REPRODUCED FROM

RISK MANAGEMENT l DERIVATIVES l REGULATION

Tail risk
Three approaches for capturing fat tails

www.risk.net
MAY 2010

tail risk

Capturing fat tails

Financial institutions are more aware of the risks posed by high-impact events since the crisis,
but the question is how to encapsulate these in models. Zari Rachev, Boryana Racheva-Iotova
and Stoyan Stoyanov discuss three approaches for capturing fat tails

In this post-crisis era,

1 Fitted degrees of freedom parameter
30

Degrees of freedom

25
20
15
10
5
0
0

100 200 300 400 500 600 700 800

Note: graph shows the fitted degrees of freedom

parameter of the residuals of a Garch model fitted on
the returns of the constituents of the S&P 500

DJIA return

DJIA index

2 Fitted degrees of freedom parameter of

the Garch residuals of the returns of the
DJIA index changing through time
4
1.5 10

1.0
0.5
Oct 23,
1997
0.1

DOF

Oct 5,
2005

Sep 25,
2009

Oct 16,
2001

Oct 5,
2005

Sep 25,
2009

Oct 16,
2001

Oct 5,
2005

Sep 25,
2009

0.1
Oct 23,
1997
30
20
10
0
Oct 23,
1997

72

Oct 16,
2001

Reproduced from Risk May 2010

there is universal agreement that financial assets

are indeed fat-tailed and that investment managers
must take extreme events into account as part of their everyday risk management
processes. But deliberation continues at a high level on how risk management approaches
and practices should change. While academic research has provided a vast offering of
modern risk methods and analytic techniques, the race to integrate them into standard
risk platforms has just begun.
This work has shown there is more to the production of accurate risk estimates than
simply acknowledging asset returns have a higher probability of extreme events than
had been thought. There are numerous phenomena that, if left out of the equation, will
render risk measures such as value-at-risk virtually useless in accurately estimating
levels of risk and the probability of extreme price movements. Skewness, auto-regression and volatility clustering are recognised phenomena that must be considered.
However, the issue of varying tail-thickness from asset to asset and across time is
widely ignored.
Focusing on daily returns across multi-asset class portfolios, the purpose of this article is
to compare and contrast the more popular fat-tailed methodologies currently being
discussed. These approaches include the classical Students t model, extreme value theory
(EVT) and stable Paretian distributions. All these methods have been widely researched,
with long academic histories outside the financial arena, and now provide the foundations
for a range of commercial applications by leading risk management service providers.
Real-world models

When constructing realistic models, it is necessary to assume a distributional

hypothesis capable of describing both fat tails and asymmetry. Several classes of
distributions have been used to capture fat tails, both in academia and by practitioners.
Perhaps the most popular is the classical Students t distribution. Other examples
include extreme value distributions, stable distributions, operator stable distributions,
the class of tempered stable distributions that include stable distributions as a limiting
case, and the class of infinitely divisible distributions that include all previous classes
except extreme value distributions.
All these classes of models, except extreme value distributions, share one feature they
include the normal distribution as a special (limiting) case. In effect, if the data is
Gaussian, the fitted distribution would be close to, or would coincide with, the normal
distribution. Therefore, these families of models can be regarded as an extension to the
classical Gaussian framework and not an alternative to it.
As far as modelling asymmetry is concerned, neither the Gaussian distribution nor the
classical Students t can account for skewness. Instead, we have to turn to models such as
the stable Paretian distribution, which look at the respective left and right tails. One way
to capture the difference between the upside and the downside potential is by calculating
expected tail loss and expected tail return.
It is important to note that the degree of tail thickness varies across assets and asset
classes. We carried out an empirical study that included the stocks in the S&P 500
universe during the 12-year period from January 1, 1992 to December 12, 2003. We fitted
a Garch model to clean the volatility clustering effect and then fitted the classical Students
t model on the residual. The degrees of freedom (DOF) parameter, which governs the tail
behaviour, is shown in figure 1.

