and Spanish equities on a
spot PE basis, current valuations are generally on the cheap side of median
with many reading below the 25th percentile. Across the board European
equities look notably cheaper than US and UK equities relative to their
domestic bond markets. They also look cheaper when earnings are adjusted
over the cycle.
That said there are probably a couple of points that are worth considering. First
of all they certainly dont appear to be quite as cheap as when we did this
analysis a year ago. Secondly, we are still hugely dependent on future earnings
growth. Despite the recent improvement in European macro economic data
there has been little evidence of this pushing through to earnings. Its possible
that earnings struggle to mean revert back to their pre-crisis trend in some
growth impaired European countries. However if you believe in mean reversion
the trade is definitely to be short US equities against being long Europe.

Mean reversion across asset classes


Having highlighted how potentially important mean reversion has been as a
tool for assessing longer-term return potential (with a certain amount of
hindsight) we now provide our usual analysis across the various different
assets.
In Figure 116 we show what nominal and real returns could be over the next
decade if assets revert back to their long-term average valuations. A brief
appendix is posted at the end of this section that takes us through our
methodology for the mean reversion exercise. It basically assumes that
earnings, PE valuations, inflation, real yields and economic growth return to
their long-run averages/trend. As well as US based assets we have also looked
at European credit markets in this exercise.
The results are only meant to be a relative value guide and work best on a
relative basis across asset classes and the longer the time horizon you view
them over. As discussed earlier, we have mainly concentrated on USD assets
in this section. This enables us to delve deeper into history to analyse the longterm rhythm of returns. In reading the results, hopefully one will be able to
understand the type of returns that a sophisticated Developed Market sees
through time.

Page 66

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Figure 116: Potential Annualised Returns Based on Full Mean Reversion over Different Time Horizons
Actual LT Annualised Return*

US Assets

High Yield

iBoxx EUR

iBoxx GBP

Mean Reversion Expected


Real Returns

Nominal

Real

3yr

5yr

10yr

3yr

5yr

10yr

Equity (Trend Earnings/Average PE)

8.5%

6.7%

-11.4%

-4.6%

0.8%

-13.8%

-7.1%

-1.8%

Equity (Trend Earnings/Average PE since 1958)

8.5%

6.7%

-3.7%

0.2%

3.2%

-6.3%

-2.4%

0.6%

Treasury (10yr)

5.1%

3.3%

-2.3%

0.2%

2.1%

-5.0%

-2.5%

-0.6%

Treasury (30yr)

4.6%

1.5%

-2.1%

0.4%

2.3%

-4.7%

-2.2%

-0.3%

IG Corporate Bond

5.7%

2.5%

-0.8%

1.5%

3.3%

-3.5%

-1.1%

0.7%

BBB Bond

6.6%

3.8%

-0.6%

1.8%

3.7%

-3.3%

-0.9%

1.0%

Property

3.4%

0.3%

-5.7%

-2.5%

0.0%

-8.3%

-5.0%

-2.6%

Gold

2.0%

0.3%

-23.7%

-14.1%

-6.1%

-25.7%

-16.3%

-8.5%

Oil

2.2%

-0.1%

-24.9%

-15.0%

-6.6%

-27.0%

-17.2%

-9.0%

All Commodities (1919 Reversion)

1.5%

-0.9%

3.7%

3.3%

2.9%

0.9%

0.5%

0.3%

USD High Yield

8.5%

5.6%

-0.1%

2.3%

4.2%

-2.8%

-0.4%

1.5%

Treasury (Duration Matched)

6.4%

3.6%

-1.9%

0.4%

2.1%

-4.6%

-2.3%

-0.5%

EUR High Yield

-0.7%

1.9%

3.8%

-3.2%

-0.7%

1.3%

Treasury (Duration Matched)

-2.5%

-0.3%

1.4%

-5.1%

-2.9%

-1.2%

Corporate Bond

-1.4%

0.8%

2.4%

-3.9%

-1.7%

0.0%

BBB Bond

-0.2%

1.6%

2.9%

-2.8%

-0.9%

0.4%

Non-Financial Bond

-1.9%

0.4%

2.2%

-4.4%

-2.0%

-0.2%

Non-Financial BBB Bond

-0.9%

1.1%

2.5%

-3.4%

-1.4%

0.1%

Bund (Duration Matched)

-3.2%

-0.7%

1.2%

-5.7%

-3.2%

-1.2%

Corporate Bond

-1.5%

1.3%

3.4%

-4.4%

-1.7%

0.4%

1.8%

3.4%

4.5%

-1.2%

0.3%

1.5%

Non-Financial Bond

-3.0%

0.2%

2.7%

-5.9%

-2.7%

-0.3%

Non-Financial BBB Bond

-0.3%

1.9%

3.6%

-3.2%

-1.1%

0.6%

Gilt (Duration Matched)

-3.3%

-0.3%

1.9%

-6.1%

-3.2%

-1.0%

Corporate Bond

-0.2%

1.9%

3.5%

-2.9%

-0.8%

0.8%

1.6%

3.0%

4.0%

-1.2%

0.3%

1.4%

-1.1%

1.3%

3.2%

-3.8%

-1.4%

0.5%

0.7%

2.4%

3.7%

-2.0%

-0.3%

1.1%

-2.3%

0.2%

2.1%

-5.0%

-2.5%

-0.6%

BBB Bond

iBoxx USD

Mean Reversion Expected


Nominal Returns

BBB Bond
Non-Financial Bond
Non-Financial BBB Bond
Treasury (Duration Matched)
Note: * - Based on longest available series in our historical returns analysis.
Source: Deutsche Bank, GFD

For equities we use two slightly different methods. Method 1 simply looks at
mean reverting earnings back to their long-term trend and PE ratios back to
their long-term average. Method 2 recognises that earnings growth may have
increased (albeit slightly) post 1958 and uses the trend line of earnings seen
since then and the (again slightly higher) average PE ratio seen since. We have
noted in previous studies, including the 2011 version, that up until 1958,
dividend yields were always above bond yields. This situation reversed for the
next 50 years when in November 2008 S&P 500 dividends briefly crossed
above bond yields again. Since this point the two have crossed a few times.
The recent move higher in 10 year Treasury yields following heightened
expectations of Tapering has seen them briefly trade above 3%, which is once
again higher than the c.2.0% dividend yield currently offered by the S&P 500.
The jury is still out however as to whether the post 1958 move to lower
dividends and perhaps higher earnings growth has actually been positive or
negative for equity returns. We think its actually been negative as there is no
conclusive evidence that earnings have broken permanently higher (and not
just cyclically) from their long-term trend post-1958. Basically returns seem to
be higher when investors receive dividends rather than when companies retain

Deutsche Bank AG/London

Page 67

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

dividends and attempt to expand their businesses. Weve written about this in
length in previous studies for those that want to explore the arguments further.
Overall this leaves us preferring method 1 but weve included both results in
the exercise for those that think its a slightly different market now to that seen
prior to 1958 and the great dividend crossover.
If we use method 1, annualised real returns on this method show a negative
trend over the next decade. The returns are slightly better if you use method 2
as we reach positive territory 10 years out but they are still sub-standard
relative to long-term history. Furthermore unlike last year when yields were
close to their all-time lows Treasuries, potentially offer slightly more attractive
mean reversion returns than equities depending which equity mean reversion
method is preferred.
Before we move on from equities we should stress that the biggest problem
with valuations today is that earnings/profits are at a very high share of GDP
relative to history. If this does eventually mean revert, our low future return
numbers are absolutely justifiable. If however weve moved to a permanent
new plateau of higher earnings relative to the size of the economy then our
numbers are too low. Again there is an argument that higher nominal GDP is
needed for equities to grow-in to their current valuations.
Despite the fact the potential Treasury returns look more attractive than they did
in last years report, helped in no small part by the more than 100bps rise in
yields weve seen since early May, the mean reversion results suggest that real
returns are still likely to be negative. This exercise suggests that in nominal terms
returns could be in the 2% p.a. region if we mean revert over the next decade.
Credit still arguably provides some protection from future mean reversion in
government yields. However expected returns are well below the long-term
average levels. In terms of total returns the LT IG corporate and BBB indices
provide a nominal 3.31% and 3.7% annual return respectively on a mean
reversion basis over 10 years (LT average around 6%p.a.), with returns also
positive in real terms (+0.7% and +1.0% respectively p.a.). Expanding this out
to the iBoxx indices (with shorter durations) we see a broadly similar outcome
for the USD market but a somewhat less impressive result for EUR and GBP
credit where real total returns are generally negative, particularly with mean
reversion over shorter horizons. That said EUR and GBP would still be expected
to outperform appropriate government bonds.
Like IG, the extra yield in HY also more than offsets any likely future rises in
government bond yields. However, as we also saw with IG, the very low
underlying yield environment suggests that mean reversion produces future
HY total returns some way below their long-term averages. Interestingly with
spreads now actually tighter than their long-term averages mean reversion
excess returns do not look that much more attractive than the levels we have
shown for IG credit. Given that expected excess returns are less generous than
they were when we conducted this exercise a year ago we also have to be
mindful of default expectations, which tend to have more of a direct impact on
HY than IG.
For property, using Robert Shillers long-term data back to 1900, the asset
class still appears slightly expensive on a mean reversion basis. In nominal
terms our mean reversion suggests house prices could be flat over the next
decade. However this means that in real terms our analysis is still suggesting
negative returns. Whats interesting is that this is the first year since 2005
where valuations on this basis actually look worse than they did a year ago.
Clearly affected by what has been a general improvement in US real estate
over the past year, with most key indicators showing a strong pick-up from
where they were 12 months ago.
Page 68

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Overall, the asset class that continues to stand out in this exercise is
Commodities. If mean reversion of long-term data back over the last century
was your only guide then Oil and Gold are likely to have poor decades in
nominal (-6.6% to -6.1% p.a.) and real (-9.0% to -8.5% p.a.) terms. The
negative expectations for gold exist despite that fact weve seen a notable
decline over the past 12 months as talk has moved from further money
printing to potential QE tapering and ultimately monetary tightening at some
point. This time last year gold was close to $1,800 whilst today we are around
the $1,300 level having fallen to as low as $1,200 as recently as late June.
Its worth noting however that whilst oil and gold are likely to have poor
decades if mean reversion is our only guide, the overall commodity index is
showing positive returns on both a nominal and real basis and irrespective of
the mean reversion horizon. Figure 118-Figure 120 show the reasons for these
discrepancies within commodities. Gold and oil are still at the upper end of
their historic long-term range in real terms whereas the overall index and
copper are not as excessively rich relative to history. Clearly this observation
ignores any structural changes that may have occurred.
Figure 117: All Commodities Real Adjusted
400

All Commodities

Figure 118: Gold Real Adjusted


1,800

350

1,600

300

1,400
1,200

250

1,000

200

800

150

600

100

400

50

200

0
1871 1886 1901 1916 1931 1946 1961 1976 1991 2006
Source: Deutsche Bank, GFD

0
1871 1886 1901 1916 1931 1946 1961 1976 1991 2006
Source: Deutsche Bank, GFD

Figure 119: Oil Real Adjusted


140

Figure 120: Copper Real Adjusted


Oil

1,400

120

1,200

100

1,000

80

800

60

600

40

400

20

200

0
1871 1886 1901 1916 1931 1946 1961 1976 1991 2006
Source: Deutsche Bank, GFD

Gold

Copper

0
1871 1886 1901 1916 1931 1946 1961 1976 1991 2006
Source: Deutsche Bank, GFD

Mean reversion conclusion

Deutsche Bank AG/London

Page 69

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

As ever, the results from this section should be used as a valuation tool and
not a forecast. That being said mean reversion is one of the most useful
investment tools for long-term investors. The caveat would be that this cycle
continues to be unique relative to history and there are risks that some
countries/assets could permanent move away from their long-term trend path.
So some caution is required in what is nevertheless a technique that is a key
influence when we consider future asset price performance.

