Tail Hedging Strategies

Issam S. S TRUB ∗

The Cambridge Strategy (Asset Management) Ltd

Abstract

This article introduces an algorithm for tail risk hedging and compares it to other existing meth-
ods. This algorithm adjusts the exposure level based on a measure of tail risk obtained by applying
Extreme Value Theory (EVT) to estimate Conditional Value at Risk (CVaR). This method is applied to
the S&P 500 and MSCI Emerging Markets equity indexes between 2001 and 2015 and its performance
is compared to cash and options based tail hedging strategies. The cash based methods are shown to
significantly increase risk adjusted returns and reduce drawdowns, while the options based strategy suf-
fers a decrease in performance from 2003 on due to the increase in cost of puts with respect to calls
after that date. The tail hedging technique presented in the article is fully investable as its turnover is
limited; additionally it can replace long/short equity hedge funds for investors who do not have access to
alternative investments.

1 Introduction
The notion of tail risk has risen to the forefront of the financial lexicon in recent years, in large part due
to the significant losses experienced by equity markets during the 2008 financial crisis and the subsequent
Eurozone turmoil. The expression tail risk originates from the frequency plot of the daily returns of an
asset and its modelling by a probability distribution. The left and right tails denote the extremities of the
distribution curve where the largest negative and positive returns are plotted. Naturally, investors are mostly
concerned with the left tail and tail risk refers to the probability of an asset experiencing large negative daily
returns; thus, tail hedging primarily consists in limiting the magnitude of large negative returns. Indeed, a
large drawdown can have a devastating impact on an investor’s wealth; for example, it may force a company
to inject cash in its pension plan at an inauspicious time in order to meet the plan’s liabilities. In addition to
mitigating drawdowns, tail hedging has other beneficial aspects from an investor standpoint; an investor is
more likely to be willing to invest in a more aggressive asset class if risks are limited through tail hedging
and to stick with his investment through occasional crises. These strategies allow an investor to benefit from
the higher returns associated with asset classes such as Emerging and Frontier Markets equities. Also, tail
hedges typically provide liquidity during market downturns, which can then be used to purchase assets at
distressed prices.
As a consequence, the demand for tail hedging is growing and a number of strategies are available to
investors. Most of them can be described as options or cash based. The options based strategies consist
∗
Director, Quantitative Research & Risk Management, [email protected]

1
in buying puts and possibly selling calls in order to generate profits during a fall in asset prices. The
cash based strategies consist in dynamically allocating a portion of a portfolio to cash, with the objective
of reducing the exposure to equity markets in down markets and increasing it during up markets. Such
strategies originated three decades ago with the development of portfolio insurance in L ELAND (1980);
RUBINSTEIN and L ELAND (1981); B RENNAN and S OLANKI (1981); B RENNAN and S CHWARTZ (1989)
and later on constant proportion portfolio insurance by B LACK and J ONES (1987); P EROLD and S HARPE
(1988); B LACK and P EROLD (1992) in which put options on a portfolio were replicated through futures
contracts.
This article develops a cash based tail hedging strategy and evaluates its performance by applying it to
both U.S. and Emerging Markets equity indexes and comparing it with other existing cash and options based
strategies. Before presenting the strategy, we begin by analysing the features of tail risk and volatility in
equity markets.

2 Volatility and Tail Risk in Equity Markets

2.1 Volatility Structure of Equity Markets
In order to reduce the exposure to equity markets during downturns, an indicator of such downturns is re-
quired. While equity market returns are difficult to predict, the volatility of these markets tends to exhibit
significant persistence and clustering, as noted in M ANDELBROT (1963, 1967). Indeed, an analysis of daily
returns for the S&P 500 and MSCI Emerging Markets index over the past 15 years shows little autocor-
relation in daily returns (Figure 1), while the autocorrelation of squared returns is significant (Figure 2),
highlighting the persistence in volatility levels over time. Volatility forecasting and its potential applications
in portfolio construction are studied in A KGIRAY (1989); C AMPBELL and H ENTSCHEL (1992); B RAILS -
FORD and FAFF (1996); F LEMING et al. (2001, 2003).
Also, higher volatility is associated with lower returns while lower volatility periods generally present
higher returns. This is confirmed by an analysis of the monthly returns and volatility of the S&P 500 and
MSCI Emerging Markets indexes during the 2001-2015 period. The data is sorted into deciles according to
volatility and the average return and volatility for each decile are reported in Table 1. For both the U.S. and
Emerging Markets, the top deciles in terms of volatility correspond to the lowest returns while the bottom
deciles present higher returns. For example, the top three deciles have average returns of -1.18% and -1.96%
while for the bottom three deciles, the average returns are 1.87% and 2.67%. Note that while the average
volatility in each decile is roughly equal for U.S. and Emerging Markets (contrary to the popular opinion
that Emerging Markets are riskier), the difference in return between the top three and bottom three deciles
is noticeably higher for Emerging Markets, suggesting that a tail hedging strategy based on volatility levels
could improve risk adjusted performance more significantly for Emerging Markets than for U.S. equities.
The Capital Asset Pricing Model (CAPM) introduced in S HARPE (1964); L INTNER (1965); B LACK
(1972) predicts that the expected return of a well diversified portfolio increases with volatility, but the data
in Table 1 contradicts this prediction as it shows that lower volatility leads to higher returns and higher
volatility to lower returns for both the S&P 500 and MSCI Emerging Markets indexes. This observation
that higher market volatility is not compensated by higher returns is also applicable to individual stocks, as
demonstrated in numerous studies on equity markets around the world such as FAMA and F RENCH (1992,
1993, 2004); G LOSTEN et al. (1993); BAKSHI and K APADIA (2003); P OON and G RANGER (2003); F RENCH
et al. (2004); A NG et al. (2006, 2009); B LITZ and VAN V LIET (2007). This has led to the development of

2
 S&P 500 MSCI Emerging Markets
Decile Avg. Return (%) Avg. Volatility (%) Avg. Return (%) Avg. Volatility (%)
1 -2.68 40.81 -4.50 37.39
2 -2.01 24.70 -0.35 22.68
3 1.15 18.78 -1.03 18.59
4 -1.03 16.24 0.51 15.94
5 0.35 14.87 0.83 14.70
6 0.38 13.03 2.30 13.63
7 1.63 11.28 1.27 12.18
8 0.84 9.93 3.22 10.83
9 2.28 8.76 1.98 9.73
10 2.49 7.16 2.80 8.00

Table 1: Average decile monthly return and volatility for the S&P 500 and MSCI Emerging Markets indexes
from 2001 to 2015.

minimum-variance portfolios and other low volatility strategies which have been shown to generate higher
risk adjusted returns (C LARKE et al. (2006, 2010, 2011); C HONG and P HILLIPS (2012)); minimum variance
portfolios are constructed by selecting stocks from the universe of risky assets so as to present the lowest
possible volatility, irrespective of expected returns.
Since higher volatility is linked to lower returns, a possible tail hedging strategy would be to attempt to
maintain volatility at a constant intermediate level; indeed, this would lead to increasing (decreasing) expo-
sure when volatility is low (high) which corresponds to high (low) returns (T HOMAS and S HAPIRO (2009)).
Many such targeted volatility strategies have been developed recently and are offered to investors as an alter-
native to traditional equity investments (BANERJEE and S RIVASTAVA (2012); G IESE (2012); T HOMAS and
C OLLIE (2012); YAN (2012)). These strategies offer several advantages: by targeting a specific volatility
level, a form of tail hedging is implemented as exposure is reduced during market downturns since these
periods are associated with higher volatility; another positive feature is that a managed volatility portfolio is
not subject to the wide fluctuations in equity market volatility (during the past fifteen years, annual volatility
varied from 10.02% to 40.88% for the S&P 500 and from 11.08% to 39.95% for the MSCI Emerging Mar-
kets index); in fact, a survey of institutional investors by RUSSELL I NVESTMENTS and P&I (2012) showed
that market volatility ranks as the top concern for corporate and public pension plans as well as endowments
and foundations.

