Are you sure?
This action might not be possible to undo. Are you sure you want to continue?
∗
Gilles CRITON
†
and Olivier SCAILLET
‡
Second version, September 2009
Abstract
We propose a timevarying coeﬃcient model in order to analyze the dynamic in estimated
alpha and betas. We showed that the proportion of “skilled”funds are higher with our model
than with a static linear factor model. Indeed, our timevarying coeﬃcient model captures the
dynamic part of alpha that reﬂects the dynamic strategy that we can ﬁnd within Hedge Funds.
Furthermore this model is not only capable of looking into anticipation for Hedge Fund managers
but is equally well suited for the analysis of beta exposure. We show that whatever the strategy,
the increase in risk behavior is mainly concentrated on the credit spread risk factor and bond
risk factor.
∗
We thank J´erˆ ome Teiletche as well as seminar participants at Pictet, Unigestion, SoFiE (2009).
†
‡
1
1 Introduction
Nowadays, it is well known that investing in Mutual Funds, on average, underperform passive
investment strategies. One class commonly called Hedge Funds are deﬁned as pooled invest
ment vehicles that are privately organized, and not widely available to the general investing
public. Hedge Funds are private partnerships that use advanced investment strategies and can
use derivatives, leverage and short selling. Over the last few years, Hedge Funds have become
the favorites of many private as well as institutional investors. Hedge Funds follow investment
strategies that are substantially diﬀerent from the nonleveraged, longonly strategies conven
tionally followed by investors. During alternative investment seminars and conferences, Hedge
Fund managers boast about their ability to produce something they refer to as “alpha”or “ab
solute return”in the sense that performance is not due to primary asset classes performance.
They do not aim to track and try to beat a certain benchmark, but instead are focused on
pure return generation. Hence Hedge Funds generate alpha, as opposed to depending on beta
and performance should normally result from active management decisions combined with the
skills of their advisors. However statistical analysis shows that many of them retain signiﬁcant
exposure to diﬀerent types of market risk factors.
Consequently, it is essential for qualiﬁed investors to determine whether or not these strategies
are sensitive to the market and whether they can ﬁnd alpha through the manager’s skill. For
these reasons, an increasing focus has been directed to the performance of Hedge Funds and
their factor exposures.
Mostly, owing to the theory of CAPM or APT, fund performances are assessed by a parametric
model with the hypothesis of normality and linearity of coeﬃcients, as well as, the nondynamic
behavior of beta. Some researchers expanded the parametric model to the world of Hedge
Funds. Fung and Hsieh (2001, 2004a) developed factors used to replicate trendfollowing
strategies. Agarwal and Naik (2000) suggested an approach using optionbased returns in
order to capture the nonlinearity in beta. Recently, Bollen and Whaley (2008) tested for
structural change and used two econometric models in order to capture the dynamic in beta.
In this paper, we attempt to go a step further than the previous research.
First, by developing the result from Bollen and Whaley (2008) by testing up to ﬁve multiple
structural changes. We showed that the relative number of breaks by fund has increased over
the last few years, resulting in the increase of the use of leverage.
2
Secondly, by developing a new model
1
for Hedge Funds which can take into consideration their
distinctly nonnormal characteristics, the dynamic in manager skill, as well as, the dynamic
trading represented respectively by alpha and beta. This model allows us to relax traditional
parametric models and to exploit possible hidden structure. For example, what manager skill
is the most important? Is it the skill to pick and choose the right stocks, bonds, any other
ﬁnancial products, or is it the skill to anticipate market events? If we consider that alpha
estimated from a linear model contains the two skills, how can we separate and study them?
Can we ﬁnd a diﬀerent proportion of positivealpha funds? This model speciﬁcally allows us to
look into how managers behave in relation to strong events and whether or not they succeed
to generate alpha. Moreover, Hedge Fund managers are likely to change trading strategies
in order to obtain “absolute return”
2
. Therefore, the exposure to a risk factor can change
during special events and can have some deep implications for a Fund of Funds or a portfolio.
Often it has been merely stated that during an event, risk to a speciﬁc factor can increase or
decrease. Our study also portrays how the beta exposure or risk behaves in relation to market
reaction, as well as, alpha during two agitated periods; the equity bubble crisis (the Equity
Bubble hereafter)
3
and LTCM crisis (LTCM hereafter)
4
.