Mean Reversion Assumptions


As an appendix to this section we outline the methodology and the variables
that we have mean reverted in order to calculate potential returns for the
various asset classes discussed in this study.
Inflation
The starting point, which is essential for calculating possible future returns
across all asset classes (including equities), is to get a future CPI time series.
For this we have just reverted the YoY growth in CPI to its long-term average
(around 3.2%).
Equities
For equities although we have used slightly different methodologies the broad
principles were the same. Essentially we first calculate a mean reverted price
series. We do this by first reverting real earnings back to their long-term trend
line. We mean revert the current PE ratio back to its long-term average.
Combining the reverted earnings and PE ratios we then calculate a price. In
order to calculate total returns we have assumed real dividends revert back to
their long-term trend line. By combining the prices and the dividends we
calculate total returns. As already mentioned we used two slightly different
methodologies the specific of which are outlined in the bullets below.

Method 1: We revert earnings, PE ratios and dividends back to their longterm trend/averages using all available data back to 1871.

Method 2: We revert earnings, PE ratios and dividends back to their longterm trend/averages based on data since 1958. As already mentioned this
recognises that earnings growth may have increased (albeit slightly) post
1958 and the previously discussed dividend crossover.

Treasury/Government bond mean reversion


For Treasuries and other Government bond series we have reverted to the
long-term average real yield which has been calculated by subtracting YoY CPI
from the nominal bond yield. We can then use these yields to calculate
prospective returns.
Corporate bond mean reversion (IG and HY)
For corporate bonds we mean revert credit spreads to their long-term average
level. These spreads coupled with the already calculated Treasury/Government
bond yields give us an overall corporate bond yield that can be used to
calculate possible future returns. We have used appropriate duration matched
Treasury/Government yields for the various different corporate bond series.
For the iBoxx indices, which only have data back to 1999, we have created a
longer-term spread series by regressing the iBoxx spread data against the
Moodys long-term spread series. The results of the regression can be used to
calculate a longer-term spread series, which can be used to calculate the longterm average level that is then used for mean reversion purposes.

Page 70

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

For further details on how we have calculated bond returns (both Government
and corporate) please refer to a previous version of this report (100 Years of
Corporate Bond Returns Revisited, 5 November 2008).
US property and commodity mean reversion
For both US property and the various commodity series we have calculated a
real adjusted price series and simply mean reverted to the long-term average
level of this series.

Deutsche Bank AG/London

Page 71

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Historical US Asset Returns


We now look at long-term US returns going back to the start of the 19th
century (where possible). Figure 121 and Figure 122 show why we invest in
assets over the medium to long term. Using data going back over 200 years, it
is quite clear that history tells us that holding cash on deposit has been a
recipe for wealth erosion. We split the data up by nominal and real returns
through different time periods. We also show returns annualised within each
decade and also by 50 year buckets. This hopefully helps us see both cyclical
and secular trends.
Over the entire sample period, Equities outperform Corporate Bonds, which
outperform Government Bonds, which outperform Cash, which interestingly
has outperformed Commodities. Since 1900, where we have data for the
widest selection of assets, Equities outperform 30yr Governments by 4.83%
p.a., Corporates by 3.78% p.a., Cash by 5.90% p.a., and Commodities by
6.69% p.a. (on a nominal basis).
In this years study there seem to be more assets that have seen their recent
performance (particularly the last 5 years) diverge from the longer-term trend.
Commodities probably continue to exhibit the greatest divergence between
recent and long-term performance. Over the last 5 and 10 year periods Copper
and Oil have actually provided the largest returns across the assets analysed in
this study, more than 15% p.a. for both assets over the last 5 years (real
adjusted). However their long-term performance has actually struggled to
exceed inflation as can be seen in Figure 121 and Figure 122 later in this section.
The last 5 years for equities have shown a huge swing from where we were
last year, owing to the fact that this period now starts at the beginning of 2009
(close to the trough for equities). Therefore the 12.5% p.a. return for the last 5
years is comfortably higher than the c.6.7% p.a. long-term average. Weve
seen the opposite move for Treasuries. The back-up in yield during Q2 this year
has seen both 10yr and 30yr yields rise above where they were at the
beginning of 2009 and have ultimately seen real returns move into negative
territory for the past 5 years. The longer-term average sits in the 3-3.5% p.a.
range (real adjusted) for 10yr Treasuries.
Property (US) is an asset class that has only just outpaced inflation (0.29% p.a.
real) over the long term. We would stress that this is a price-only series and
doesnt include potential rental yields but its a reminder that real adjusted capital
returns in the asset class can be minimal over longer time periods, especially in
markets like the US where overall ample space and a lack of restrictive planning
prevents their being a national supply shortage relative to demand.
Non-financial IG Corporate Bonds have steadily out-performed Government
Bonds over all medium-term time periods. The levels of defaults historically
seen in IG very rarely erode the additional spread the asset class provides.
Periods of under-performance are much more likely to be driven by temporary
spread widening. These spread changes tend to be highly cyclical whereas
equity and Treasury valuations tend to exhibit a more secular pattern.
HY is still a fairly new market in the context of this study, with new issuance
(rather than simply fallen angels) only existing from the mid 1980s. In this time,
weve been through longer and less frequent business cycles than long-term
history, but also through two deep default cycles (2000-2003 and 2007-2009),
with the former far worse for HY (especially in Europe) than it was for the
overall economy. So we would argue that we dont have enough data yet to
assess what a likely long-term return number for HY should be. However the
Page 72

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

excess return of 2.12% p.a. over Government Bonds since 1989 (2.32% p.a.
since 1986) might be argued to be disappointing relative to the lower risk
returns seen in IG credit. Much of this disappointment has been obscured by
the high total returns in fixed income which has given the asset class a healthy
8.48% p.a. nominal return over the last 25 years (since 1989) and 8.79% p.a.
since 1986. This is relevant as HY investors are more total return biased than
the more excess return biased IG investors.
In the following section (starting on page 76) we extend the analysis of
historical asset returns to equity and bond markets around the world.

Deutsche Bank AG/London

Page 73

Corp Bond

AAA Bond

BBB Bond

Treasury
(10yr)

Treasury
(30yr)

HY Bond

Treasury (HY
House Prices
Matched) Treasury Bill (Price Only)

Gold

Copper

Oil

Wheat

Commodities
(Economist)

last 5yrs (2009-2013)

14.89%

9.06%

4.63%

12.28%

1.58%

-1.25%

15.97%

1.32%

0.09%

last 10yrs (2004-2013)

6.01%

7.12%

6.40%

7.36%

4.71%

5.68%

7.86%

3.76%

1.58%

0.97%

9.71%

19.38%

22.43%

5.46%

7.73%

0.28%

12.90%

11.84%

12.75%

5.37%

last 15yrs (1999-2013)

3.77%

7.53%

7.04%

7.79%

5.07%

5.57%

6.91%

4.44%

6.22%

2.16%

3.23%

11.09%

10.62%

15.68%

6.42%

last 25yrs (1989-2013)

9.69%

9.22%

8.89%

9.52%

7.26%

8.12%

8.48%

6.36%

5.51%

3.41%

2.92%

4.72%

3.03%

7.61%

1.48%

last 50yrs (1964-2013)

9.64%

7.73%

7.37%

8.12%

7.14%

1.67%

6.65%

5.22%

4.58%

7.63%

4.85%

7.45%

2.13%

last 75yrs (1939-2013)

10.62%

6.03%

5.57%

6.62%

4.26%

5.48%

4.94%

3.95%

4.54%

5.04%

4.61%

5.71%

2.99%

last 100yrs (1914-2013)

9.84%

5.88%

3.83%

5.15%

5.00%

3.61%

3.57%

4.30%

3.12%

3.84%

1.90%

last 125yrs (1889-2013)

9.09%

4.66%

2.89%

3.44%

3.43%

2.04%

3.93%

1.48%

last 150yrs (1864-2013)

8.78%

2.47%

4.74%

3.53%

2.56%

1.38%

2.23%

1.18%

last 175yrs (1839-2013)

1.45%

8.72%

4.78%

3.74%

2.43%

1.48%

last 200yrs (1814-2013)