The plot illustrates that tail behaviour

can be quite diverse from very fattailed, with a DOF below five, to less fattailed, with a DOF above 15. In fact, the
study found 21% of the S&P 500 stocks
are very fat-tailed, with a DOF below
four and just 35% have a DOF above
seven. Since the DOF parameter governs
the potential for extreme events, realistic
estimates are essential in ensuring that
portfolio tail risk contributors and
diversifiers are properly identified.
Not only does tail behaviour vary
across assets, it also varies through time.
In relatively calm periods, asset returns
are almost Gaussian, while in turbulent
periods, the tails become fatter. Figure 2
illustrates this behaviour in the Dow
Jones Industrial Average (DJIA) index
returns from October 1997 to October
2009. The top and middle plots show the
value and return of the DJIA, respectively. The bottom plot shows the fitted
DOF parameter of the residuals of a
Garch model fitted on a 500-day rolling
window. Clearly, the tail behaviour
changes through time. In the period from
October 2003 to about January 2006, the
tail behaviour is almost Gaussian as the
fitted DOF is above 30.
It is crucial for the model to take into
account the differences in tail behaviour
both across assets and through time so
they can be reflected in the risk statistics
on both a marginal and aggregate level.
Applications of fat-tailed models

Risk management software vendors are

offering approaches based on fat-tailed
models including classical Students t
distribution, EVT and stable Paretian
distributions.
n The Students t distribution.
A typical approach in building a
framework based on the classical
Students t distribution is:
n An auto-regressive component to
capture auto-regressive behaviour.
n Volatility clustering by means of Garch
or alternative short- or long-memory
Arch-type processes.
n The classical symmetric Students t
distribution with the DOF parameter
fixed for all variables (typically a DOF of
four or five) to capture fat tails.
First introduced in 1908, Students t
distribution is probably the most commonly used alternative to the normal
distribution as a model for asset returns.
Like the normal distribution, classical
Students t densities are symmetric and have

a single peak. Unlike the normal distribution, Students t densities are more peaked
around the centre and have fatter tails.
While these two properties make them
acceptable for asset returns modelling,
the real reason behind the widespread
use of Students t is its ease of use
numerical methods are easily implementable and are widely available.
The Garch component could be
replaced with alternative models using
an exponential or logarithmic decay of
the observation weights when calculating
the volatility based on a pre-defined
parameter (such as 0.94 for the exponentially weighted moving average decay
parameter (see Zumbach, 2006)). This
forces the relative importance of the
observations in the past to be the same
for all risk drivers and across time. While
this universal parameter makes these
models simpler and easier to grasp, there
is an important trade-off between
simplicity and precision: these models
are less accurate and only work on
average in a universe of risk drivers.
The most significant limitation in a
classical Students t distribution-based
framework, however, is that the residual
in the time-series model is assumed to
have a Students t distribution with the
DOF parameter fixed (typically to four
or five).1 This value is assumed to be one
and the same for all risk drivers, irrespective of their type and the time period
under consideration. This assumption is
not realistic as empirical analyses
indicate that tail behaviour varies across
different risk drivers.2 Fixing the DOF
parameter does not allow for a smooth
transition between Gaussian data and
fat-tailed data. As a result, the risk will
be significantly overestimated for assets
with returns that are close to being
normally distributed.
Finally, the classical Students t model
is symmetric. In cases where there is a
significant asymmetry in the data, it will
not be reflected in the risk estimate. By
forcing the tails to be identical, it becomes
impossible to reveal which assets are true
tail risk contributors and diversifiers.
n EVT generalised Pareto
distribution.
The key characteristics based on suggested
approaches include:
n Volatility clustering by means of a
Garch model.
n EVT to explain the fat tails of the
residuals from the Garch model.
n Skewness captured by using a