2.2 Non Normality of Equity Returns

The managed volatility strategies attempt to mitigate tail risk based on the level of volatility of the market.
However, relying on volatility to measure risk usually assumes a normal distribution of returns, which is

3
 1

0.8
Sample Autocorrelation

0.6

0.4

0.2

-0.2
0 5 10 15 20
Lag

0.8
Sample Autocorrelation

0.6

0.4

0.2

-0.2
0 5 10 15 20
Lag

Figure 1: Top: Sample autocorrelation function of daily returns for the S&P 500 between 2001 and 2015.
Bottom: Sample autocorrelation function of daily returns for the MSCI Emerging Markets between 2001
and 2015.

4
 1

0.8
Sample Autocorrelation

0.6

0.4

0.2

-0.2
0 5 10 15 20
Lag

0.8
Sample Autocorrelation

0.6

0.4

0.2

-0.2
0 5 10 15 20
Lag

Figure 2: Top: Sample autocorrelation function of squared daily returns for the S&P 500 between 2001 and
2015. Bottom: Sample autocorrelation function of squared daily returns for the MSCI Emerging Markets
between 2001 and 2015.

5
likely to result in an underestimation of risk levels. Indeed, most equity returns are non normal and present
a higher tail risk than that of a normal distribution; this fact, which has been widely noted in the risk man-
agement literature after the 2008 financial crisis, was already expressed five decades ago in M ANDELBROT
(1963, 1967). Figure 3 shows the density chart of daily returns of the MSCI Emerging Markets over the
past twenty years, as well as a normal distribution with the same mean and standard deviation. The top
graph shows the normal distribution approximates well the center of the historical return distribution, but
the bottom graph, which focuses on the left tail of returns, underlines the significant underestimation of tail
risk that occurs when relying on volatility and normal distribution assumptions; in particular, negative daily
returns in excess of -4% are not accounted for by the normal distribution. The situation is similar for the
S&P 500. Figure 4 shows a Q-Q plot of standardised daily returns for the MSCI Emerging Markets and
S&P 500 indexes from 2001 and 2015; this graph confirms that the left and right tails are much thicker than
those of a normal distribution. This larger than normal tail risk is partly responsible for the outsized draw-
downs experienced in market downturns (J OHANSEN and S ORNETTE (2001)), thus being able to accurately
measure and control it is likely to yield significant improvements in risk adjusted performance.

3 Tail Risk Measures

A common measure of tail risk is Value at Risk (VaR) (B EDER (1995); D UFFIE and PAN (1997); J ORION
(2006)), which is defined as the minimum loss experienced over a given time horizon with a given prob-
ability. When applied to historical daily returns, VaR can be computed by ordering the daily returns and
selecting the quantile corresponding to the confidence level chosen (for example 95%). Unfortunately, VaR
is concerned only with the number of losses that exceed the VaR confidence level and not the magnitude
of these losses; to obtain a more complete measure of large losses, one needs to examine the entire shape
of the left tail of the return distribution beyond the VaR threshold, which leads to the Conditional Value at
Risk (CVaR) also referred to as Expected Shortfall, Tail VaR or Mean Shortfall (A RTZNER et al. (1999);
C HRISTOFFERSEN (2003); H ARMANTZIS et al. (2006); M C N EIL et al. (2005)). CVaR can be defined as
the average expected loss at a given confidence level; for example, at the 95% confidence level, the CVaR
represents the average expected loss on the worst 5 days out of 100 whereas the VaR is the minimum loss
on those days. In mathematical terms, the CVaR for a daily return distribution F at a confidence level α is
given by:

CVaRα = −E{X|X 6 −VaRα } (1)

where the VaR is defined by:
VaRα = −F −1 (1 − α) (2)
Computing CVaR requires an explicit expression of the portfolio return distribution function F which
is usually unknown in practice. However, if historical daily returns are assumed to follow a normal (or
Gaussian) distribution, VaR and CVaR can be easily obtained from the standard deviation σ and mean µ of
returns; for example, at the 95% level, standard deviation, VaR and CVaR are related by:

VaR ' 1.65 × σ − µ and CVaR ' 2.07 × σ − µ (3)

6
 0.6
Daily Returns
Normal Distribution
0.5

0.4
Density

0.3

0.2

0.1

0
-5 0 5 10
Daily Return (%)

0.25
Daily Returns
Normal Distribution
0.2

0.15
Density

0.1

0.05

0
-9 -8 -7 -6 -5 -4 -3 -2 -1
Daily Return (%)

Figure 3: Top: Density chart of MSCI Emerging Markets daily returns between 2001 and 2015 and corre-
sponding normal distribution. Bottom: Same as above but focusing on the left tail of the return distribution;
the underestimation of tail risk by the normal distribution is apparent.

7
 15
MSCI Emerging Markets
Normal Distribution
10 S&P 500

-5

-10
-4 -2 0 2 4

Figure 4: Q-Q plot of the standardised daily returns of the MSCI Emerging Markets and S&P 500 indexes
from 2001 to 2015. The lowest and highest quantiles are clearly not normally distributed unlike the middle
quantiles.

8
3.1 Extreme Value Theory
As noted earlier, relying on a normal distribution assumption is dangerous in practice as this is likely to
result in underestimating tail risk; tail risk can be more accurately measured using tools originating from
Extreme Value Theory (EVT), a branch of statistics dedicated to modelling extreme events introduced in
BALKEMA and DE H AAN (1974); P ICKANDS (1975). The central result in Extreme Value Theory states
that the extreme tail of a wide range of distributions can be approximately described by the Generalised
Pareto Distribution (GPD) with shape and scale parameters ξ and β:
−1
(
1 − (1 + ξy
β ) ξ , for ξ 6= 0;
Gξ,β (y) = −y
(4)
1 − exp β , for ξ = 0.

where β > 0, and the support of Gξ,β is y > 0 when ξ > 0 and 0 6 y 6 − βξ when ξ < 0.
The shape and scale parameters ξ and β can be estimated using Maximum Likelihood Estimation (MLE)
by fitting a GPD distribution to the tail of the return distribution after a given threshold u. Once this is done,
the CVaR can be computed:

VaRα + β − ξu
CVaRα = (5)
1−ξ
where the VaR for a GPD can be estimated by:
 −ξ !
β αN
VaRα = u + −1 (6)
ξ Nu

with N the total number of observations and Nu the number of observations exceeding the threshold u.
Note that the preceding result requires that observations be independent and identically distributed,
which is often not the case for daily returns as they present some level of autocorrelation (Figure2). There-
fore, we start by filtering the daily returns and then apply Extreme Value Theory to the standardised residuals
(see M C N EIL and F REY (2000); N YSTR ÖM and S KOGLUND (2005)), with a Generalised Pareto Distribu-
tion being fitted to the tails through Maximum Likelihood Estimation. Once this is done, we obtain the
shape and scale parameters and replace these values in Equation (5) to compute the CVaR at the required
confidence level. Extreme Value Theory has been used during the previous decade for risk management in
finance with a notable increase in the number of publications on the subject after the recent financial crisis.
We refer to BALI (2003); B EIRLANT et al. (2004); C ASCON and S HADWICK (2009); C OLES (2001); DE
H AAN and F ERREIRA (2006); E MBRECHTS (2011); G HORBEL and T RABELSI (2008, 2009); G OLDBERG
et al. (2008, 2010); G UMBEL (2004); H UANG et al. (2012); L ONGIN (2000); M C N EIL and F REY (2000);
N YSTR ÖM and S KOGLUND (2005) for a sample of publications dealing with Extreme Value Theory and its
applications to financial risk modelling.

3.2 Filtered Historical Simulation

Using Extreme Value Theory for tail risk estimation requires fitting a Generalised Pareto Distribution to
the left tail of the daily returns; in practice, if 250 days are considered and the 95% confidence level is
desired, this means that the GPD has to be fitted to about 12 daily returns, a number which is typically
too low to guarantee convergence of the Maximum Likelihood Estimation method and which will cause a

9
high sensitivity to changes in historical returns. To circumvent these issues, simulations can be employed,
generating a much larger number of daily returns to which left tail a GPD can be fitted more easily.
Choosing the appropriate simulation method is not necessarily straightforward. Indeed, if Monte Carlo
simulations (M ETROPOLIS and U LAM (1949)) are selected, a distribution of returns has to be specified,
usually a normal distribution, which negates the advantage of using Extreme Value Theory to estimate tail
risk. Therefore, some form of historical simulation is highly preferable as it makes no assumption on the
return distribution, instead relying on the past returns. However, as noted in P RITSKER (2006), such a
method presents two potential issues.
First, the required sample size to obtain a statistically significant distribution is usually considered to be
at least 250 days; this, in turn, raises the potential issue of not being sensitive enough to recent returns which
presumably contain the most relevant information to predict future returns. To circumvent this problem,
the weighted historical simulation (WHS) method was developed in B OUDOUKH et al. (1998); this method
assigns probabilistic weights to the daily returns which decay exponentially with a chosen decay factor over
time; thus recent returns have more influence than the more distant ones. Unfortunately, it is not clear how
to select the correct time constant; also, an unintended consequence is that extreme events, which by nature
occur rarely, might end up being discounted.
Second, the historical simulation method assumes that daily returns are independent and identically dis-
tributed through time, which is not particularly realistic. Indeed, it is commonly observed that the volatility
of returns evolves through time and that periods of high and low volatility do not occur at randomly spaced
intervals but rather tend to be clustered together. The filtered historical simulation (FHS) method presented
in BARONE -A DESI et al. (1999) is an attempt to combine the advantages of the historical and parametric
methods; the variance-covariance method attempts to capture conditional heteroskedasticity but assumes a
normal distribution while the historical method does not assume a specific distribution but does not cap-
ture conditional heteroskedasticity. The FHS method relies on a model based approach for the volatility,
while remaining model free in terms of the distribution. In particular, this method has the notable advantage
of being able to simulate extreme losses even if they are not present in the historical returns used for the
simulation.

3.3 Practical Implementation

We begin by implementing the FHS method on a series of daily returns Rt with standard deviation σt .
As mentioned above, the historical simulation method assumes that daily returns are i.i.d. through time;
however, significant autocorrelation can often be found in the daily squared returns. To produce a sequence
of i.i.d. observations, we fit an AR(1) first order autoregressive model to the daily returns:

Rt+1 = c + aRt + εt where εt = σt zt (7)

where we choose the standardised returns {zt } as following a Student’s t-distribution rather than a normal
one to account for increased tail risk as the t-distribution has fatter tails.
To model the distribution of the returns standard deviation, we use a GARCH type model (B OLLERSLEV
(1986, 1987); B OLLERSLEV et al. (1992); E NGLE (1982, 2001); TAYLOR (1986)) such as the GARCH(1,1):

2
σt+1 = ω + αε2t + βσt2 , with α + β < 1 (8)

10
 Alternately, the GARCH model can be replaced by its extension, the exponential GARCH (EGARCH)
model developed in N ELSON (1991); N ELSON and C AO (1992) to capture the asymmetry in volatility in-
duced by large positive and negative returns. Indeed, volatility usually increases more after a large drop than
after a large increase due to the leverage effect (B LACK (1976)). This model is defined by:
2
ln σt+1 = ω + α(φεt + γ(|εt | − E|εt |)) + β ln σt2 (9)

Once an AR(1)/GARCH(1,1) model has been fitted to the daily returns, the autocorrelation of the
squared returns is usually lower and these observations can now be used in a historical simulation method.
The i.i.d. property is important for bootstrapping, as it allows the sampling procedure to safely avoid the pit-
falls of sampling from a population in which successive observations are serially dependent. We simulate a
number of independent random trials (10,000 in this article) over a time horizon of 252 trading days; unlike
Monte Carlo simulations we do not make a specific distributional assumption regarding the standardised
returns {zt } and instead use the past returns data. Given a sequence of past returns, we can compute past
standardised returns from observed returns and estimated standard deviations as the quotient of the residual
of the AR(1) model over the standard deviation. Once the historical standardised returns are known, we gen-
erate future returns by drawing standardised returns with replacement. Eventually, we end up with 10,000
daily return series, each covering 252 trading days. These daily returns are aggregated to generate a distri-
bution of 2,520,000 daily returns and a GPD is fitted to the left tail, yielding the CVaR. The high number of
residuals ensures the stability of the method, as the left tail contains 126,000 returns for a 95% confidence
level, which almost guarantees the convergence of the Maximum Likelihood Estimation algorithm used to
fit the GPD. One of the advantages of using Extreme Value Theory to compute the CVaR is that the tail
risk of the return distribution is measured more accurately, being less likely to be underestimated than when
relying only on the volatility based method described earlier.