8.41%

4.92%

2.16%

0.82%

since 1800

8.48%

5.07%

2.02%

0.86%

since 1900

9.44%

5.65%

since 1920

9.94%

6.12%

5.87%

since 1930

9.38%

6.05%

since 1971

10.18%

9.25%

4.75%

4.61%

3.54%

3.38%

3.76%

2.52%

3.74%

1.98%

2.75%

6.60%

5.33%

5.10%

3.62%

3.56%

4.58%

3.10%

3.32%

1.01%

2.22%

5.79%

6.52%

5.31%

4.98%

3.59%

3.91%

5.14%

3.54%

4.36%

1.91%

3.12%

8.75%

9.74%

7.75%

7.68%

5.24%

4.79%

8.77%

4.35%

8.37%

3.06%

5.05%

RETURNS BY DECADE

Deutsche Bank AG/London

1800-1809

11.09%

9.12%

0.00%

-1.62%

1810-1819

4.91%

6.23%

0.00%

-4.63%

1820-1829

6.94%

5.53%

0.00%

-1.63%

1830-1839

5.34%

2.75%

0.67%

1.38%

1840-1849

7.83%

7.47%

5.02%

-0.03%

-2.57%

1850-1859

1.62%

3.98%

5.08%

0.00%

2.35%

1860-1869

18.34%

6.30%

5.04%

1.81%

1.90%

-12.73%

-1.80%

2.91%

1870-1879

7.73%

3.67%

4.11%

-1.78%

-2.05%

-14.26%

5.23%

-3.89%

1880-1889

5.68%

5.48%

3.04%

0.00%

-1.66%

-0.70%

-5.09%

-0.63%

1890-1899

5.37%

3.93%

2.33%

0.00%

-1.26%

4.88%

-1.21%

-0.54%

1900-1909

9.92%

4.37%

1.63%

2.17%

3.04%

1.97%

0.00%

-3.55%

-1.43%

6.06%

1.56%

1910-1919

4.35%

2.60%

2.52%

2.52%

3.28%

3.15%

0.00%

3.34%

13.33%

7.19%

9.09%

1920-1929

14.78%

6.71%

6.52%

7.25%

5.48%

6.05%

3.88%

0.65%

0.00%

-0.48%

-4.98%

-6.18%

-4.99%

1930-1939

-0.47%

6.41%

7.48%

6.27%

3.95%

5.49%

0.58%

-1.21%

5.41%

-3.51%

-1.81%

-2.22%

-1.25%

1940-1949

8.99%

3.92%

2.92%

5.42%

2.70%

2.42%

0.48%

8.12%

1.47%

4.00%

0.28%

7.64%

5.17%

1950-1959

19.26%

0.16%

-0.08%

0.59%

0.39%

-0.50%

2.02%

2.97%

-1.38%

5.96%

1.46%

-0.69%

-0.02%

1960-1969

7.76%

0.57%

0.42%

0.89%

2.76%

0.51%

4.06%

1.85%

0.04%

5.43%

0.78%

-2.96%

1.09%

1970-1979

5.77%

5.34%

5.02%

5.83%

6.08%

3.71%

6.48%

7.99%

32.23%

6.28%

28.04%

11.43%

15.61%

1980-1989

17.47%

13.72%

13.03%

14.42%

12.78%

12.64%

9.13%

6.78%

-2.85%

0.57%

-5.40%

-0.74%

-0.28%

1990-1999

18.21%

9.31%

8.84%

9.99%

7.98%

8.40%

10.59%

7.27%

4.95%

2.69%

-4.02%

-2.12%

1.67%

-6.31%

-1.15%

2000-2009

-0.95%

8.89%

8.91%

8.66%

6.63%

7.03%

6.57%

6.04%

2.74%

3.30%

14.32%

13.96%

11.91%

6.67%

7.75%

2010-2013

12.17%

8.03%

7.04%

8.55%

4.55%

6.99%

9.14%

2.73%

0.08%

1.85%

6.24%

0.44%

7.99%

11.69%

1.06%

0.13%

-1.83%

0.00%

-0.16%

5.70%

RETURNS BY HALF CENTURY


1800-1849

7.20%

6.20%

1850-1899

7.61%

4.67%

1900-1949

7.39%

4.79%

1950-1999

13.55%

5.69%

5.33%

2000-2013

2.63%

8.64%

8.37%

3.91%

0.48%

3.25%

3.72%

2.24%

2.49%

1.35%

-0.09%

0.89%

2.34%

1.80%

6.21%

5.91%

4.84%

5.30%

4.43%

4.00%

3.17%

4.72%

-0.03%

2.87%

8.63%

6.04%

7.02%

1.97%

2.88%

11.95%

9.92%

10.78%

8.08%

5.79%

7.30%

5.08%

Note: 2013 Returns are calculated up to 31August. So for example the last 5 years data is actually for 4 years and 8 months, 10 years for 9 years and 8 months. Source: Deutsche Bank

12 September 2013

Equity

Long-Term Asset Return Study: A Nominal Problem

Page 74

Figure 121: Nominal Returns for US Assets over Different Time Horizons

Corp Bond

AAA Bond

BBB Bond

Treasury
(10yr)

Treasury
(30yr)

HY Bond

Treasury (HY
House Prices
Matched) Treasury Bill (Price Only)

Gold

Copper

Oil

Wheat

Commodities
(Economist)

last 5yrs (2009-2013)

12.53%

6.82%

2.48%

9.98%

-0.51%

-3.27%

13.59%

-0.76%

-1.96%

last 10yrs (2004-2013)

3.55%

4.63%

3.94%

4.87%

2.28%

3.23%

5.36%

1.36%

-0.78%

-1.10%

7.45%

16.93%

19.92%

3.30%

5.52%

-2.04%

10.28%

9.25%

10.14%

2.93%

last 15yrs (1999-2013)

1.34%

5.01%

4.53%

5.26%

2.61%

3.10%

4.40%

2.00%

3.76%

-0.23%

0.81%

8.49%

8.03%

12.98%

3.93%

last 25yrs (1989-2013)

6.81%

6.36%

6.03%

6.65%

4.45%

5.28%

5.63%

3.57%

3.04%

0.70%

0.22%

1.97%

0.33%

4.79%

-1.18%

-0.99%

last 50yrs (1964-2013)

5.28%

3.45%

3.10%

3.83%

2.88%

last 75yrs (1939-2013)

6.54%

2.12%

1.68%

2.69%

1.59%

2.41%

1.04%

0.42%

3.36%

0.68%

3.18%

-1.93%

0.12%

1.07%

0.12%

0.69%

1.16%

0.75%

1.81%

-0.81%

last 100yrs (1914-2013)

6.43%

2.59%

1.89%

0.00%

1.74%

0.39%

0.35%

1.06%

-0.09%

0.61%

-1.27%

-0.31%

last 125yrs (1889-2013)

6.11%

1.81%

last 150yrs (1864-2013)

6.31%

2.36%

0.61%

0.60%

-0.75%

1.09%

-1.29%

-0.33%

1.17%

0.23%

-0.93%

-0.09%

-1.11%

last 175yrs (1839-2013)

6.54%

2.68%

-0.86%

1.66%

0.38%

-0.55%

last 200yrs (1814-2013)

6.62%

since 1800

6.68%

3.19%

0.48%

-0.84%

3.32%

0.33%

since 1900

6.16%

2.49%

-0.81%

since 1920

7.04%

3.32%

3.07%

since 1930

6.03%

2.80%

since 1971

5.73%

4.84%

1.62%

1.48%

0.44%

0.29%

0.66%

-0.55%

0.63%

-1.07%

-0.32%

3.78%

2.54%

2.32%

0.88%

0.82%

1.82%

0.38%

0.59%

-1.66%

-0.48%

2.55%

3.26%

2.09%

1.77%

0.42%

0.73%

1.92%

0.37%

1.16%

-1.21%

-0.04%

4.35%

5.30%

3.39%

3.33%

0.99%

0.55%

4.37%

0.13%

3.99%

-1.10%

0.80%

RETURNS BY DECADE
1800-1809

11.09%

9.12%

0.00%

-1.62%

1810-1819

4.56%

5.87%

-0.34%

-4.96%

1820-1829

9.05%

7.62%

1.98%

0.31%

1830-1839

3.23%

0.70%

-1.35%

-0.65%

1840-1849

10.82%

10.45%

7.94%

2.75%

0.13%

1850-1859

0.07%

2.39%

3.47%

-1.53%

0.79%

1860-1869

13.58%

2.02%

0.81%

-2.29%

-2.20%

-16.24%

-5.75%

-1.24%

1870-1879

10.20%

6.04%

6.50%

0.47%

0.19%

-12.30%

7.64%

-1.69%

1880-1889

5.68%

5.48%

3.04%

0.00%

-1.66%

-0.70%

-5.09%

-0.63%

1890-1899

5.23%

3.79%

2.19%

-0.13%

-1.39%

4.74%

-1.34%

-0.67%

1900-1909

7.36%

1.93%

-0.74%

-0.22%

0.63%

-0.41%

-2.34%

-5.80%

-3.73%

3.58%

-0.81%

1910-1919

-2.78%

-4.41%

-4.48%

-4.49%

-3.78%

-3.90%

-6.84%

-3.72%

5.59%

-0.14%

1.64%

1920-1929

15.87%

7.72%

6.48%

7.06%

4.87%

1.61%

0.95%

0.46%

-4.08%

-5.29%

-4.09%

1930-1939

1.60%

1940-1949

3.45%

1950-1959

16.67%

1960-1969

5.11%

1970-1979

4.08%

7.53%

8.27%

8.62%

9.72%

8.48%

6.11%

7.69%

2.67%

0.85%

7.60%

-1.50%

0.24%

-0.19%

0.81%

-1.37%

-2.31%

0.05%

-2.52%

-2.79%

-4.63%

2.62%

-3.69%

-1.29%

-4.83%

2.17%

-0.18%

-2.02%

-2.25%

-1.60%

-1.80%

-2.67%

-0.20%

0.74%

-3.52%

3.66%

-0.75%

-2.84%

-2.19%

-1.89%

-2.05%

-1.59%

0.23%

-1.96%

1.51%

-0.65%

-2.41%

2.84%

-1.69%

-5.34%

-1.39%

-1.51%

-1.91%

-2.20%

-1.45%

-1.21%

-3.43%

-0.85%

0.56%

23.14%

-1.03%

19.23%

3.76%

7.65%

1980-1989

11.78%

8.21%

7.56%

8.88%

7.32%

7.19%

3.84%

1.61%

-7.55%

-4.30%

-9.98%

-5.54%

-5.10%

1990-1999

14.83%

6.19%

5.73%

6.85%

4.90%

5.30%

7.43%

4.20%

1.95%

-0.25%

-6.77%

-4.92%

-1.23%

-8.99%

-3.97%

2000-2009

-3.42%

6.17%

6.19%

5.95%

3.97%

4.36%

3.91%

3.39%

0.17%

0.72%

11.46%

11.11%

9.12%

4.00%

5.06%

2010-2013

10.06%

6.00%

5.03%

6.51%

2.59%

4.97%

7.08%

0.80%

-1.80%

-0.07%

4.24%

-1.45%

5.96%

9.59%

-0.85%

RETURNS BY HALF CENTURY

Page 75

1800-1849

7.70%

6.70%

1850-1899

6.85%

3.93%

1900-1949

4.91%

2.37%

1950-1999

9.17%

1.62%

1.27%

2000-2013

0.25%

6.12%

5.86%

3.19%

0.60%

-1.37%

-0.70%

-0.86%

-0.23%

0.87%

1.33%

-0.11%

0.13%

-0.98%

-2.40%

-1.44%

-0.02%

-0.55%

2.11%

1.83%

0.79%

1.24%

0.40%

-0.01%

-0.81%

0.68%

-3.88%

-1.10%

6.11%

3.57%

4.53%

-0.39%

0.49%

9.35%

7.37%

8.21%

5.57%

3.33%

4.81%

2.64%

Note: 2013 Returns are calculated up to 31August. So for example the last 5 years data is actually for 4 years and 8 months, 10 years for 9 years and 8 months. Source: Deutsche Bank

12 September 2013

Equity

Long-Term Asset Return Study: A Nominal Problem

Deutsche Bank AG/London

Figure 122: Real Returns for US Assets over Different Time Horizons

Page 76
0%
0%

10%

10%

10%

10%

4%

4%

5%

5%

8%

8%

-1%

-1%

0%

0%

6%

6%

-6%

-6%

-5%

-5%

-11%

-11%

-10%

-10%

4%

4%

-16%

-16%

-15%

-15%

2%

2%

-21%

-21%

-20%

-20%

0%

0%

-26%

-26%

-25%

-25%

DM

EM
Italy

0%
France

2%

0%
Canada

4%

2%

Germany

6%

4%
DM

US

6%

5%

18%
16%
14%
12%
10%
8%
6%
4%
2%
0%
Germany

10%

5%

DM

France

10%

20%
18%
16%
14%
12%
10%
8%
6%
4%
2%
0%
-2%

Spain

8%

Philippines
Thailand
Norway
Sweden
Malaysia
Korea
Taiwan
Australia
Denmark
US
UK
Mexico
Ireland
South Africa
Canada
Switzerland
Netherlands
Germany
Belgium
India
Austria
France
Japan
Portugal
Italy
Spain
Greece