Thirdly, with the methodology used for mutual Funds by Barras, Scaillet and Wermers (2008)
called False Discovery Rate, we look into the proportion of the fund’s population which has
an increase in diﬀerent market exposures, as well as, the proportion of skilled, unskilled and
zeroalpha funds.
We showed that the proportion of “skilled”funds are higher with our model than with a static
linear factor model. We explained such a diﬀerence by our model’s ability to capture the
dynamic part of alpha that reﬂects Hedge Fund managers’ market timing ability. We found
that the quantity of new alpha substantially increased the proportion of positive and negative
alpha funds. Nevertheless, we showed that the majority of Hedge Funds are zeroalpha funds
as Barras, Scaillet and Wermers (2008) portrayed for Mutual Funds. Furthermore, we look
into the possibility for a strategy to have a common increasing trend to a market exposure
during a crisis. For all strategies, the main result was a common trend to an increase in the
credit spread risk factor.
1
This timevarying coeﬃcient model has been fully explored by the seminal work of Cleveland, Grosse and Shyu
(1991).
2
positive returns in all market conditions, up or down.
3
February, March 2000
4
July, Auguste, and September 1998
3
To the best of our knowledge, this is the ﬁrst paper which tries to look into how the risk of
Hedge Funds and CTAs exposure changes. This paper also looks into the skill of manager
anticipation, part of alpha.
The rest of our paper is organized as follows; Section 2 reviews related literature about Hedge
Funds modeling, and the dynamic in beta
5
. The data is described in section 3. Section 4
summarizes the risk factors deﬁned in Fung and Hsieh’s paper (2001, 2004). Section 5 tests
for structural change by applying the method of Bai and Perron (1998). Section 6 deﬁnes
our timevarying coeﬃcient model. Our methodology in which we apply our model is deﬁned
in section 7. Section 8 provides results of our timevarying coeﬃcient model, as well as, the
application of the False discovery Rate (FDR hereafter) to alpha and beta bringing, in this
way, a new tool for Hedge Fund analysis. Section 9 presents our conclusion.
2 Literature review
It is wellknown that there are some drawbacks to model Hedge Funds. Various fundamental
and statistical multifactor models for Hedge Funds have been analyzed by Agarwal and Naik
(2000), Edwards and Caglayan (2001), Fung and Hsieh (1997, 2001, 2004a), Lhabitant (2001),
Liang (2001), Schneeweis and Spurgin (1998), among others.
Research has speciﬁcally been focused on identifying dynamic trading strategies in order to
mimic the actual trades of Hedge Funds (for example: Fung and Hsieh (1997), Agarwal and
Naik (2000)). Contrary to mutual funds, performance analyzes of Hedge Funds are completely
diﬀerent. One reason for this, concerns the returns of Hedge Funds which usually have low
correlations with market indices, and therefore a traditional CAPM analysis using the Jensen
measure is typically nonappropriate. Another reason concerns the linear regression which
works well for mutual funds because its strategies tend to follow a buyandhold pattern.
Brown, Goetzmann and Ibbotson (1999) have shown that Hedge Funds pursue many diﬀerent
styles which are completely diﬀerent to the buy and hold strategy followed by Mutual Funds
and therefore can put the hypothesis of linearity to question. Liang (2001) also documents
that Hedge Fund investment strategies are dramatically diﬀerent from those of Mutual Funds.
To the best of our knowledge, the ﬁrst paper which suggested another approach in order to bet
ter analyze the returns of Hedge Funds was written by Fung and Hsieh (1997). They attempted
5
the readers can ﬁnd a detailed literature review into the book from Agarwal and Naik (2005).