generalised Pareto distribution (GPD) to

fit the two tails separately.
EVT has been applied for a long time
when modelling the frequency of
extreme events, including extreme
temperatures, floods, winds and other
natural phenomena. From a general
perspective, extreme value distributions
represent distributional limits for
properly normalised maxima of random
independent quantities with equal
distributions, and therefore can be
applied in finance as well.
A common approach to EVT-type
modelling is the peaks-over-threshold
method that follows GPD and models
those events in the data that exceed a high
threshold. This is presented in a nonnormal framework by Embrechts,
Klppelberg & Mikosch (1997).
GPD is the limiting distribution of the
exceedances of a given return distribution
over a certain threshold when the
threshold (which can be viewed as the
right tail of the original distribution) goes
to infinity. It is a model for the tail only,
left or right. GPD does not model the
body of the corresponding distribution.
As a consequence, the parameters of
GPD can be fitted using only information
from the respective tail. There are two big
challenges stemming from this restriction:
n An extremely large sample is needed to
get a sufficient number of observations
from the tail.
n We need to know where the body of the
distribution ends and where the tail
begins.
The large sample size is a significant
challenge. Academic publications indicate
that minimum requirements are in the
range of 5,000 to 10,000 observations.
However, it should also be noted that
using a sample this large minimises any
current fat-tailed market behaviour and
would only be suitable for long-term
projections. For financial time series,
where the standard time window for risk
estimation is two years of daily data, a
sample size of just 500 observations is far
too short to ensure an accurate GPD fit.
Goldberg, Miller & Weinstein (2008)
suggest 1,000 days, with approaches to
generate synthetic data where enough
observations are not available.
The second challenge separating the
body of the distribution from the tail
may seem easy to surmount by resorting
to statistical methods that would indicate
1
2

Described in section 4 of Zumbach (2006)

A large empirical study is presented in Rachev et al (2005)

risk-magazine.net

73

tail risk

Tail index

DJIA return

1.5

4 Dow Jones Industrial Average: July 8, 2005December 31, 2009

15,000
14,000

104

13,000
1.0
0.5
Oct 23,
1997

Oct 16,
2001

Oct 5,
2005

Index level

DJIA index

3 Fitted tail index of the Garch residuals

of the returns of the DJIA index changing
through time

Sep 25,
2009

0.1

10,000
9,000
7,000

Oct 5,
2005

Oct 16,
2001

6,000
Jul 8,
2005

Sep 25,
2009

1.9
1.8
Oct 23,
1997

11,000

8,000

0
0.1
Oct 23,
1997
2.0

12,000

Oct 5,
2005

Oct 16,
2001

Sep 25,
2009

where the tail begins. Unfortunately, no

reliable methods exist.
Typically, this high threshold is chosen
subjectively by looking at certain plots,
such as the Hill plot or the mean excess
plot, which are standard in EVT. As a
consequence, identifying the threshold
between the body and the tail is a matter
of subjective choice based on visual
inspection, which cannot be achieved on
a large scale. Kuipers test has been
suggested as a numerical method for
determining the optimal threshold
selection, as suggested in Goldberg,
Miller & Weinstein (2008). However, this
test is very difficult to automate for large
universes, because the resulting optimisation problem does not have good

Jan 4,
2006

Jul 3,
2006

Dec 30, Jun 28, Dec 25, Jun 22, Dec 19, Jun 17, Dec 14,
2007
2008
2008
2006
2007
2009
2009

optimality properties, with the global

minimum being hard to find.
As a result, unreliable threshold
selections will be made in the absence of
thorough visual inspection. An automated
approach remains elusive. The choice of
this threshold has a great impact on the
parameter estimates of GPD and,
therefore, on the final risk estimates. This
will be especially acute when the sample is
relatively small. This deficiency is
acknowledged in Goldberg, Miller &
Weinstein (2008) and del Castillo &
Daoudi (2008).3
In using GPD4, one is always faced
with the classical tail-estimation trade-off
problem. For the GPD estimates to be
unbiased, they must be fit with the largest
possible threshold. Unbiased estimators
are obtained when the threshold is
infinity. This implies that one should use
a very small number of extreme observations from the original sample. On the

5 Back-test of the 99% daily VAR calculated according to different fat-tailed

methodologies

12.5
10.0
7.5
5.0
2.5
0
2.5
5.0
7.5
10.0
12.5
15.0
17.5
20.0
22.5
25.0
27.5
Jul 8,
2005

74

DJIA returns
VAR Gaussian Garch
VAR stable Garch
VAR Students t DOF at five
VAR EVT 1.02% threshold