DM

Japan

8%

UK

10%

15%

30%
25%
20%
15%
10%
5%
0%
-5%
-10%
-15%

Belgium

10%

15%

12%

UK

14%

Australia

20%

South Africa

EM
Mexico
Denmark
Switzerland
Sweden
Netherlands
Australia
US
India
South Africa
UK
Canada
France
Spain
Malaysia
Germany
Belgium
Ireland
Greece
Philippines
Austria
Thailand
Portugal
Korea
Taiwan
Italy
Japan

16%
14%
12%
10%
8%
6%
4%
2%
0%
-2%
-4%

Netherlands

20%

12%

Korea

EM

Sweden

30%
25%
20%
15%
10%
5%
0%
-5%
-10%
-15%

US

15%

9%

EM
Italy

Thailand
Philippines
Norway
Sweden
Malaysia
Denmark
Ireland
Taiwan
US
Korea
Netherlands
South Africa
Germany
Switzerland
Mexico
UK
Australia
Belgium
Canada
Austria
France
Japan
India
Portugal
Italy
Spain
Greece

EM

Australia

14%

9%

DM
Spain

DM

Japan

14%

Germany

25%

France

0%

-5%

Germany

0%

-5%

DM

France

5%

Canada

10%

5%
DM

Belgium

15%

UK

10%

16%
14%
12%
10%
8%
6%
4%
2%
0%
-2%
-4%

US

20%

15%
30%
25%
20%
15%
10%
5%
0%
-5%
-10%
-15%

UK

20%

Australia

25%

Netherlands

30%

South Africa

EM
Mexico
Denmark
Sweden
India
South Africa
Switzerland
US
Netherlands
Australia
UK
Malaysia
Canada
France
Spain
Germany
Ireland
Thailand
Belgium
Austria
Philippines
Greece
Taiwan
Korea
Italy
Portugal
Japan

EM

Korea

25%

Sweden

30%
25%
20%
15%
10%
5%
0%
-5%
-10%
-15%

US

14%

12%

EM
Japan

Italy

Germany

Philippines
Thailand
Norway
Malaysia
Denmark
Sweden
India
South Africa
Ireland
US
Taiwan
Korea
Mexico
Netherlands
UK
Germany
Australia
Belgium
Austria
Canada
Switzerland
France
Japan
Portugal
Italy
Spain
Greece

EM

Australia

14%

12%

DM

Germany

14%
France

DM

Belgium

25%

US

DM

Canada

30%

Netherlands

Spain

Australia

UK

Sweden

Mexico
India
South Africa
Philippines
Denmark
Sweden
Greece
US
Australia
UK
Netherlands
Malaysia
Spain
Switzerland
Canada
Thailand
France
Ireland
Germany
Belgium
Austria
Korea
Portugal
Italy
Taiwan
Japan

DM

US

Korea
South Africa

30%
25%
20%
15%
10%
5%
0%
-5%
-10%
-15%

UK

France

Australia

12 September 2013

Long-Term Asset Return Study: A Nominal Problem

Historical International Asset Returns

International equity return charts


Figure 123: Last 5 Years Annualised Equity Returns Nominal (left), Real (middle), USD (right)
EM

30%
25%
20%
15%
10%
5%
0%
-5%
-10%
-15%

Source: Deutsche Bank, GFD

Figure 124: Last 10 Years Annualised Equity Returns Nominal (left), Real (middle), USD (right)
EM
20%
18%
16%
14%
12%
10%
8%
6%
4%
2%
0%
-2%

Source: Deutsche Bank, GFD

Figure 125: Last 50 Years Annualised Equity Returns Nominal (left), Real (middle), USD (right)

EM
18%
16%
14%
12%
10%
8%
6%
4%
2%
0%

Source: Deutsche Bank, GFD

Figure 126: Last 100 Years Annualised Equity Returns Nominal (left), Real (middle), USD (right)

15%

Source: Deutsche Bank, GFD

Deutsche Bank AG/London

Deutsche Bank AG/London


4%

3%

4%

2%
2%
1%
1%

0%
0%
0%
0%

10%
10%
15%

8%
8%
10%
10%

6%
6%
5%
5%

4%
4%
0%
0%

2%
2%
-5%
-5%

0%
0%
-10%
-10%

-2%
-2%
-15%
-15%

DM

4%

1%

1%

0%

0%

3%

3%

-1%

-1%

2%

2%

-2%

-2%

1%

1%

-3%

-3%

0%

0%

4%
8%

3%
3%
7%
7%

2%
2%
6%
6%

5%

5%

4%

4%

DM

EM

EM

2%

DM

DM

Italy

4%
2%

France

9%
8%
7%
6%
5%
4%
3%
2%
1%
0%

3%

India

20%
18%
16%
14%
12%
10%
8%
6%
4%
2%
0%
4%
DM

Japan

6%
12%

Belgium

8%

6%
Philippines
Australia
Mexico
Ireland
Norway
Korea
South Africa
Canada
Sweden
Switzerland
Malaysia
Greece
Austria
UK
Belgium
Spain
Netherlands
Denmark
Thailand
France
Italy
Germany
Taiwan
US
Portugal
Japan
India

DM

South Africa

8%
5%

US

10%

15%

Norway

10%

Canada

5%

6%

8%

Australia

7%

EM

Australia
Canada
Denmark
France
Korea
Italy
Belgium
Norway
Thailand
South Africa
Spain
Portugal
Austria
Netherlands
US
Sweden
Germany
UK
Switzerland
Malaysia
Japan
India
Ireland

6%

12%
DM

Korea
Denmark
Austria
Japan
Germany
Belgium
Netherlands
Australia
Norway
France
Canada
Italy
Switzerland
UK
Spain
US
Sweden
Malaysia
South Africa
India
Ireland

-2%

10%
8%
6%
4%
2%
0%
-2%
-4%
-6%
-8%
-10%
-12%

Denmark

-2%

EM

Netherlands

0%

Italy

2%

0%

France

4%

2%

India

4%

Japan

6%

DM

Belgium

8%

6%

US

10%

8%

10%
8%
6%
4%
2%
0%
-2%
-4%
-6%
-8%
-10%
-12%

Norway

10%

Australia

EM
Philippines
Ireland
Mexico
Greece
Austria
Spain
Switzerland
Denmark
Belgium
Norway
France
Italy
Netherlands
Germany
Australia
Japan
South Africa
Korea
Sweden
Portugal
Canada
Malaysia
Thailand
UK
Taiwan
US
India

12%

Netherlands

EM
Australia
South Africa
Korea
Canada
Denmark
France
Thailand
Italy
Norway
Ireland
Belgium
Sweden
Spain
US
Austria
Netherlands
UK
Germany
Portugal
Malaysia
Japan
Switzerland
India

7%

14%

12%
EM

Korea
Denmark
Germany
Belgium
Austria
Canada
France
Malaysia
Japan
Australia
Netherlands
Italy
Norway
US
Sweden
UK
South Africa
Ireland
Switzerland
Spain
India

14%

South Africa

DM

EM

Canada

Philippines
Mexico
South Africa
Ireland
Greece
Austria
Belgium
Spain
Denmark
Netherlands
Korea
France
Australia
Italy
Norway
Germany
UK
Switzerland
Portugal
Canada
Sweden
Malaysia
Japan
US
Thailand
Taiwan
India

12%

Denmark

9%
8%
7%
6%
5%
4%
3%
2%
1%
0%

US

DM

India

20%
18%
16%
14%
12%
10%
8%
6%
4%
2%
0%

Norway

16%

14%
DM

Netherlands

16%

Canada

South Africa
Korea
Australia
India
Thailand
Italy
Spain
Canada
Denmark
Portugal
Norway
Ireland
France
Sweden
Belgium
US
UK
Austria
Netherlands
Malaysia
Germany
Switzerland
Japan

DM

Belgium

Korea
South Africa
Denmark
Italy
Spain
Australia
Ireland
UK
France
Canada
Belgium
Norway
Sweden
Austria
US
India
Netherlands
Germany
Malaysia
Japan
Switzerland

14%

France

Japan

Australia

Italy

Denmark

South Africa

12 September 2013

Long-Term Asset Return Study: A Nominal Problem

International 10 year government bond return charts


Figure 127: Last 5 Years Annualised 10 Year Government Bond Returns Nominal (left), Real (middle), USD (right)
EM
15%

10%
10%

5%
5%

0%
0%

-5%
-5%

Source: Deutsche Bank, GFD

Figure 128: Last 10 Years Annualised 10 Year Government Bond Returns Nominal (left), Real (middle), USD (right)
10%

EM
12%

10%

8%
8%

6%
6%

4%
4%

2%
2%

0%
0%

Source: Deutsche Bank, GFD

Figure 129: Last 50 Years Annualised 10 Year Government Bond Returns Nominal (left), Real (middle), USD (right)
EM
15%

Source: Deutsche Bank, GFD

Figure 130: Last 100 Years Annualised 10 Year Government Bond Returns Nominal (left), Real (middle), USD (right)

EM
8%

Source: Deutsche Bank, GFD

Page 77

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

International Equity minus Bond return charts


Figure 131: Last 5 Yrs Annualised Equity-Bond Return
30%
25%
20%
15%
10%
5%
0%
-5%
-10%
-15%
-20%

30%
25%
20%
15%
10%
5%
0%
-5%
-10%
-15%
-20%

EM

8%

6%

4%

4%

2%

2%

0%

0%

-2%

-2%

-4%

-4%

-6%

-6%

-8%

-8%

Source: Deutsche Bank, GFD

DM

6%

EM

5%

5%

4%

4%

3%

3%

2%

2%

1%

1%

0%

0%

US

Italy

Japan

Germany

France

Belgium

6%

France

6%
5%
4%
3%
2%
1%
0%
-1%
-2%
-3%
-4%

EM

US

Spain

Australia

Korea

Netherlands

UK

Sweden

South Africa

DM

Canada

6%
5%
4%
3%
2%
1%
0%
-1%
-2%
-3%
-4%

Figure 134: Last 100 Yrs Annualised Equity-Bond Return

Australia

Figure 133: Last 50 Yrs Annualised Equity-Bond Return

Page 78

8%

EM

India
Switzerland
Sweden
Denmark
Netherlands
US
Malaysia
UK
South Africa
Germany
Spain
Austria
Australia
France
Canada
Ireland
Belgium
Thailand
Portugal
Korea
Italy
Japan

Source: Deutsche Bank, GFD

Source: Deutsche Bank, GFD

DM

6%

Thailand
India
Malaysia
Sweden
Taiwan
Norway
Philippines
US
Denmark
UK
Korea
Netherlands
South Africa
Ireland
Germany
Canada
Switzerland
Australia
Belgium
Japan
France
Mexico
Austria
Portugal
Italy
Spain
Greece