4
to analyze the returns to Hedge Funds by applying the factor or style analysis conducted by
William Sharpe with respect to Mutual funds. As we will be using their methodology in this
paper we will introduce and develop it further later on in our work. In their paper, they found
that the amount of variation in Hedge Fund returns that is explained by asset class return is
low
6
. Schneeweis and Spurgin (1998) conﬁrm this result by conducting a regression analysis
of the returns of stocks, bonds, commodities, and currency returns. They found that Rsquare
measures that range from near zero for relative value Hedge Funds to 0.67 for Hedge Funds
that pursue primarily a long equity investment strategy, which concludes that Hedge Fund
return patterns do not map as well to the ﬁnancial asset as do mutual fund returns. In order
to ameliorate the explanation of Hedge Fund returns by asset class, Fung and Hsieh (1997)
applied a factor analysis based on a Hedge Fund’s trading style. They found that ﬁve dif
ferent trading style (system/opportunistic, global/macro, value, systems/trend following and
distressed) explain about 45 percent of the crosssectional variation in Hedge Fund returns.
Following Fung and Hsieh (1997), a lot of articles have been developed which seek to under
stand the trading strategies and characteristics of Hedge Funds by regressing their returns
on explanatory factors (Agarwal and Naik (2000), Brown and al. (2001), Mitchell and Pul
vino (2001)). Agarwal and Naik (2000) extended this analysis of Hedge Fund performance by
recognizing that funds may follow dynamic nonlinear trading strategies. They used stepwise
regression to identify the independent variables, and found that a put or a call option on
an underlying variable is the most signiﬁcant factor in the case of 54 percent of their funds.
Also, Fung and Hsieh (2002) incorporated option strategies into a Sharpe style model, but
they focused on exploring the risk of ﬁxed income Hedge Fund styles and did not consider
performance explicitly. In contrary to Agarwal and Naik’s (2000) discoverers, Fung and Hsieh
found that in most cases their option strategies only played a marginal role. One reason
given by the authors for the varied results mentioned above is due to the fact that Fung and
Hsieh (2002) used active and advanced straddle strategies. The paper written by Mitchell and
Pulvino (2001) on mergerarbitrage strategies are also begun to produce useful explicit links
between HedgeFund strategies and observable asset returns. Mitchell and Pulvino (2001) sim
ulated the returns of a merger arbitrage strategy applied to announced takeover transactions
from 1968 until 1998. In summary, Fung and Hsieh (1999,2000,2001), Mitchell and Pulvino
(2001) and Agarwal and Naik (2004) show that Hedge Fund returns relate to conventional
6
RSquare measures were less than twenty ﬁve percent for almost half of the Hedge Fund studied
5
asset class returns and optionbased strategy returns. They relate another strong diﬀerence
between Mutual Funds and Hedge Funds which concerns the problem of managers changing
their investment style over time. This problem is less acute for Mutual Funds than for Hedge
Funds. Brealey and Kaplanis (2001) presented evidence that within each category funds tend
to make similar changes to their factor exposures. Their interest is in the broader issue of how
well data on Fund returns can be used to measure “changes”in factor exposure. Indeed they
found that their CUSUMs’ tests of the stability of factor loadings rejected the null hypothesis
of stable coeﬃcients at the 5 percent level in the case of three quarters of the funds. In a
similar way, Fung et al.(2006) paper was concerned with estimating factor exposures at the
time of particular crises. They study vendorprovided fundoffund indices, and performed
a modiﬁedCUSUM test to ﬁnd structural break points in fund factor loadings. They found
that the break points coincide with extreme market events
7
. Recently, the paper from Bollen
and Whaley
8
studied 2 econometric techniques that accommodate changes in risk exposure.
The primer methodology was developed by Andrews, Lee, and Ploberger (1996). They stud
ied a class of optimal tests
9
for multiple changes. The latter, uses maximum likelihood and
a Kalman ﬁlter under an AR(1) model. They found signiﬁcant changes in the risk factor
parameters in about 40 percent of their sample of Hedge Funds.
Nevertheless the 2 precedent methodologies use a hypothesis of normality which is not adapted
to the world of Hedge Funds. With strong evidence of non normality (Agarwal and Naik, 2001;
Amin and Kat, 2003; Fung and Hsieh, 1999; Lo, 2001), the mean and standard deviations are
not suﬃcient to describe the return distribution, and higher moments need to be considered.
Kat and Lu (2002), Brooks and Kat (2002) show that although Hedge Funds oﬀer high mean
returns and low standard deviations, the returns also exhibit third and fourth moment at
tributes
10
as well as positive ﬁrstorder serial correlation
11
.