other hand, using such a small number of

observations drastically increases the
variance of the estimators. GPD estimation fits become more of an art than a
science in balancing this trade-off
between being unbiased and small
variance of the estimators.
n Stable Paretian distributions.
The key characteristics of a stable Paretian
distribution implementation include:
n An auto-regressive component to
capture auto-regressive behaviour.
n Volatility clustering captured by means
of a Garch model.
n Stable Paretian distributions to explain
the fat tails and the skewness of the
residual from the Garch model with assetspecific parameters.
n Temporal behaviour of tail thickness
captured by fitting the full distribution
from historical data.
Applications of stable distributions in
the field of finance have a long history.5 In
1963, the mathematician Benoit Mandelbrot first used the stable distribution to
model empirical distributions that have
skewness and fat tails. To distinguish
between Gaussian and non-Gaussian
stable distributions, the latter are
commonly referred to as stable Paretian or
Lvy stable distributions.
Stable Paretian tails decay more slowly
than the tails of the normal distribution
and therefore better describe the extreme
events present in the data. Since they
represent a model for the entire distribution and not just the tails, reliable model
parameter estimation methods exist. The
See page 687 of del Castillo & Daoudi (2008)
While we refer to GPD estimation, it should be noted that GEV
estimation suffers from the same properties (see page 331 of Hull &
Welsh, 1985)
5
See Rachev & Mittnik (2000)
3

Jan 4,
2006

Jul 3,
2006

Dec 30,
2006

Reproduced from Risk May 2010

Jun 28,
2007

Dec 25,
2007

Jun 22,
2008

Dec 19,
2008

Jun 17,
2009

Dec 14,
2009

Comparing models

Using the DJIA, a back-testing study was

conducted to compare the three fattailed models: stable Paretian, Students t
with a DOF of five and EVT, alongside
the normal distribution model. All the
models have filters for auto-regression
and volatility clustering based on ArmaGarch, with the Students t model using
the particular method described in
Zumbach (2006). For each of the four
models, exceedances the number of
times the real loss is larger than the
calculated VAR are tracked.
The back-testing was run with the
following settings:
n Back-test period: July 8, 2005 to
December 31, 2009.
n VAR confidence level: 99%.

n Time window: 500 rolling days for

normal, classical Students t and stable

Paretian, and 3,000 rolling days for
EVT.8
n EVT threshold: 1.02% (as suggested by
Goldberg, Miller & Weinstein, 2008).
Figure 4 shows the DJIA performance
for the back-test period. Figure 5 shows
the daily forecast for the four models
versus the daily returns of the DJIA across
the full back-test period. The number of
exceedances for the four models is
reported in figure 6. Using a 95%
confidence interval, the number of
exceedances is compared. The results
show the normal-Garch model is too
optimistic, with its daily 99% VAR
forecasts being too low. In contrast, the
Students t and EVT approaches are
overly pessimistic, with their forecasts
being too high.
Figure 7 zooms in on the period
between July 1, 2006 and June 30,
2007. This 12-month history has
relatively low volatility, but includes a
drop of 3.4% on February 27, 2007.
Neither the Students t nor the EVT
model can distinguish between normal
and fat-tailed markets. As observed in
figure 7, the Students t and EVT models
simply overreact to extreme events when
they occur. The EVT model, the most
pessimistic, is unable to adjust to
current market conditions because of
the very large sample size that is
required to ensure stable estimates.
Garch alone does not help since there
are changes in the tail behaviour of the
markets as well.
The stable-Garch model appears to be
the only one with realistic VAR forecasts
with exceedances within the confidence

6 Number of exceedances in the daily 99%

VAR back-testing experiment
20

20
16

11

12
8

4
0

Normal
Garch

Stable
Classical EVT (GPD)
Paretian Students t
Garch
Garch

Note: the back-testing period is 1,130 days and the 95%

confidence interval for the number of exceedances is [4, 17]

interval. Additionally, the daily forecasts

of the stable-Garch model react with
extreme events. They change according
to new market information, sometimes
with estimates lower than the normal
model, making the spread between the
normal and stable risk estimates
indicative of the markets probability of
extreme events.
Figure 8 shows the minimum and
maximum VAR for the EVT model based
on selecting different thresholds (115%).
It shows minimum EVT and Students t
with a DOF of five almost coincide for
calm periods and are still too conservative. The percentage spread between
minimum EVT and normal does not
change, indicating indifference to market
conditions. The maximum EVT is
constantly one-and-a-half to two times
the minimum EVT, which points to the
dramatic influence of the tail threshold
selection on the risk estimates.
Forthcoming in Journal of Banking and Finance
Forthcoming in Society for Industrial and Applied Mathematics
Goldberg, Miller & Weinstein (2008) use time windows ranging
from approximately 1,500 to 7,600 days
6
7
8