DM

Figure 132: Last 25 Yrs Annualised Equity-Bond Return

Source: Deutsche Bank, GFD

Deutsche Bank AG/London

2000-2009

2010-2013

10.5%

1990-1999

10.3%

1980-1989

17.1%

1970-1979

Denmark

1960-1969

9.3%

1950-1959

8.0%

1940-1949

7.1%

1930-1939

10.1%

1920-1929

Canada

1910-1919

8.8%

1900-1909

7.0%

1890-1899

6.8%

1880-1889

10.9%

1870-1879

Belgium

1860-1869

11.9%

1850-1859

11.8%

1840-1849

Since 1900

11.6%

6.6%

1830-1839

Last 100yrs

9.5%

5.4%

1820-1829

Last 50yrs

9.0%

10.2%

1810-1819

Last 25yrs

11.2%

Austria

1800-1809

Last 10yrs

Australia

Since 1800

Last 5yrs

RETURNS BY DECADE

7.9%

13.6%

9.7%

15.4%

10.2%

10.1%

15.3%

14.0%

8.6%

17.7%

11.0%

8.9%

5.6%

EQUITY

France

8.4%

9.8%

5.6%

7.4%

9.3%

11.2%

10.5%

12.1%

7.8%

7.1%

7.4%

5.4%

5.3%

Greece

-10.6%

-6.3%

10.1%

Ireland

Germany

7.7%

10.0%

5.1%

13.3%

6.5%

16.3%

1.4%

7.4%

2.2%

3.4%

7.2%

20.6%

11.4%

1.8%

7.0%

10.0%

10.4%

12.2%

10.6%

5.6%

4.6%

7.9%

23.8%

11.1%

6.7%

12.9%
5.4%

5.6%

8.1%

16.9%

-1.5%

20.7%

24.0%

4.5%

6.8%

21.9%

14.3%

-0.3%

5.6%

-18.7%

18.1%

4.5%

-6.0%

25.8%

6.0%

2.2%

15.9%

12.1%

-0.9%

9.0%

36.2%

38.3%

-7.2%

-18.1%

15.0%

0.8%

7.3%

14.4%

-2.8%

11.5%

Italy

3.0%

0.6%

4.7%

7.0%

6.5%

30.4%

23.5%

3.7%

-3.0%

28.0%

12.6%

-1.5%

-1.7%

Japan

7.2%

2.2%

-1.8%

6.9%

14.2%

15.9%

33.9%

13.0%

12.3%

21.3%

-4.3%

-5.0%

7.1%

Netherlands

13.6%

6.5%

9.1%

10.0%

6.1%

4.4%

21.5%

19.4%

-1.6%

8.1%

Norway

18.1%

12.3%

Portugal

3.7%

2.8%

11.1%

0.6%

-3.8%

Spain

7.7%
5.1%

1.3%

3.5%

8.7%

11.4%

Sweden

16.8%

10.4%

10.5%

13.3%

Switzerland

10.0%

6.4%

8.6%

UK

12.8%

8.1%

9.2%

11.8%

9.9%

8.7%

6.9%

8.1%

5.4%

4.8%

4.3%

4.8%

3.8%

4.4%

4.9%

5.5%

3.0%

0.6%

1.5%

9.5%

1.9%

8.9%

17.2%

US

14.9%

6.0%

9.7%

9.6%

9.8%

9.4%

8.5%

11.1%

4.9%

6.9%

5.3%

7.8%

1.6%

18.3%

7.7%

5.7%

5.4%

9.9%

4.3%

14.8%

-0.5%

9.0%

Australia

4.5%

6.4%

9.7%

8.7%

6.5%

5.9%

5.3%

5.6%

6.2%

5.2%

1.6%

0.7%

6.1%

5.9%

3.9%

Austria

5.5%

5.2%

6.6%

7.4%

Belgium

5.0%

5.0%

7.3%

7.9%

6.0%

5.5%

Canada

2.6%

5.1%

8.2%

8.1%

6.1%

5.5%

Denmark

4.8%

5.3%

7.8%

10.0%

7.7%

7.1%

France

4.5%

5.0%

7.5%

8.2%

6.1%

5.6%

Germany

4.2%

5.2%

5.9%

6.9%

Greece

5.6%

4.4%

Ireland

6.4%

5.1%

7.6%

8.6%

Italy

4.5%

4.4%

8.7%

9.7%

7.0%

6.0%

Japan

1.8%

1.9%

3.9%

6.4%

6.4%

6.2%

Netherlands

4.7%

5.0%

6.6%

7.0%

5.7%

5.2%

Norway

4.2%

4.1%

7.6%

7.8%

5.6%

5.1%

Portugal

2.7%

3.8%

7.8%

Spain

4.9%

4.9%

8.3%

8.8%

Sweden

2.4%

4.3%

7.4%

7.7%

Switzerland

2.9%

3.3%

4.4%

4.6%

UK

2.9%

4.6%

7.2%

8.6%

US

1.6%

4.7%

7.3%

7.1%

3.5%

-0.2%

10.5%

13.3%

19.1%

-1.2%

27.4%

18.7%

4.3%

-3.1%

16.3%

8.1%

6.7%

32.4%

19.0%

1.3%

9.9%

2.0%

10.6%

16.0%

1.1%

6.9%

8.3%

10.2%

23.9%

14.9%

1.6%

8.9%

19.3%

7.8%

5.8%

17.5%

18.2%

-0.9%

12.2%

3.1%

4.2%

6.9%

12.4%

12.9%

6.7%

7.8%

8.1%

4.6%

8.1%

8.7%

8.5%

5.8%

6.0%

BOND

Page 79

Source: Deutsche Bank, GFD

3.3%

6.2%

19.8%

6.4%

12.9%

6.2%

5.6%

4.6%

5.3%

3.5%

2.8%

0.4%

5.4%

3.6%

4.9%

4.3%

4.4%

6.3%

12.0%

10.4%

6.0%

5.2%

5.2%

7.2%

6.9%

3.4%

2.2%

2.2%

5.8%

5.2%

3.5%

1.5%

3.7%

6.8%

13.4%

10.7%

6.8%

4.4%

4.0%

3.6%

31.5%

-19.8%

6.0%

5.8%

3.2%

3.7%

0.7%

6.2%

5.4%

8.3%

4.5%

4.1%

10.1%

18.9%

11.2%

6.1%

5.6%

1.8%

2.9%

6.7%

4.9%

6.0%

4.6%

4.3%

3.1%

-0.6%

6.6%

3.8%

2.8%

4.8%

4.3%

6.1%

14.9%

10.7%

5.9%

5.1%

7.5%

-17.3%

5.9%

5.8%

8.1%

8.2%

6.9%

5.8%

5.4%

5.3%

6.9%

11.4%

8.8%

3.2%

5.4%

5.6%

2.4%

4.8%

5.1%

9.1%

6.2%

5.5%

2.8%

7.5%

4.0%

6.3%

6.3%

4.9%

3.4%

5.5%

18.4%

10.6%

5.1%

7.7%

12.6%

5.1%

-1.4%

1.5%

4.2%

5.3%

4.7%

5.3%

5.0%

6.5%

17.3%

14.3%

5.8%

4.0%

7.0%

5.7%

6.1%

2.9%

6.2%

5.7%

5.5%

6.0%

12.3%

6.8%

9.2%

7.2%

1.8%

2.2%

6.3%

6.2%

2.6%

2.6%

-1.2%

7.1%

3.9%

7.8%

2.6%

3.0%

7.5%

9.6%

8.7%

5.9%

5.0%

6.5%

5.0%

5.0%

2.6%

2.0%

3.3%

3.9%

4.1%

2.5%

3.7%

6.2%

11.9%

11.7%

5.4%

5.0%

19.5%

10.9%

6.7%

2.5%

6.0%

5.2%

4.4%

3.7%

5.5%

3.9%

1.6%

2.5%

5.5%

5.1%

4.8%

6.0%

16.3%

12.1%

5.6%

5.1%

4.7%

6.2%

2.6%

3.8%

5.9%

11.7%

11.8%

5.6%

3.3%

4.2%

4.1%

2.7%

2.9%

5.8%

3.9%

5.9%

4.3%

2.6%

3.3%

3.4%

5.0%

9.4%

14.0%

10.2%

5.4%

3.9%

2.7%

0.4%

2.8%

6.1%

12.8%

8.0%

6.6%

4.6%

4.0%

12 September 2013

Figure 135: Developed Market Nominal Annualised Equity and Bond Returns

Long-Term Asset Return Study: A Nominal Problem

Deutsche Bank AG/London

International return tables

1990-1999

2000-2009

2010-2013

9.5%

12.3%

4.2%

14.6%

11.3%

4.5%

8.4%

11.2%

-1.4%

1980-1989

8.3%

1970-1979

8.3%

1960-1969

15.1%

1950-1959

5.0%

1940-1949

5.7%

1930-1939

5.2%

1920-1929

8.2%

1910-1919

Canada

1900-1909

4.8%

1890-1899

4.7%

1880-1889

4.4%

1870-1879

8.7%

1860-1869

Belgium

1850-1859

7.7%

1840-1849

Since 1900

7.3%

1830-1839

Last 100yrs

6.0%

4.4%

1820-1829

Last 50yrs

6.5%

3.3%

1810-1819

Last 25yrs

6.3%

8.1%

1800-1809

Last 10yrs

8.9%

Austria

Since 1800

Last 5yrs
Australia

8.6%

8.6%

5.6%

3.3%

0.5%

12.2%

-1.0%

5.5%

0.0%

0.6%

0.1%

15.2%

9.1%

-0.3%

4.4%

7.1%

2.7%

5.6%

8.3%

3.5%

2.7%

-1.6%

16.3%

8.8%

4.7%

10.8%
3.5%

EQUITY

Denmark
France

3.7%

8.0%

3.7%

5.4%

4.6%

2.8%

3.1%

10.3%

5.8%

5.0%

4.5%

-20.1%

-17.6%

Greece

-11.9%

-8.4%

3.8%

Ireland

Germany

5.3%
6.1%

9.6%

5.2%

3.6%

-3.3%

10.6%

8.3%

-4.3%

-8.8%

17.4%

0.6%

-2.2%

14.1%

12.2%

-2.1%

-32.6% -89.3%

6.5%

-9.5%

23.1%

3.5%

-2.6%

12.8%

9.6%

-2.5%

7.1%

14.3%

25.4%

-10.1%

-19.0%

14.6%

-0.5%

4.9%

11.8%

-5.2%

10.3%

Italy

1.3%

-1.4%

1.6%

0.6%

6.1%

-12.8%

19.9%

0.2%

-14.3%

15.7%

8.3%

-3.7%

-3.5%

Japan

7.5%

2.2%

-2.3%

3.8%

10.4%

-25.1%

30.2%

7.1%

3.1%

18.5%

-5.3%

-4.7%

7.0%

Netherlands

11.2%

4.5%

6.7%

6.1%

2.0%

-2.6%

18.3%

16.6%

-3.7%

5.4%

Norway

16.5%

10.4%

Portugal

2.0%

0.8%

1.0%

5.3%

-1.9%

-5.8%

Spain

-0.5%

1.1%

5.1%

4.0%

Sweden

15.3%

8.8%

7.9%

8.1%

Switzerland

10.1%

5.8%

7.1%

UK

9.7%

5.3%

6.1%

5.7%

5.5%

4.8%

4.9%

4.6%

6.3%

7.2%

3.7%

6.9%

3.7%

3.9%

5.4%

5.9%

3.0%

-0.2%

-5.8%

12.9%

1.4%

5.9%

12.5%

US

12.5%

3.5%

6.8%

5.3%

6.4%

6.2%

6.7%

11.1%

4.6%

9.1%

3.2%

10.8%

0.1%

13.6%