4 Tail Hedging Strategies

We apply the preceding technique to develop tail hedging strategies for the S&P 500 and the MSCI Emerging
Markets indexes. We target a pre-specified volatility level through time with the aim of maintaining the
volatility of the tail hedged portfolio around this level. In order to match a volatility target with the CVaR
computed using Extreme Value Theory, we convert the volatility into a CVaR number assuming a normal
distribution using Equation (3) and we set a maximum exposure level of 150%. Note that this normality
assumption is only used to compute the target CVaR from a target volatility in what represents a best case
scenario as the current CVaR of the index is computed using Extreme Value Theory.
The equity allocation is then given by:
 
Target CVaR
Equity Allocation = min , Max. Exposure (10)
Current CVaR
When the equity allocation is less than 100%, the remaining part of the portfolio is invested in cash,
and assumed to be earning the overnight USD LIBOR rate; similarly, when the equity allocation exceeds
100%, financing is assumed to cost the overnight USD LIBOR rate. The equity allocation is computed every
week, and to limit transaction cost impact, the actual equity allocation is only modified when it differs from
its previous value by more than ± 10%. This strategy will be referred to as Tail Hedged strategy in the
following.

11
 250
Tail Hedged S&P 500 12%
S&P 500 Risk Control 12%
S&P 500 Collar Index
200 S&P 500 Total Return

150

100

50
2001 2003 2005 2007 2009 2011 2013 2015

500

400

300

200

100 Tail Hedged MSCI EM

MSCI EM RIsk Control 12.5%
MSCI EM Total Return
0
2001 2003 2005 2007 2009 2011 2013 2015

Figure 5: Top: Comparison of the historical performance of the S&P 500 Total Return index, the S&P 500
Risk Control 12% index, the S&P 500 95-110 Collar index and the Tail Hedged S&P 500 12%. Bottom:
Comparison of the historical performance of the MSCI Emerging Markets Total Return index, the MSCI
EM Risk Control 12.5% index, and the Tail Hedged MSCI EM 12.5%.

12
 This process is applied to the S&P 500 Total Return index (Bloomberg ticker: SPTR Index) from
2001 to 2015 on with a 12% volatility target, and the performance of the resulting tail hedged strategy is
compared to that of the S&P 500 Daily Risk Control 12% index (Bloomberg ticker: SPXT12DT Index)
and the CBOE S&P 500 95-110 Collar index (Bloomberg ticker: CLL Index). The S&P 500 Daily Risk
Control 12% index is based on the S&P 500 Total Return index with an exposure dynamically adjusted
to target a 12% level of volatility, bounded by a maximum exposure level of 150%. The CBOE S&P 500
95-110 Collar index consists in holding the S&P 500 Total Return index while buying a 3 month 5% out of
the money put on the S&P 500, and financing this purchase by selling a 1 month 10% out of the money call.
These two indexes serve as benchmarks for the main tail hedging strategies currently available to investors:
allocation between equities and cash based on historical volatility, and options based strategies. The tail
hedging strategy is also applied to the MSCI Daily Total Return Net Emerging Markets index (Bloomberg
ticker: NDUEEGF Index) from 2001 to 2015, using the same method as for U.S. equities and targeting a
volatility of 12.5%. The performance of this tail hedged strategy is compared to the MSCI EM Risk Control
12.5% index which adjusts exposure based on historical volatility with a 150% limit similarly to the S&P
500 Daily Risk Control 12% index. The charts for the S&P 500 and MSCI Emerging Markets strategies are
presented in Figure 5. The annualised return, annualised volatility, maximum drawdown, 95% VaR, 95%
CVaR and Sharpe ratio for each of the strategies can be found in Tables 2, 3 and Tables 4, 5 respectively.

4.1 Performance analysis

The return data for both the S&P 500 TR and MSCI TR Emerging Markets demonstrate the effectiveness of
the managed volatility and tail risk approaches in maintaining volatility at the selected level. The annualised
volatilities over 2001–2015 for the S&P 500 Risk Control and Tail Hedged strategies are 11.95% and 11.88%
which are in line with the target of 12% annualised volatility; similarly, the MSCI EM Risk Control and
Tail Hedged strategies have annualised volatilities of 12.62% and 12.49% over 2001–2015, for a target of
12.5%. While not targeting a specific volatility level, the S&P 500 Collar index has an annualised volatility
of 11.69%. The yearly fluctuations in volatility are significantly reduced by the managed volatility and tail
hedged strategies. The volatility of the S&P 500 Total Return index varies widely from 10.05% to 41.02%
while the S&P 500 Risk Control index and Tail Hedged strategy volatility lie between 9.61% and 13.55%,
and between 6.83% and 17.09%, respectively. The S&P 500 Collar index experiences a wider range in
volatility with a low of 8.42% and a high of 15.88%, showing that options based strategies may not offer
as much volatility control. For the MSCI TR Emerging Markets index, the volatility varies from 11.35%
to 41.14% while the volatilities of the MSCI EM Risk Control index and Tail Hedge strategy are bounded
between 11.46% and 13.20% and 8.28% and 15.38%, respectively.
This significant reduction in volatility is not necessarily followed by a decreased return as would be ex-
pected; in fact, the Tail Hedged strategy generate higher returns than the S&P 500 TR index. The annualised
return for the 2001–2015 period is 4.99% for the S&P 500 TR, 4.76% for the S&P 500 Risk Control index
and 5.69% for the Tail Hedged strategy. For the MSCI TR Emerging Markets index, the annualised return is
8.37% while the MSCI EM Risk Control index and Tail Hedged strategy registered 6.71% and 6.98%. The
Sharpe ratios generated by the Risk Control and Tail Hedged strategies are higher than the original index
ratios by a factor of 2 for the S&P 500 TR and by 25% for the MSCI TR Emerging Markets index. This
consistent outperformance of tail hedging strategies can be explained by the limitations placed on many long
only managers in terms of maximum cash levels, which are usually set around 10% or 20%. This means
that this potential market inefficiency is not accessible to many investors.

13
 100
Turnover Tail Hedged S&P 500 12%
Turnover Tail Hedged MSCI EM 12.5%
Annualised Turnover (%)

80

60

40

20

0
2002 2004 2006 2008 2010 2012 2014

140
Tail Hedged S&P 500 12%
Tail Hedged MSCI EM 12.5%
120
Equity Allocation (%)

100

80

60

40

20
2001 2003 2005 2007 2009 2011 2013 2015

Figure 6: Top: Historical evolution of the annualised turnover over a rolling 6 months window for the tail
hedged strategies. Bottom: Historical evolution of the equity allocation for the tail hedged strategies.

14
 45
Credit Suisse Fear Barometer
40 Average Value 2001 - 2004
Average Value 2005 - 2015
35

30

25

20

15

10
2001 2003 2005 2007 2009 2011 2013 2015

Figure 7: Historical evolution of the Historical evolution of the Credit Suisse Fear Barometer index.