7 Back-test period: July 1, 2006June 30, 2007

3
2
1
0
%

instability of parameter estimation,

inherent in EVT, is not present for stable
distributions because the entire sample is
taken into account, rather than just the
tail of the distribution. The operator
stable version of the stable Paretian
distribution allows for varying tailfatness from asset to asset. A tailtempering process ensures a finite second
moment (variance). Tail tempering is
achieved by imposing an additional
exponential, or faster than exponential,
decay in the tail very far away from the
centre of the distribution.
A detailed description of the stable
methodology is available in Rachev et al
(2009).
The tail-tempering process is detailed in
Kim et al (2008), Young et al (2010)6 and
Bianchi et al (2010)7.
Like the Students t distribution, stable
Paretian distributions have a parameter
responsible for the tail behaviour, which is
called the tail index or index of stability.
In contrast to the DOF parameter, the
index of stability is between zero and two.
The closer it is to two, the more Gaussianlike the distribution is. Therefore, small
values of the index of stability imply a
fatter tail.
The fact the tail index changes through
time is demonstrated in figure 3 with
DJIA returns in the period from October
1997 to October 2009. The tail index is
very close to two in the upward market
from 2003 to 2005, but then starts
decreasing right before the market crash
and is smallest at the crash itself. This
implies the tail thickness is smallest in the
bullish market from 2003 to 2005 and is
largest during the crisis periods.

1
2
DJIA returns
VAR Gaussian Garch
VAR stable Garch
VAR Students t DOF at five
VAR EVT 1.02% threshold

3
4
5
6
Jul 3,
2006

Oct 1,
2006

Dec 30,
2006

Mar 30,
2007

risk-magazine.net

Jun 28,
2007

75

TAIL RISK

8 Minimum and maximum EVT bundaries in red versus the Students t

10.0
7.5
5.0
2.5
0
2.5
5.0
7.5
10.0
12.5
15.0
17.5
20.0
22.5
25.0
27.5
Jul 8,
2005

DJIA returns
VAR Students t DOF at five
VAR EVT-max (115% threshold)
VAR EVT-min (115% threshold)
Jan 4,
2006

Jul 3,
2006

Dec 30,
2006

Jun 28,
2007

Dec 25,
2007

Conclusion

Suddenly, it seems everyone agrees fat

tails exist and are crucial to properly
estimate and proactively manage risk.
Vendors and practitioners are now trying
harder to incorporate fat-tailed risk
assessment into their risk management
systems. Models based on normal
distributions cannot capture them. To
accurately measure risk, we need to focus
on how the behaviour of assets relates
and contributes to fat tails.
Empirical properties including autoregressive behaviour, clustering of
volatility, skewness and fat tails, including
differences in thickness across assets and
through time, can all be pronounced in
daily asset returns. Modelling these
phenomena is a two-step process: fit a
time-series model to explain the clustering of volatility and auto-correlation, and

Jun 22,
2008

Dec 19,
2008

Jun 17,
2009

Dec 14,
2009

employ a fat-tailed model for the residual.

We considered three approaches that all
start with capturing the fat tails. Taking
into account these results and the
preceding model descriptions, a final
model comparison is as follows:
n EVT approach. This is overly pessimistic
at all times. There is a slow relative response
when a market becomes more likely to have
large losses. Automated methods to fit the
tails are impractical for daily risk
management use because of the pessimism
penalty, data requirements and tail-fitting
implementation issues. The approach is
potentially capable of reviewing downside
risk from time to time, but not over time. It
will require manual set-up and intervention
to produce reasonable results.
n Classical Students t approach. While
not as pessimistic as EVT, it still requires
taking a large penalty and has little ability