10.2%

5.7%

5.2%

7.4%

-2.8%

15.9%

1.6%

3.4%

Australia

2.3%

3.7%

6.7%

3.3%

2.2%

1.9%

5.5%

5.9%

6.8%

0.5%

-4.3%

5.3%

7.0%

-1.4%

Austria

3.5%

3.1%

4.3%

3.9%

Belgium

2.9%

2.6%

5.0%

3.9%

0.3%

Canada

0.9%

3.2%

6.0%

3.8%

2.9%

Denmark

2.9%

3.4%

5.6%

4.9%

3.2%

3.0%

France

2.8%

3.2%

5.5%

3.6%

-1.9%

-1.5%

Germany

2.5%

3.3%

3.9%

4.0%

Greece

4.1%

2.1%

Ireland

6.0%

3.7%

5.1%

2.5%

Italy

2.8%

2.4%

5.5%

3.2%

-2.7%

-2.5%

Japan

2.1%

1.9%

3.4%

3.3%

-0.9%

-0.6%

Netherlands

2.5%

3.0%

4.2%

3.2%

2.2%

1.9%

Norway

2.8%

2.4%

5.4%

2.9%

1.5%

1.4%

Portugal

1.0%

1.8%

3.6%

Spain

3.1%

2.4%

4.7%

1.5%

Sweden

1.1%

2.8%

4.9%

2.7%

Switzerland

3.0%

2.7%

3.0%

1.9%

UK

0.0%

1.9%

4.1%

2.7%

US

-0.6%

2.2%

4.4%

2.9%

6.5%

8.4%

-0.9%

6.5%

7.1%

12.6%

-13.9%

16.0%

14.1%

1.3%

-5.0%

11.3%

4.1%

-2.0%

23.0%

15.6%

-0.6%

9.0%

-2.8%

7.0%

13.6%

0.2%

7.1%

4.5%

-2.6%

15.9%

11.0%

-0.3%

5.8%

16.7%

5.1%

-1.5%

11.8%

14.8%

-3.4%

10.1%

-3.1%

1.7%

-2.9%

3.8%

10.4%

3.5%

5.6%

3.2%

1.2%

2.0%

4.8%

5.9%

3.9%

3.7%

BOND

Deutsche Bank AG/London

Source: Deutsche Bank, GFD

0.3%

4.4%

4.3%

3.9%

5.9%

4.1%

0.8%

3.0%

6.7%

0.2%

-0.2%

4.3%

-6.9%

2.2%

1.6%

-0.8%

6.9%

8.2%

3.9%

2.6%

6.7%

7.1%

-1.0%

-0.9%

1.0%

-0.7%

6.8%

8.4%

4.6%

2.6%

29.7%

-20.2%

6.2%

6.4%

3.3%

2.6%

-7.7%

7.3%

3.4%

3.7%

0.6%

-1.4%

0.5%

11.7%

9.0%

4.1%

3.6%

6.2%

4.2%

5.5%

4.7%

4.5%

2.7%

-11.1%

-1.3%

0.8%

-22.4%

-0.8%

0.4%

-2.8%

7.5%

8.7%

4.0%

3.2%

9.5%

-20.4%

3.6%

3.4%

3.0%

5.3%

4.5%

4.0%

3.5%

2.0%

5.7%

9.8%

10.6%

4.4%

5.4%

2.5%

1.6%

3.3%

9.1%

5.9%

7.6%

0.7%

10.5%

2.4%

2.0%

1.0%

1.2%

-0.9%

-6.7%

8.8%

8.0%

2.5%

6.6%

13.4%

5.3%

-2.1%

-8.7%

-4.0%

4.9%

-30.0%

2.1%

1.5%

-5.8%

6.1%

9.9%

3.4%

2.2%

10.5%

-0.9%

2.6%

-5.8%

10.3%

2.1%

-31.8%

3.0%

6.4%

-2.0%

6.7%

6.1%

2.1%

2.1%

6.1%

8.1%

3.4%

0.7%

-7.7%

9.2%

5.3%

0.0%

-1.2%

-0.9%

0.3%

6.7%

6.2%

3.6%

2.4%

6.7%

5.3%

4.2%

1.7%

-8.6%

7.9%

2.8%

0.0%

-2.4%

0.1%

-2.1%

3.4%

9.0%

3.4%

3.8%

2.2%

5.1%

4.1%

0.4%

9.5%

1.9%

3.7%

6.0%

5.5%

3.8%

-0.7%

-4.5%

6.5%

-0.7%

-0.9%

-7.6%

5.9%

7.8%

2.6%

3.1%

4.0%

2.4%

-1.8%

-0.1%

-2.7%

3.8%

8.5%

3.7%

2.4%

5.5%

-0.4%

1.5%

-0.3%

0.8%

0.6%

3.7%

3.3%

2.8%

0.5%

-0.7%

1.3%

-3.2%

6.6%

6.5%

3.4%

0.9%

-2.5%

-1.8%

0.2%

-1.2%

7.3%

4.9%

4.0%

2.4%

6.1%

12 September 2013

RETURNS BY DECADE

Long-Term Asset Return Study: A Nominal Problem

Page 80

Figure 136: Developed Market Real Annualised Equity and Bond Returns

1980-1989

1990-1999

2000-2009

2010-2013

11.4%

1970-1979

10.9%

1960-1969

15.8%

1950-1959

Denmark

1940-1949

9.4%

1930-1939

8.5%

1920-1929

9.3%

1910-1919

13.3%

1900-1909

Canada

1890-1899

9.9%

1880-1889

7.8%

1870-1879

7.3%

1860-1869

9.7%

1850-1859

Belgium

1840-1849

Since 1900
11.0%

1830-1839

Last 100yrs
10.7%

1820-1829

Last 50yrs
11.1%

7.4%

1810-1819

Last 25yrs
9.7%

6.0%

1800-1809

Last 10yrs
10.9%

9.0%

Since 1800

Last 5yrs
16.5%

Austria

8.0%

13.6%

7.0%

18.2%

5.6%

6.4%

15.3%

14.0%

8.5%

13.8%

9.0%

12.4%

5.4%

14.6%

16.8%

0.0%

11.3%

0.2%

3.4%

13.5%

17.8%

10.1%

5.4%

4.9%

8.7%

9.5%

12.3%

8.1%

9.0%

4.5%

11.5%

21.3%

9.8%

10.5%

10.6%
3.3%

EQUITY
Australia

France

8.5%

8.5%

6.1%

8.3%

9.3%

6.3%

6.2%

10.9%

8.3%

7.8%

9.5%

-21.1%

-18.2%

Greece

-11.6%

-5.8%

7.7%

Ireland

13.7%

1.3%

7.8%

1.8%

1.1%

4.2%

Germany

Italy
Japan

10.0%

5.1%

5.6%

7.5%

-6.9%

-1.7%

19.9%

3.2%

10.3%

17.6%

12.9%

3.3%

-36.5% -90.5%

0.3%

10.0%

-29.1%

25.9%

7.3%

10.3%

16.1%

10.5%

2.7%

6.9%

17.5%

28.5%

-4.2%

-19.7%

12.2%

0.7%

9.3%

8.0%

2.1%

-3.7%

5.2%

6.1%

-7.6%

23.6%

3.6%

-5.4%

22.3%

6.1%

-25.6%

33.9%

13.0%

16.9%

27.7%

-0.9%

-4.1%

5.7%

6.5%

11.3%

21.4%

17.7%

1.9%

5.9%

7.9%

4.2%

-5.7%

5.5%

3.1%

-0.9%

9.7%

Netherlands

12.3%

7.0%

9.9%

11.7%

Norway

21.1%

13.3%

Portugal

2.5%

3.3%

Spain

5.7%
7.6%

15.1%

6.3%
4.9%

0.1%

4.0%

8.2%

9.8%

Sweden

20.7%

11.3%

10.1%

12.7%

3.8%

17.3%

-0.7%

21.2%

13.9%

8.0%

-5.0%

6.0%

-1.5%

8.2%

16.3%

8.1%

9.1%

27.2%

15.4%

3.0%

12.1%

Switzerland

13.1%

9.5%

10.7%

UK

14.2%

6.5%

8.5%

10.5%

8.6%

7.6%

6.4%

8.1%

5.6%

5.5%

4.3%

4.8%

3.9%

6.4%

2.9%

5.5%

3.1%

0.6%

12.7%

11.0%

15.6%

5.6%

9.8%

-1.1%

12.4%

-0.2%

5.2%

17.2%

6.7%

9.3%

20.0%

14.9%

1.6%

US

14.9%

6.0%

9.7%

9.6%

9.8%

9.4%

8.5%

11.1%

4.9%

6.9%

5.3%

7.8%

1.6%

18.3%

7.7%

5.7%

5.4%

7.8%

9.9%

4.3%

14.8%

-0.5%

9.0%

19.3%

7.8%

5.8%

17.5%

18.2%

-0.9%

12.2%

5.4%

5.0%

5.2%

5.6%

6.3%

5.3%

1.6%

-1.8%

8.6%

1.5%

0.3%

3.1%

4.2%

6.8%

8.7%

10.9%

10.1%

8.4%

8.1%

4.7%

16.3%

9.2%

7.0%

9.6%

3.5%

BOND
Australia

10.0%

8.5%

10.0%

8.3%

Austria

4.0%

5.6%

7.3%

9.3%

Belgium

3.5%

5.3%

8.1%

8.9%

4.2%

3.8%

Canada

5.7%

7.3%

8.8%

8.2%

6.0%

5.5%

Denmark

3.2%

5.7%

8.6%

10.4%

7.2%

6.7%

France

3.0%

5.4%

8.3%

8.2%

1.3%

1.4%

Germany

2.7%

5.6%

6.6%

9.0%

Greece

4.1%

4.7%

Ireland

8.3%

3.4%

0.3%

-11.7%

Italy

3.0%

4.8%

8.1%

7.8%

1.1%

1.0%

Japan

0.0%

2.7%

4.9%

9.2%

2.3%

2.6%

Netherlands

3.2%

5.3%

7.3%

8.6%

6.1%

5.5%

Norway

7.1%

5.1%

8.0%

8.2%

5.1%

4.7%

Portugal

1.2%

4.1%

7.5%

Spain

3.4%

5.2%

7.8%

7.2%

Sweden

5.6%

5.1%

7.0%

7.1%

Switzerland

5.4%

6.1%

6.4%

7.8%

UK

3.7%

3.0%

6.4%

7.3%

US

1.6%

4.7%

7.3%

7.1%

Source: Deutsche Bank, GFD

3.2%

4.1%

13.3%

1.8%

2.8%

6.3%

7.0%

5.5%

4.6%

5.3%

3.6%

2.8%

-6.8%

-6.4%

5.5%

-0.3%

4.3%

4.5%

12.6%

9.4%

9.2%

9.8%

2.7%

8.3%

4.2%

7.0%

3.3%

2.0%

1.4%

6.5%

4.1%

3.6%

3.2%

2.4%

5.9%

13.5%

8.2%

10.2%

4.4%

6.0%

5.8%

3.2%

3.7%

-2.6%

9.9%

2.0%

5.2%

4.5%

3.2%

13.9%

16.5%

10.0%

9.9%

3.1%

6.0%

4.6%

4.3%

3.1%

-7.8%

-2.0%

-1.9%

-16.3%

1.3%

3.0%

9.6%

10.8%

9.4%

9.7%

2.6%

13.2%

-37.6%

5.9%

7.1%

16.7%

8.4%

5.4%

9.6%

2.9%

8.7%

4.3%

4.9%

10.6%

9.1%

3.2%

5.3%

6.0%

2.2%

4.8%

5.1%

9.1%

6.2%

5.5%

2.8%

7.5%

4.0%

6.3%

-1.0%

-18.1%

-57.5%

-1.2%

6.4%

-3.4%

2.0%

6.3%

14.0%

4.5%

-0.7%

-7.5%

0.5%

4.9%

-25.8%

5.3%

4.9%

3.9%

12.1%

9.6%

9.6%

1.6%

5.4%

0.9%

6.1%

2.9%

6.1%

-1.9%

-32.3%

6.0%

12.3%

11.2%

14.9%

11.0%

2.8%