However, the annualised return of the S&P 500 Collar index is lower at 2.97% showing the cost of an
options based strategy. In fact, while the S&P 500 Collar index outperforms the S&P 500 Risk Control
index and Tail Hedged strategy during 2001–2003, from 2004 to 2015 its annualised return is significantly
lower than for the S&P 500 Risk Control index and Tail Hedged strategy. One explanation for this perfor-
mance difference over these two periods is apparent when considering the Credit Suisse Fear Barometer
(Bloomberg ticker: CSFB Index); this index measures investment sentiment by pricing a zero-cost collar,
implemented by selling a 10% out of the money S&P 500 call option to purchase an out of the money put.
The index level represents how far out of the money the purchased S&P 500 put is. As can be seen on Figure
7, for the 2001–2004 period, the average value of the index is 15.45 meaning that selling a 10% out of the
money call would on average finance the purchase of a 15.45% out of the money put. During the 2005–2015
period, the average value of the index is noticeably higher at 23.68. Such an increase in the price of puts
compared to calls on the S&P 500 would undoubtedly lead to lower returns for the S&P 500 Collar index
since the difference between the price of the 3 month 5% out of the money put and the 1 month 10% out
of the money call would increase and act as a drag on performance. As long as the price of put options
remains elevated compared to call options, cash based strategies will tend to outperform their option based
counterparts.
The managed volatility and tail hedged strategies also offer substantial drawdown protection with a
reduction from an original maximum drawdown of 55.25% for the S&P 500 TR to 27.69% for the S&P 500
Risk Control index and 25.35% for the Tail Hedged strategy; the situation is similar in Emerging Markets
where an original 65.25% maximum drawdown sustained by the MSCI TR Emerging Markets is limited
to 34.69% for the MSCI EM Risk Control index and 33.62% for the Tail Hedged index. This drawdown

15
protection seems to come from a reduction in tail risk (measured by 95% CVaR) during the worst years. For
example, in 2008, the 95% CVaR was 6.44% for the S&P 500 TR while it was more than three times lower
for the S&P 500 Risk Control index (1.91%) and the Tail Hedged strategy (1.97%). The decrease in tail
risk is even more pronounced in Emerging Markets; for the MSCI TR Emerging Markets, the 95% CVaR
in 2008 is 6.60% compared to 1.88% for the MSCI EM Risk Control index and 2.21% for the Tail Hedged
strategy.
The historical evolution of the equity allocation and the annualised turnover over a rolling 6 months
window for each of the Tail Hedged strategies are plotted on Figure 6. Due to the rebalancing rule which
requires a change greater than 10% in absolute value, the allocation remains constant for extended periods
of time, thereby limiting the potential impact of transaction costs. The turnover remains below 100% most
of the time and its average value is 18% for the Tail Hedged S&P 500 and 17% for the Tail Hedged MSCI
EM. This confirms the potential practical applications of the Tail Hedged strategies and their investability.

5 Conclusion
This article develops a novel tail hedging strategy based on tail risk estimation using Extreme Value Theory
and compares this method with other cash and options based techniques. The primary objectives of these
strategies are to protect investors against the significant drawdowns that occur in equity markets and to limit
the variations in volatility. The implementation of these techniques on U.S. and Emerging Markets equity
indexes during 2001–2015 confirm their ability to manage volatility and reduce drawdowns, with the options
based strategy being less effective than its cash based counterparts. An added benefit of tail hedging is a
noticeable improvement in risk adjusted return measured in terms of the Sharpe ratio provided by the cash
based strategies; the relatively poor performance of the options based strategy is found to occur during the
2004–2015 period and is mostly due to a significant increase in the price of puts with respect to calls which
acts as a performance drag. This illustrates one of the main drawbacks of using options for tail hedging,
namely that the cost of put options seems to have increased in recent years and will hamper the performance
of such strategies as long as this situation persists.
On the other hand, the cash based strategies are not as sensitive to such changes in market conditions
and provide a more consistent risk and return profile. The tail hedging strategy introduced in this article,
which focuses on tail risk, appears to outperform the volatility based strategies in terms of risk adjusted
return and drawdown during the 2008 financial crisis. Tail risk measured as 95% CVaR seems to give an
earlier warning of major downturns than historical volatility which may explain the better performance of
the tail risk based strategy. The exposure adjustments for this strategy are relatively infrequent; as a result,
adding a transaction cost should have a minimal impact on the overall performance and one would expect
the performance of this strategy to remain similar to the backtested results when implemented in practice.

16
 Ann. Return (%) Ann. Volatility (%) Max. Drawdown (%)
S&P TR Collar Risk Ctrl Tail Hedge S&P TR Collar Risk Ctrl Tail Hedge S&P TR Collar Risk Ctrl Tail Hedge

2001 -11.20 4.92 -8.88 -4.54 21.39 14.93 11.49 11.26 29.09 14.47 19.12 15.66

2002 -22.15 -10.55 -14.16 -13.73 26.06 13.77 12.39 14.74 33.01 18.51 19.28 20.76

2003 27.20 15.40 17.59 15.42 17.07 12.54 10.56 9.23 13.78 11.10 7.10 7.40

2004 12.37 5.40 8.87 12.36 11.16 9.11 11.75 10.39 7.42 7.90 9.19 6.71

2005 6.70 2.69 3.24 7.39 10.26 8.40 11.81 12.22 7.01 6.20 8.25 8.12

2006 14.37 11.76 19.11 14.89 10.05 8.64 11.57 12.04 7.46 6.84 8.36 9.42

2007 7.53 0.78 5.11 5.89 15.95 10.68 13.22 14.42 9.87 7.54 7.28 9.49

2008 -40.45 -25.21 -15.58 -9.17 41.02 15.09 13.19 17.09 48.71 28.89 17.86 15.08

17
2009 32.26 19.92 15.05 8.15 27.45 15.88 9.61 6.83 27.19 15.10 7.30 7.31

2010 13.80 3.47 10.74 7.66 18.12 11.93 11.71 10.01 15.63 13.71 9.69 9.23

2011 -0.76 -9.58 -2.98 -2.00 23.02 13.66 13.55 14.23 18.65 17.50 15.12 14.10

2012 22.78 9.18 8.91 11.69 12.98 8.73 10.60 7.06 9.58 8.03 8.55 5.06

2013 26.70 21.12 28.20 24.31 11.22 9.64 11.84 10.29 5.58 5.63 5.21 5.12

2014 11.19 10.65 10.71 10.11 10.93 8.42 12.56 10.63 7.28 4.79 7.99 7.64

2015 3.66 -3.69 -6.53 2.18 15.34 9.60 12.86 13.05 12.04 9.23 13.14 11.21

2001–2015 4.99 2.97 4.76 5.69 19.90 11.69 11.95 11.88 55.25 38.52 27.69 25.35

Table 2: Annualised return, annualised volatility and maximum drawdown for the S&P 500 Total Return index, the S&P 500 95-110
Collar index, the Daily Risk Control 12% index and the Tail Hedged S&P 500 12%.
 95% VaR (%) 95% CVaR (%) Sharpe Ratio
S&P TR Collar Risk Ctrl Tail Hedge S&P TR Collar Risk Ctrl Tail Hedge S&P TR Collar Risk Ctrl Tail Hedge