to discriminate between assets and how

they change over time. This is further
compounded by ignoring skewness. It is
easily implemented for day-to-day risk
management use but with overly
conservative results and limited insight
into the key contributors to specific risk.
n Generalised stable Paretian approach.
The risk estimates do not impose a risk
penalty during normal market times
and are responsive when fat-tailed
behaviour becomes more probable.
There is strong differentiation between
assets and over time. It is a practical
approach for day-to-day risk
management with high discrimination
of key risk drivers in the tail.
Additional new commercial approaches
are likely to appear in 2010. It seems clear
many of these will provide alternative
lenses through which users may view and
analyse risk, but will not be replacements
or upgrades to their core models. How
and when users should fit this into their
process is unclear. Are market participants expecting a separate model to evaluate tail risk, for example?
The debate will continue around what
is best to implement and how. We suggest
that if you have a secondary model
specifically for use in times of crisis, by
the time you implement it, the chances
are it will be too late. n
Zari Rachev is chair-professor of statistics,
econometrics and finance at KIT in Germany and
FinAnalytica chief scientist. Boryana Rachev-Iotova
is president of FinAnalytica. Stoyan Stoyanov is a
professor of finance at Edhec Business School and
the scientific director for Edhec-Risk Institute in Asia

References
Bianchi M, S Rachev, Y Kim and
F Fabozzi, 2010
Tempered infinitely divisible
distributions and processes
Forthcoming in Theory of Probability
and Its Applications, Society for Industrial and Applied Mathematics, available
at www.statistik.uni-karlsruhe.de/download/doc_secure1/TID20080729.pdf
Del Castillo J and J Daoudi, 2008
Estimation of the generalized Pareto
distribution
Statistics & Probability Letters 79(5),
pages 684688
Embrechts P, C Klppelberg and
Mikosch, 1997
Modelling external events for insurance
and finance
Springer, Berlin

76

Reproduced from Risk May 2010

Goldberg L, G Miller and J Weinstein,

2008
Beyond value at risk: forecasting
portfolio loss at multiple horizons
Journal of Investment Management
6(2), pages 7398
Hull P and A Welsh, 1985
Adaptive estimates of parameters of
regular variation
Annals of Statistics 122(1), pages
331341
Kim Y, S Rachev, M-L Bianchi and
F Fabozzi, 2008
Financial market models with Lvy
processes and time-varying volatility
Journal of Banking and Finance 32(7),
pages 1,3631,378

Mandelbrot B, 1963
The variation of certain speculative
prices
Journal of Business 36, pages 394419
Rachev S, R Martin, B Racheva-Iotova
and S Stoyanov, 2009
Stable ETL optimal portfolios and
extreme risk management
In Risk Assessment: Decisions in Banking and Finance, Springer-Physika,
pages 235262
Rachev S and S Mittnik, 2000
Stable Paretian models in finance
John Wiley & Sons, Series in Financial
Economics

Rachev S, S Stoyanov, A Biglova and

F Fabozzi, 2005
An empirical examination of daily stock
return distributions for US stocks
In Data Analysis and Knowledge Organization, Springer-Verlag, Berlin, pages
269281
Young S, S Rachev, M-L Bianchi and
F Fabozzi, 2010
Tempered stable and tempered infinitely
divisible Garch models
Forthcoming in Journal of Banking and
Finance, available at www.statistik.unikarlsruhe.de/download/RDTS-GARCHMathFin.20090127.pdf
Zumbach G, 2006
A gentle introduction to the RM2006
methodology
RiskMetrics Technology Paper

FinAnalytica is a leading provider of real-world portfolio and

risk management solutions for quantitative analysts and portfolio
managers.
FinAnalytica's Cognity software suite incorporates the latest and
most transparent advances in analytics, including comprehensive
treatment of real-world fat-tailed and skewed asset returns.
Cognity holdings-based and returns-based market risk solutions
offer research and risk professionals the most incisive view of risk
through patented fat-tailed processes that accurately differentiate
downside risks and upside opportunities. Cognity demonstrates
to investors a clear grasp of forward-looking, real-world risks.
With offices in New York, London and Sofia, FinAnalytica
supports leading asset managers, hedge funds, fund of
funds, endowments, and pension funds globally. For further
information, please visit www.finanalytica.com
New York
100 Park Avenue, 16th Floor
New York, NY 10017 USA
Telephone: +1 (212) 880 2670
Fax: +1 (212) 880 2679
e-mail: sales@finanalytica.com
London
1 Liverpool Street
London EC2M 7QD UK
Telephone: +44 (0)20 7956 2778
Fax: +44 (0)20 7956 2001
e-mail: sales@finanalytica.com
Sofia, Bulgaria
1407 Sofia
21 Srebarna Str., Floor 5
Telephone: +359 2 962 4645
e-mail: sales@finanalytica.com