0.6%

6.1%

6.2%

2.6%

2.6%

-2.0%

7.9%

6.8%

0.5%

2.7%

3.4%

14.7%

9.6%

7.3%

9.6%

2.5%

4.5%

5.0%

5.1%

2.6%

-0.8%

6.2%

2.2%

-0.9%

2.5%

3.7%

10.2%

8.7%

9.5%

8.9%

3.8%

7.2%

7.7%

10.5%

0.1%

6.9%

5.2%

7.5%

3.7%

5.5%

3.9%

1.6%

2.5%

5.5%

-3.7%

3.2%

6.6%

10.6%

7.6%

9.4%

2.6%

3.3%

4.0%

2.6%

3.8%

8.3%

7.3%

8.3%

7.5%

5.0%

5.7%

4.5%

2.7%

2.9%

16.9%

4.3%

5.6%

8.9%

5.0%

-0.2%

3.4%

3.4%

8.6%

10.4%

10.2%

5.4%

2.3%

2.7%

0.4%

2.8%

6.1%

12.8%

8.0%

6.6%

4.6%

4.0%

12 September 2013

RETURNS BY DECADE

Long-Term Asset Return Study: A Nominal Problem

Deutsche Bank AG/London

Figure 137: Developed Market USD Annualised Equity and Bond Returns

Page 81

1970-1979

1980-1989

1990-1999

2000-2009

2010-2013

1960-1969

1950-1959

1940-1949

1930-1939

1920-1929

1910-1919

1900-1909

1890-1899

1880-1889

1870-1879

1860-1869

1850-1859

1840-1849

1830-1839

1820-1829

1810-1819

1800-1809

Since 1800

Since 1900

Last 100yrs

5.1%

12.8%

5.6%

7.8%

10.3%

26.1%

35.9%

18.3%

6.8%

10.7%

9.3%

5.1%

17.4%

13.9%

14.7%

12.9%

3.9%

0.9%

3.0%

27.3%

-2.4%

8.7%

14.6%

Last 50yrs

9.9%

Last 25yrs

2.1%

4.6%

Last 10yrs

14.9%

29.2%

Last 5yrs

23.3%
40.7%

India

16.0%

13.0%

16.1%

Korea

14.2%

10.4%

6.6%

Malaysia

17.3%

10.9%

8.9%

Mexico

13.7%

18.3%

Philippines

25.4%

16.3%

South Africa

15.4%

16.6%

15.5%

Taiwan

14.4%

5.8%

4.6%

Thailand

24.0%

9.0%

8.0%

India

-0.3%

2.6%

9.5%

7.1%

Korea

4.5%

5.2%

10.4%

17.9%

Malaysia

2.3%

4.7%

6.5%

6.8%

Mexico

8.7%

9.4%

14.5%

8.5%

11.9%

13.3%

16.3%

13.9%

South Africa

6.7%

8.5%

Taiwan

0.6%

2.7%

Thailand

0.8%

6.0%

EQUITY

21.1%

16.7%

16.0%

24.1%

BOND

Philippines

14.1%

11.5%

5.7%

7.9%

5.3%

5.0%

5.7%

6.3%

5.3%

5.5%

4.6%

3.0%

5.1%

7.4%

4.2%

4.6%

4.1%

5.6%

3.1%

3.7%

2.3%

4.8%

0.5%

2.0%

5.9%

4.8%

8.0%

4.8%

4.2%

3.5%

3.0%

5.3%

4.2%

4.9%

4.4%

14.1%

8.5%

3.0%

28.5%

27.2%

22.1%

15.6%

7.7%

5.6%

6.8%

9.0%

7.6%

6.1%

4.1%

4.9%

7.4%

8.8%

15.2%

13.6%

17.5%

13.7%

12.1%

8.7%

6.9%

-0.1%

7.9%

3.4%

Source: Deutsche Bank, GFD

Figure 139: Emerging Market Real Annualised Equity and Bond Returns
1970-1979

1980-1989

1990-1999

2000-2009

2010-2013

1960-1969

8.1%

8.6%

1.0%

0.0%

1.5%

23.2%

6.9%

4.7%

21.1%

-6.9%

6.1%

14.7%

India

-9.4%

-5.1%

1.6%

-0.8%

Korea

2.0%

2.4%

6.1%

9.2%

Malaysia

0.5%

2.2%

3.6%

3.4%

Mexico

5.2%

5.2%

9.1%

5.0%

Philippines

8.3%

8.1%

10.5%

10.5%

Thailand

1950-1959

4.2%

2.7%

Taiwan

1940-1949

7.6%

4.3%

South Africa

1930-1939

10.7%

13.1%

21.3%

1920-1929

10.4%

Philippines

1910-1919

13.9%

13.8%

1900-1909

-0.2%

10.0%

1890-1899

0.5%

Mexico

1880-1889

4.1%

5.9%

1870-1879

3.4%

11.0%

8.2%

1860-1869

12.7%

15.3%

1850-1859

13.7%

Malaysia

1840-1849

8.1%

13.8%

2.4%

1830-1839

5.5%

7.7%

7.4%

1820-1829

1.7%

4.5%

11.4%

1810-1819

2.6%

9.0%

5.4%

Korea

Last 50yrs

6.5%

Last 25yrs

-6.3%

-0.9%

Last 10yrs

8.2%

20.3%

Last 5yrs

12.7%
22.3%

India

1800-1809

Since 1800

Since 1900

Last 100yrs

RETURNS BY DECADE

EQUITY

12.2%

7.4%

5.4%

8.3%

Deutsche Bank AG/London

BOND

South Africa

2.0%

3.0%

Taiwan

-0.5%

1.2%

Thailand

0.1%

4.0%

Source: Deutsche Bank, GFD

6.3%

5.5%

2.6%

0.2%

2.3%

0.4%

2.4%

3.1%

3.5%

1.3%

6.0%

-4.5%

-3.0%

5.3%

4.4%

11.4%

5.3%

-5.2%

-1.2%

1.6%

1.6%

-1.6%

-2.6%

-4.0%

4.2%

2.3%

-5.4%

13.4%

10.5%

13.6%

9.5%

4.5%

3.1%

1.2%

5.4%

3.6%

3.9%

2.0%

2.2%

-2.4%

0.5%

8.1%

7.5%

8.5%

5.7%

4.5%

5.9%

-1.5%

5.3%

3.5%

12 September 2013

RETURNS BY DECADE

Long-Term Asset Return Study: A Nominal Problem

Page 82

Figure 138: Emerging Market Nominal Annualised Equity and Bond Returns

3.7%

1.9%

5.7%

8.1%

12.9%

Malaysia

4.1%

6.6%

5.8%

6.7%

Mexico

9.9%

7.8%

2010-2013

-0.1%

6.6%

2000-2009

-4.2%

Korea

1990-1999

India

1980-1989

6.9%

1970-1979

4.4%

11.3%

1960-1969

7.2%

25.9%

1950-1959

16.5%

Thailand

1940-1949

Taiwan

1930-1939

8.9%

1920-1929

7.5%

11.7%

1910-1919

18.9%

13.5%

1900-1909

27.0%

South Africa

1890-1899

Philippines

1880-1889

17.5%

1870-1879

16.2%

1860-1869

14.2%

1850-1859

Mexico

1840-1849

8.0%

1830-1839

12.5%

1820-1829

18.5%

1810-1819

Malaysia

1800-1809

4.5%

Since 1800

9.5%

11.2%

Since 1900

9.0%

17.1%

Last 100yrs

Last 25yrs

Last 50yrs

Last 10yrs

9.2%

Korea

Last 5yrs
India

12.2%

14.1%

34.3%

24.9%

-0.7%

9.6%

10.4%

2.1%

8.9%

19.8%

14.5%

EQUITY

16.0%

10.6%

14.3%

2.3%

3.6%

11.0%

4.2%

12.6%

2.0%

0.7%

24.3%

-6.0%

10.0%

BOND

Philippines

14.1%

16.2%

South Africa

5.8%

4.4%

Taiwan

2.4%

4.0%

Thailand

3.0%

8.6%

Source: Deutsche Bank, GFD

2.6%

2.6%

6.8%

3.7%

3.3%

4.7%

2.1%

2.7%

2.6%

2.3%

3.7%

3.8%

6.0%

0.5%

2.9%

-0.5%

4.3%

-3.2%

3.8%

7.8%

7.3%

21.4%

18.0%

9.8%

7.5%

10.5%

6.7%

3.9%

7.3%
10.9%
14.6%

7.8%

5.8%

4.5%

4.5%

2.6%

5.6%

3.8%

4.8%

-0.6%

7.6%

2.6%

0.0%

5.3%

4.9%

5.9%

3.0%

7.6%

10.9%

9.5%

10.1%
6.7%

7.9%

9.1%

12 September 2013

RETURNS BY DECADE

Long-Term Asset Return Study: A Nominal Problem

Deutsche Bank AG/London

Figure 140: Emerging Market USD Annualised Equity and Bond Returns

Page 83

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

Appendix 1
Important Disclosures
Additional information available upon request
For disclosures pertaining to recommendations or estimates made on securities other than the primary subject of this
research, please see the most recently published company report or visit our global disclosure look-up page on our
website at http://gm.db.com/ger/disclosure/DisclosureDirectory.eqsr

Analyst Certification
The views expressed in this report accurately reflect the personal views of the undersigned lead analyst(s). In addition,
the undersigned lead analyst(s) has not and will not receive any compensation for providing a specific recommendation
or view in this report. Jim Reid/Nick Burns/Seb Barker

Page 84

Deutsche Bank AG/London

12 September 2013
Long-Term Asset Return Study: A Nominal Problem

(a) Regulatory Disclosures


(b) 1. Important Additional Conflict Disclosures
Aside from within this report, important conflict disclosures can also be found at https://gm.db.com/equities under the
Disclosures Lookup and Legal tabs. Investors are strongly encouraged to review this information before investing.