2001 2.22 1.57 1.19 1.17 2.92 1.77 1.63 1.53 -0.52 0.33 -0.77 -0.40

2002 2.48 1.38 1.24 1.50 3.13 1.72 1.54 1.82 -0.85 -0.77 -1.14 -0.93

2003 1.52 1.16 0.95 0.81 2.10 1.47 1.27 1.12 1.59 1.23 1.67 1.67

2004 1.29 0.94 1.20 1.14 1.43 1.17 1.55 1.31 1.11 0.59 0.75 1.19

2005 1.02 0.81 1.22 1.23 1.24 0.91 1.51 1.49 0.65 0.32 0.27 0.60

2006 1.00 0.88 1.09 1.21 1.33 1.14 1.58 1.64 1.43 1.36 1.65 1.24

2007 1.80 1.08 1.47 1.64 2.50 1.68 2.11 2.29 0.47 0.07 0.39 0.41

2008 4.68 1.62 1.49 1.80 6.44 2.35 1.91 1.97 -0.99 -1.67 -1.18 -0.54

18
2009 2.91 1.57 1.11 0.72 3.88 2.02 1.31 0.96 1.18 1.25 1.57 1.19

2010 1.71 1.32 1.45 0.98 2.68 1.80 1.81 1.47 0.76 0.29 0.92 0.77

2011 2.50 1.59 1.50 1.31 3.53 1.95 2.22 2.23 -0.03 -0.70 -0.22 -0.14

2012 1.26 0.81 1.14 0.69 1.69 1.06 1.52 0.96 1.76 1.05 0.84 1.66

2013 1.20 1.00 1.24 1.10 1.55 1.40 1.66 1.42 2.38 2.19 2.38 2.36

2014 1.24 0.84 1.46 1.19 1.72 1.21 2.03 1.69 1.02 1.26 0.85 0.95

2015 1.45 0.98 1.27 1.24 2.19 1.30 1.98 1.87 0.24 -0.38 -0.51 0.17

2001–2015 1.91 1.24 1.24 1.22 4.50 2.49 2.62 2.54 0.25 0.25 0.40 0.48

Table 3: 95% VaR, 95% CVaR and Sharpe ratio for the S&P 500 Total Return index, the S&P 500 95-110 Collar index, the Daily Risk
Control 12% index and the Tail Hedged S&P 500 12%.
 Ann. Return (%) Ann. Volatility (%) Max. Drawdown (%)
MSCI EM TR Risk Ctrl Tail Hedge MSCI EM TR Risk Ctrl Tail Hedge MSCI EM TR Risk Ctrl Tail Hedge

2001 6.76 -1.99 8.73 15.07 12.43 10.21 26.54 24.44 16.70

2002 -14.04 -13.08 -10.83 17.39 12.77 12.43 28.81 22.23 22.05

2003 63.46 58.97 52.08 12.61 12.08 11.65 4.92 4.60 5.29

2004 17.33 23.21 8.42 14.84 13.19 12.52 16.36 12.83 16.46

2005 55.40 52.09 56.95 12.71 12.57 13.27 8.95 9.20 10.28

2006 15.23 13.25 7.57 18.47 13.02 13.43 24.28 15.48 21.23

2007 16.68 11.05 13.35 22.97 13.20 14.39 22.03 12.57 13.82

2008 -53.03 -25.00 -22.70 41.14 13.03 14.49 63.14 32.17 30.79

19
2009 94.28 30.02 25.26 23.49 11.46 8.28 12.94 9.49 5.86

2010 20.98 14.43 13.23 17.16 11.85 11.40 17.93 11.47 12.28

2011 -3.40 -3.58 -3.70 22.50 12.88 14.78 30.06 20.00 20.56

2012 2.73 1.87 4.31 14.02 11.61 8.43 17.48 15.76 10.05

2013 -6.89 -13.08 -6.53 13.35 12.67 11.76 16.56 18.65 15.21

2014 1.27 -0.84 -1.24 11.35 13.00 12.03 17.04 20.52 19.57

2015 -15.67 -14.69 -13.76 16.02 12.80 15.38 26.73 25.15 26.56

2001–2015 8.37 6.71 6.98 19.61 12.62 12.49 65.25 34.69 33.62

Table 4: Annualised return, annualised volatility and maximum drawdown for the MSCI Daily Total Return Net Emerging Markets
index, the MSCI EM Risk Control 12.5% index, and the Tail Hedged MSCI EM 12.5%.
 95% VaR (%) 95% CVaR (%) Sharpe Ratio
MSCI EM TR Risk Ctrl Tail Hedge MSCI EM TR Risk Ctrl Tail Hedge MSCI EM TR Risk Ctrl Tail Hedge

2001 1.50 1.43 1.11 2.17 1.92 1.43 0.45 -0.16 0.86

2002 1.84 1.35 1.26 2.57 1.82 1.88 -0.81 -1.02 -0.87

2003 1.07 1.09 0.99 1.52 1.47 1.51 5.03 4.88 4.47

2004 1.68 1.57 1.25 2.24 2.05 2.04 1.17 1.76 0.67

2005 1.24 1.08 1.32 1.78 1.79 1.89 4.36 4.14 4.29

2006 1.96 1.38 1.45 3.08 2.16 2.41 0.82 1.02 0.56

2007 2.38 1.29 1.69 3.68 2.16 2.40 0.73 0.84 0.93

2008 4.63 1.49 1.60 6.60 1.88 2.21 -1.29 -1.92 -1.57

20
2009 1.90 1.06 0.71 2.68 1.58 1.00 4.01 2.62 3.05

2010 1.91 1.41 1.28 2.48 1.62 1.66 1.22 1.22 1.16

2011 2.58 1.50 1.51 3.24 2.00 2.18 -0.15 -0.28 -0.25

2012 1.45 1.18 0.81 1.90 1.55 1.06 0.19 0.16 0.51

2013 1.25 1.34 1.15 1.83 1.92 1.57 -0.52 -1.03 -0.56

2014 1.21 1.40 1.35 1.51 1.80 1.65 0.11 -0.06 -0.10

2015 1.61 1.16 1.55 2.23 1.91 2.07 -0.98 -1.15 -0.90

2001–2015 1.88 1.35 1.27 4.42 2.81 2.82 0.43 0.53 0.56

Table 5: 95% VaR, 95% CVaR and Sharpe ratio for the MSCI Daily Total Return Net Emerging Markets index, the MSCI EM Risk
Control 12.5% index, and the Tail Hedged MSCI EM 12.5%.
References
A KGIRAY, V. (1989). Conditional heteroskedasticity in time series of stock returns: Evidence and forecasts.
Journal of Business, 62(1):55–80.

A NG, A., H ODRICK, R. J., X ING, Y., and Z HANG, X. (2006). The cross-section of volatility and expected
returns. Journal of Finance, 61(1):259–299.

A NG, A., H ODRICK, R. J., X ING, Y., and Z HANG, X. (2009). High idiosyncratic volatility and low returns:
International and further U.S. evidence. Journal of Financial Economics, 91(1):1–23.

A RTZNER, P., D ELBAEN, F., E BER, J. M., and H EATH, D. (1999). Coherent measures of risk. Mathematical
Finance, 9(3):203–228.

BAKSHI, G. and K APADIA, N. (2003). Delta-hedged gains and the negative market volatility risk premium.
Review of Financial Studies, 16(2):527–566.

BALI, T. G. (2003). An extreme value approach to estimating volatility and value at risk. Journal of
Business, 76(1):83–108.

BALKEMA, A. and DE H AAN, L. (1974). Residual life time at great age. Annals of Probability, 2(5):792–
804.

BANERJEE, A. and S RIVASTAVA, V. (2012). Limiting Risk Exposure with S&P Risk Control Indices. S&P
Dow Jones Indices White Paper.

BARONE -A DESI, G., G IANNOPOULOS, K., and VOSPER, L. (1999). VaR without correlations for portfolios
of derivative securities. Journal of Futures Markets, 19(5):583–602.

B EDER, T. S. (1995). VaR: Seductive but dangerous. Financial Analysts Journal, 51(5):12–24.

B EIRLANT, J., G OEGEBEUR, Y., S EGERS, J., and T EUGELS, J. (2004). Statistics of Extremes: Theory and
Applications. Wiley.

B LACK, F. (1972). Capital market equilibrium with restricted borrowing. Journal of Business, 45(3):444–
455.

B LACK, F. (1976). Studies in stock price volatility changes. Proceedings of the 1976 American Statistical
Association, Business and Economic Statistics Section., pages 177–181.

B LACK, F. and J ONES, R. C. (1987). Simplifying portfolio insurance. Journal of Portfolio Management,
14(1):48–51.

B LACK, F. and P EROLD, A. F. (1992). Theory of constant proportion portfolio insurance. Journal of
Economic Dynamics and Control, 16(3–4):403–426.

B LITZ, D. C. and VAN V LIET, P. (2007). The volatility effect. Journal of Portfolio Management, 34(1):102–
113.

21
B OLLERSLEV, T. (1986). Generalized autoregressive conditional heteroskedasticity. Journal of Economet-
rics, 31(3):307–327.

B OLLERSLEV, T. (1987). A conditionally heteroskedastic time series model for speculative prices and rates
of return. Review of Economics and Statistics, 69(3):542–547.

B OLLERSLEV, T., C HOU, R., and K RONER, K. F. (1992). ARCH modeling in finance: A review of the
theory and empirical evidence. Journal of Econometrics, 52(1–2):5–59.

B OUDOUKH, J., R ICHARDSON, M., and W HITELAW, R. (1998). The best of both worlds: A hybrid ap-
proach to calculating value at risk. Risk, 11(5):64–67.

B RAILSFORD, T. J. and FAFF, R. W. (1996). An evaluation of volatility forecasting techniques. Journal of

Banking and Finance, 20(3):419–438.

B RENNAN, M. J. and S CHWARTZ, E. S. (1989). Portfolio insurance and financial market equilibrium.
Journal of Business, 62(4):455–472.

B RENNAN, M. J. and S OLANKI, R. (1981). Optimal portfolio insurance. Journal of Financial and Quanti-
tative Analysis, 16(3):279–300.

C AMPBELL, J. Y. and H ENTSCHEL, L. (1992). No news is good news: An asymmetric model of changing
volatility in stock returns. Journal of Financial Economics, 31(3):281–318.

C ASCON, A. and S HADWICK, W. F. (2009). A new approach to tail risk. Journal of Investment Consulting,
10(1):33–48.

C HONG, J. and P HILLIPS, G. M. (2012). Low-(economic) volatility investing. Journal of Wealth Manage-
ment, 15(3):75–85.

C HRISTOFFERSEN, P. F. (2003). Elements of Financial Risk Management. Academic Press.

C LARKE, R. G., DE S ILVA, H., and T HORLEY, S. (2006). Minimum-variance portfolios in the U.S. equity
market. Journal of Portfolio Management, 33(1):10–24.

C LARKE, R. G., DE S ILVA, H., and T HORLEY, S. (2010). Know your VMS exposure. Journal of Portfolio
Management, 36(2):52–59.

C LARKE, R. G., DE S ILVA, H., and T HORLEY, S. (2011). Minimum-variance portfolio composition. Journal
of Portfolio Management, 37(2):31–45.

C OLES, S. (2001). An Introduction to Statistical Modeling of Extreme Values. Springer.

DE H AAN, L. and F ERREIRA, A. (2006). Extreme Value Theory: An Introduction. Springer.

D UFFIE, D. and PAN, J. (1997). An overview of value at risk. Journal of Derivatives, 4(3):7–49.

E MBRECHTS, P. (2011). Modelling Extremal Events. Springer.

E NGLE, R. F. (1982). Autoregressive conditional heteroskedasticity with estimates of the variance of United
Kingdom inflation. Econometrica, 50(4):987–1007.

22
E NGLE, R. F. (2001). GARCH 101: The use of ARCH/GARCH models in applied econometrics. Journal
of Economic Perspectives, 15(4):157–168.

FAMA, E. F. and F RENCH, K. R. (1992). The cross-section of expected stock returns. Journal of Finance,
47(2):427–465.

FAMA, E. F. and F RENCH, K. R. (1993). Common risk factors in the returns on stocks and bonds. Journal
of Financial Economics, 33(1):3–56.

FAMA, E. F. and F RENCH, K. R. (2004). The capital asset pricing model: Theory and evidence. Journal of
Economic Perspectives, 18(3):25–46.

F LEMING, J., K IRBY, C., and O STDIEK, B. (2001). The economic value of volatility timing. Journal of
Finance, 56(1):329–352.

F LEMING, J., K IRBY, C., and O STDIEK, B. (2003). The economic value of volatility timing using “realized”
volatility. Journal of Financial Economics, 67(3):474–509.

F RENCH, K. R., S CHWERT, G. W., and S TAMBAUGH, R. F. (2004). Expected stock returns and volatility.
Journal of Financial Economics, 19(1):3–29.

G HORBEL, A. and T RABELSI, A. (2008). Predictive performance of conditional Extreme Value Theory in
Value-at-Risk estimation. International Journal of Monetary Economics and Finance, 1(2):121–148.

G HORBEL, A. and T RABELSI, A. (2009). Measure of financial risk using conditional extreme value copulas
with EVT margins. Journal of Risk, 11(4):51–85.

G IESE, G. (2012). Optimal design of risk control strategy indexes. Journal of Indexes, 15(5):44–51.

G LOSTEN, L. R., JAGANNATHAN, R., and RUNKLE, D. E. (1993). On the relation between the expected
value and the volatility of the nominal excess return on stocks. Journal of Finance, 48(5):1779–1801.

G OLDBERG, L. R., M ENCHERO, J., H AYES, M. Y., and M ITRA, I. (2010). Extreme risk analysis. Journal
of Performance Measurement, 14(3):17–30.

G OLDBERG, L. R., M ILLER, G., and W EINSTEIN, J. (2008). Beyond value-at-risk: forecasting portfolio
loss at multiple horizons. Journal of Investment Management, 6(2):73–98.

G UMBEL, E. J. (2004). Statistics of Extremes. Dover.

H ARMANTZIS, F. C., M IAO, L., and C HIEN, Y. (2006). Empirical study of value-at-risk and expected
shortfall models with heavy tails. Journal of Risk Finance, 7(2):117–135.

H UANG, W., L IU, Q., R HEE, S. G., and W U, F. (2012). Extreme downside risk and expected stock returns.
Journal of Banking & Finance, 36(5):1492–1502.