(c) 2. Short-Term Trade Ideas


Deutsche Bank equity research analysts sometimes have shorter-term trade ideas (known as SOLAR ideas) that are
consistent or inconsistent with Deutsche Banks existing longer term ratings. These trade ideas can be found at the
SOLAR link at http://gm.db.com.

(d) 3. Country-Specific Disclosures


Australia and New Zealand: This research, and any access to it, is intended only for wholesale clients within the
meaning of the Australian Corporations Act and New Zealand Financial Advisors Act respectively.
Brazil: The views expressed above accurately reflect personal views of the authors about the subject company(ies) and
its(their) securities, including in relation to Deutsche Bank. The compensation of the equity research analyst(s) is
indirectly affected by revenues deriving from the business and financial transactions of Deutsche Bank. In cases where
at least one Brazil based analyst (identified by a phone number starting with +55 country code) has taken part in the
preparation of this research report, the Brazil based analyst whose name appears first assumes primary responsibility for
its content from a Brazilian regulatory perspective and for its compliance with CVM Instruction # 483.
EU
countries:
Disclosures
relating
to
our
obligations
under
MiFiD
can
be
found
at
http://www.globalmarkets.db.com/riskdisclosures.
Japan: Disclosures under the Financial Instruments and Exchange Law: Company name - Deutsche Securities Inc.
Registration number - Registered as a financial instruments dealer by the Head of the Kanto Local Finance Bureau
(Kinsho) No. 117. Member of associations: JSDA, Type II Financial Instruments Firms Association, The Financial Futures
Association of Japan, Japan Investment Advisers Association. This report is not meant to solicit the purchase of specific
financial instruments or related services. We may charge commissions and fees for certain categories of investment
advice, products and services. Recommended investment strategies, products and services carry the risk of losses to
principal and other losses as a result of changes in market and/or economic trends, and/or fluctuations in market value.
Before deciding on the purchase of financial products and/or services, customers should carefully read the relevant
disclosures, prospectuses and other documentation. Moodys, Standard & Poors, and Fitch mentioned in this
report are not registered credit rating agencies in Japan unless Japan or Nippon is specifically designated in the
name of the entity.
Malaysia: Deutsche Bank AG and/or its affiliate(s) may maintain positions in the securities referred to herein and may
from time to time offer those securities for purchase or may have an interest to purchase such securities. Deutsche Bank
may engage in transactions in a manner inconsistent with the views discussed herein.
Russia: This information, interpretation and opinions submitted herein are not in the context of, and do not constitute,
any appraisal or evaluation activity requiring a license in the Russian Federation.

(e) Risks to Fixed Income Positions


Macroeconomic fluctuations often account for most of the risks associated with exposures to instruments that promise
to pay fixed or variable interest rates. For an investor that is long fixed rate instruments (thus receiving these cash
flows), increases in interest rates naturally lift the discount factors applied to the expected cash flows and thus cause a
loss. The longer the maturity of a certain cash flow and the higher the move in the discount factor, the higher will be the
loss. Upside surprises in inflation, fiscal funding needs, and FX depreciation rates are among the most common adverse
macroeconomic shocks to receivers. But counterparty exposure, issuer creditworthiness, client segmentation, regulation
(including changes in assets holding limits for different types of investors), changes in tax policies, currency
convertibility (which may constrain currency conversion, repatriation of profits and/or the liquidation of positions), and
settlement issues related to local clearing houses are also important risk factors to be considered. The sensitivity of fixed
income instruments to macroeconomic shocks may be mitigated by indexing the contracted cash flows to inflation, to
FX depreciation, or to specified interest rates - these are common in emerging markets. It is important to note that the
index fixings may by construction lag or mis-measure the actual move in the underlying variables they are intended
to track. The choice of the proper fixing (or metric) is particularly important in swaps markets, where floating coupon
rates (i.e., coupons indexed to a typically short-dated interest rate reference index) are exchanged for fixed coupons. It is
also important to acknowledge that funding in a currency that differs from the currency in which the coupons to be
received are denominated carries FX risk. Naturally, options on swaps (swaptions) also bear the risks typical to options
in addition to the risks related to rates movements.

Deutsche Bank AG/London

Page 85

David Folkerts-Landau
Global Head of Research
Marcel Cassard
Global Head
CB&S Research

Ralf Hoffmann & Bernhard Speyer


Co-Heads
DB Research

Guy Ashton
Chief Operating Officer
Research

Richard Smith
Associate Director
Equity Research

Asia-Pacific

Germany

North America

Michael Spencer
Regional Head

Andreas Neubauer
Regional Head

Steve Pollard
Regional Head

International Locations
Deutsche Bank AG
Deutsche Bank Place
Level 16
Corner of Hunter & Phillip Streets
Sydney, NSW 2000
Australia
Tel: (61) 2 8258 1234

Deutsche Bank AG
Groe Gallusstrae 10-14
60272 Frankfurt am Main
Germany
Tel: (49) 69 910 00

Deutsche Bank AG London


1 Great Winchester Street
London EC2N 2EQ
United Kingdom
Tel: (44) 20 7545 8000

Deutsche Bank Securities Inc.


60 Wall Street
New York, NY 10005
United States of America
Tel: (1) 212 250 2500

Deutsche Bank AG
Filiale Hongkong
International Commerce Centre,
1 Austin Road West,Kowloon,
Hong Kong
Tel: (852) 2203 8888

Deutsche Securities Inc.


2-11-1 Nagatacho
Sanno Park Tower
Chiyoda-ku, Tokyo 100-6171
Japan
Tel: (81) 3 5156 6770

Global Disclaimer
The information and opinions in this report were prepared by Deutsche Bank AG or one of its affiliates (collectively Deutsche Bank). The information herein is believed to be reliable and has been obtained from public
sources believed to be reliable. Deutsche Bank makes no representation as to the accuracy or completeness of such information.
Deutsche Bank may engage in securities transactions, on a proprietary basis or otherwise, in a manner inconsistent with the view taken in this research report. In addition, others within Deutsche Bank, including
strategists and sales staff, may take a view that is inconsistent with that taken in this research report.
Opinions, estimates and projections in this report constitute the current judgement of the author as of the date of this report. They do not necessarily reflect the opinions of Deutsche Bank and are subject to change
without notice. Deutsche Bank has no obligation to update, modify or amend this report or to otherwise notify a recipient thereof in the event that any opinion, forecast or estimate set forth herein, changes or
subsequently becomes inaccurate. Prices and availability of financial instruments are subject to change without notice. This report is provided for informational purposes only. It is not an offer or a solicitation of an
offer to buy or sell any financial instruments or to participate in any particular trading strategy. Target prices are inherently imprecise and a product of the analyst judgement. As a result of Deutsche Banks March 2010
acquisition of BHF-Bank AG, a security may be covered by more than one analyst within the Deutsche Bank group. Each of these analysts may use differing methodologies to value the security; as a result, the
recommendations may differ and the price targets and estimates of each may vary widely. The financial instruments discussed in this report may not be suitable for all investors and investors must make their own
informed investment decisions. Stock transactions can lead to losses as a result of price fluctuations and other factors. If a financial instrument is denominated in a currency other than an investors currency, a change
in exchange rates may adversely affect the investment. Past performance is not necessarily indicative of future results. Deutsche Bank may with respect to securities covered by this report, sell to or buy from
customers on a principal basis, and consider this report in deciding to trade on a proprietary basis.
Derivative transactions involve numerous risks including, among others, market, counterparty default and illiquidity risk. The appropriateness or otherwise of these products for use by investors is dependent on the
investors own circumstances including their tax position, their regulatory environment and the nature of their other assets and liabilities and as such investors should take expert legal and financial advice before
entering into any transaction similar to or inspired by the contents of this publication. Trading in options involves risk and is not suitable for all investors. Prior to buying or selling an option investors must review the
Characteristics and Risks of Standardized Options, at http://www.theocc.com/components/docs/riskstoc.pdf . If you are unable to access the website please contact Deutsche Bank AG at +1 (212) 250-7994, for a
copy of this important document.
The risk of loss in futures trading, foreign or domestic, can be substantial. As a result of the high degree of leverage obtainable in futures trading, losses may be incurred that are greater than the amount of funds
initially deposited.
Unless governing law provides otherwise, all transactions should be executed through the Deutsche Bank entity in the investors home jurisdiction. In the U.S. this report is approved and/or distributed by Deutsche
Bank Securities Inc., a member of the NYSE, the NASD, NFA and SIPC. In Germany this report is approved and/or communicated by Deutsche Bank AG Frankfurt authorized by the BaFin. In the United Kingdom this
report is approved and/or communicated by Deutsche Bank AG London, a member of the London Stock Exchange and regulated by the Financial Services Authority for the conduct of investment business in the UK
and authorized by the BaFin. This report is distributed in Hong Kong by Deutsche Bank AG, Hong Kong Branch, in Korea by Deutsche Securities Korea Co. This report is distributed in Singapore by Deutsche Bank AG,
Singapore Branch or Deutsche Securities Asia Limited, Singapore Branch, and recipients in Singapore of this report are to contact Deutsche Bank AG, Singapore Branch or Deutsche Securities Asia Limited, Singapore
Branch in respect of any matters arising from, or in connection with, this report. Where this report is issued or promulgated in Singapore to a person who is not an accredited investor, expert investor or institutional
investor (as defined in the applicable Singapore laws and regulations), Deutsche Bank AG, Singapore Branch or Deutsche Securities Asia Limited, Singapore Branch accepts legal responsibility to such person for the
contents of this report. In Japan this report is approved and/or distributed by Deutsche Securities Inc. The information contained in this report does not constitute the provision of investment advice. In Australia, retail
clients should obtain a copy of a Product Disclosure Statement (PDS) relating to any financial product referred to in this report and consider the PDS before making any decision about whether to acquire the product.
Deutsche Bank AG Johannesburg is incorporated in the Federal Republic of Germany (Branch Register Number in South Africa: 1998/003298/10). Additional information relative to securities, other financial products or
issuers discussed in this report is available upon request. This report may not be reproduced, distributed or published by any person for any purpose without Deutsche Banks prior written consent. Please cite source
when quoting.
Copyright 2013 Deutsche Bank AG

GRCM2013PROD030218