J OHANSEN, A. and S ORNETTE, D. (2001). Large stock market price drawdowns are outliers. Journal of
Risk, 4(2):69–110.

J ORION, P. (2006). Value at Risk. McGraw-Hill.

23
L ELAND, H. E. (1980). Who should buy portfolio insurance? Journal of Finance, 35(2):581–594.

L INTNER, J. (1965). The valuation of risk assets and the selection of risky incvestments in stock portfolios
and capital budgets. Review of Economics and Statistics, 47(1):13–37.

L ONGIN, F. M. (2000). From value at risk to stress testing: the extreme value approach. Journal of Banking
& Finance, 24(7):1097–1130.

M ANDELBROT, B. (1963). The variation of certain speculative prices. Journal of Business, 36(4):394–419.

M ANDELBROT, B. (1967). The variation of some other speculative prices. Journal of Business, 40(4):393–
413.

M C N EIL, A. J. and F REY, R. (2000). Estimation of tail-related risk measures for heteroscedastic financial
time series: an extreme value approach. Journal of Empirical Finance, 7(3–4):271–300.

M C N EIL, A. J., F REY, R., and E MBRECHTS, P. (2005). Quantitative Risk Management : Concepts, Tech-
niques, and Tools. Princeton University Press.

M ETROPOLIS, N. and U LAM, S. (1949). The Monte Carlo method. Journal of the American Statistical
Association, 44(247):335–341.

N ELSON, D. B. (1991). Conditional heteroskedasticity in asset returns: A new approach. Econometrica,

59(2):347–370.

N ELSON, D. B. and C AO, C. Q. (1992). Inequality constraints in the univariate GARCH model. Journal of
Business and Economic Statistics, 10(2):229–235.

N YSTR ÖM, K. and S KOGLUND, J. (2005). Efficient filtering of financial time series and Extreme Value
Theory. Journal of Risk, 7(2):63–84.

P EROLD, A. and S HARPE, W. F. (1988). Dynamic strategies for asset allocation. Financial Analysts Journal,
44(1):16–27.

P ICKANDS, J. (1975). Statistical inference using extreme order statistics. Annals of Statistics, 3:119–131.

P OON, S.-H. and G RANGER, C. W. J. (2003). Forecasting volatility in financial markets: A review. Journal
of Economic Literature, 41(2):478–539.

P RITSKER, M. (2006). The hidden dangers of historical simulation. Journal of Banking and Finance,
30(2):561–582.

RUBINSTEIN, M. and L ELAND, H. E. (1981). Replicating options with positions in stock and cash. Finan-
cial Analysts Journal, 37(4):63–72.

RUSSELL I NVESTMENTS and P&I (2012). De-risking Risk: A Three-Part Survey of Institutional Investors.
Available at http://supplement.pionline.com/risk-trending/index.php.

S HARPE, W. F. (1964). Capital asset prices: a theory of market equilibrium under conditions of risk. Journal
of Finance, 19(3):425–442.

24
TAYLOR, S. J. (1986). Modeling Financial Time Series. Wiley.

T HOMAS, M. and C OLLIE, R. (2012). Volatility Management. Russell Viewpoint.

T HOMAS, R. and S HAPIRO, R. (2009). Managed volatility: A new approach to equity investing. Journal of
Investing, 18(1):15–23.

YAN, X. (2012). Lending certainty to volatility: Russell’s new Volatility Control Indexes. Russell Research.

25
About the author:

DR ISSAM STRUB: Dr Strub is a senior member of the Cambridge Strategy research group where he
works on quantitative strategies as well as asset allocation and risk management tools; he has authored a
number of research articles in financial and scientific journals and has been an invited speaker at financial
conferences and roundtables. Prior to joining the Cambridge Strategy, Dr Strub was a graduate student at the
University of California, Berkeley, where he conducted research in an array of fields ranging from Partial
Differential Equations and Fluid Mechanics to Scientific Computing and Optimisation; he obtained a Ph.D.
in Engineering from the University of California in 2009.

Disclaimer: Some services are not available to private inexperienced investors. Services may also not be available to certain investors due to regulatory or other constraints either in the UK or
elsewhere. The information contained herein is not targeted at the residents of any particular country and is not intended for distribution to, or use by, any person in any country or jurisdiction where
such distribution or use would be contrary to local law or regulatory requirements. Information contained herein may be subject to change without notice. Investments in securities for financial
instruments (which include contracts for differences, futures, options spot and forward foreign exchange and off-exchange contracts) can fluctuate in value and you should be aware that you may
not realise the initial amount invested and may incur additional liabilities. As investments in securities or financial instruments may entail above average risks you should carefully consider whether
your financial circumstances permit you to invest and if necessary seek the advice of an independent financial adviser. Foreign exchange denominated securities and financial instruments are subject
to fluctuations in exchange rates that may have a positive or negative effect on the value, price or income derived from the securities or financial instrument concerned. Past performance is not a
reliable indicator of future performance. You are advised that The Cambridge Strategy (Asset Management) Limited is unable to provide advice as to tax consequences of a particular investment
or investment strategy and you are advised to seek professional advice in this respect. This document is issued by the Cambridge Strategy (Asset Management) Limited and is not a solicitation
or instruction to invest. It is provided for information purposes only. The Cambridge Strategy (Asset Management) Limited is authorised and regulated by the Financial Conduct Authority and
registered with the SEC (US) and the SFC (HK). The Cambridge Strategy (Asset Management) Limited is registered with the CFTC as a commodity pool operator and commodity trading advisor.
The Cambridge Strategy (Asset Management) Limited is exempt from the requirement to hold an Australian financial services licence under the Corporations Act 2001 (Cth) (Class Order 03/1099)
in respect of the provision of financial services. It is regulated by the Financial Conduct Authority (FCA) under UK laws, which differ from Australian laws. Australian Investors: These materials
are provided by a representative of The Cambridge Strategy (Asset Management) Limited (’Cambridge’) and is intended for wholesale clients as defined in the Corporations Act 2001 (Cth). The
information in this presentation is current unless stated otherwise and Cambridge is not under any obligation to update the information to the extent that it is or becomes out of date or incorrect.
It is confidential and has been prepared by Cambridge solely for use in connection with its Programmes. This information must not be made available, published or distributed to any third party
without the prior consent of Cambridge. This information has been prepared without taking into account anyone’s objectives, financial situation or needs so before acting on it each person should
consider its appropriateness to their circumstances before making any investment decision. In particular, a person should consider the Programme’s investment objectives, risks, fees and other
charges. Each person should carefully read and consider any offer documentation before making an investment decision. The information in this presentation is indicative and may change with
market fluctuations. It does not purport to be a comprehensive statement or description of any markets or securities referred to within. Cambridge assumes no fiduciary responsibility or liability for
any consequences financial or otherwise arising from any reliance on this information. Each person should make their own appraisal of the risks and should consult to the extent necessary, their
own legal, financial, tax, accounting and other professional advisors in this respect to any investment in the Programmes. This document is not a solicitation or instruction to invest. It is provided
for information purposes only.

26