Furthermore these results about the distribution of Hedge Funds could be diﬀerent depending
on their strategies (Anson, 2006).
7
the collapse of LongTerm Capital Management in September 1998, and the peak of the technology bubble in
March 2000
8
At the moment where we write this article, we do not know when this paper will be publish in Journal of Finance.
9
In the sense that they maximize a weighted average power, namely the AvgW and ExpW.
10
Skewness and kurtosis.
11
They showed (as Lo & al. (2004)) that monthly Hedge Fund returns may exhibit high levels of autocorrelation.
6
3 Database
For this study, we use the Center for International Securities and Derivatives Markets (CISDM)
and HedgeFund.Net database. The primer covers the period from January 1994 through to
July 2007 with the advantage of the inclusion of dead funds. One of the main studies in this
article is to analyze whether or not Hedge Funds and CTA were able to generate alpha during
market events or on the contrary, whether they generated negative alpha. In order not to
create a bias we have included all funds in our analysis even though dead funds mean that
they did not necessarily generate alpha or worst, generate negative alpha. The sample with
all funds contains approximately 9800 funds (Hedge Funds, CTA and Fund of Funds), and
approximately 3000 live funds. The latter is the largest commercial database of active Hedge
Fund, Fund of Fund and CTA products with over 8500. It covers the period from May 1975
through to October 2008 but we are only concerned with the period from January 1991 until
October 2008. It has approximately 3000 Funds of Funds, 4900 Hedge Funds, and 600 CTAs.
For every fund, we have collected the returns, the strategy and fund type
12
. Returns are net
of management and performance based fees. We select funds for a minimum of 30 months
covering the period May 1998 until June 2000. We have created 2 groups in order to analyze
them separately; one for Hedge Funds and the other for CTAs. Nevertheless, it is more
pertinent to study Hedge Funds and CTAs depending on their strategies. There is a vast
selection of academic literature on how to classify Hedge Funds. Fung and Hsieh (1997)
and Brown and Goetzmann (2003) have identiﬁed between ﬁve and eight investment styles,
whereas, Bianchi Drew, Veeraraghavan, and Whelan (2005) have found the presence of only
three diﬀerent Hedge Fund styles. In contrast, Hedge Fund database providers classify Hedge
Funds into between 11 and 31 investment styles. Therefore choosing a number of strategies is
not an easy task, especially because we are not persuaded by the precedent results.
13
. Thus,
it seems better for this study to follow the 23 strategies deﬁned by the provider plus the CTA
as another strategy as well as Fund of Funds.
{please insert Table I here}
HedgeFund.net uses 31 strategies that we have grouped in order to obtain the same 23 strate
gies from the CISDM.
12
This database combines four main group, Hedge Funds, Funds of Funds, CTA, and CPO.
13
it will be the topic we are going to turn to next.
7
{please insert Table II here}
The main analyzes consider the merge database
14
due to the speciﬁcity of these products that
they tend to avoid direct regulation (by the SEC or any regulatory authorities).
4 Factors
Hedge Funds can be divided into four main groups, market directional, corporate restructuring
fund, convergence trading fund and opportunistic funds. We can add to each of them a major
risk factor. The exposure to the stock market is the major risk aﬀecting market directional
funds
15
. The major risk aﬀecting corporate restructuring funds
16
is exposure to the event
risk
17
which is the same for the convergence trading fund
18
. Into every group, each strategy
can also have a speciﬁc exposure to another risk factors and therefore risk exposure can change
drastically in regards to the strategy.
Consequently, deﬁning factors is not an easy task. Precedent articles have shown that the
relation between Hedge Fund returns and market returns are nonlinear. Moreover, Hedge
Funds have a systematic risk but it is not possible to capture this risk with standard asset
benchmarks. Therefore many researchers have suggested factors in order to explain Hedge
Fund returns (Fung and Hsieh (1999, 2000, 2001, 2004), Mitchell and Pulvino (2001) and
Agarwal and Naik (2004)). Nevertheless, it is not the only problem that we have with factors.
Indeed, knowing the right number of factors is major in order to capture risk correctly. If
some risk factors are missing, the model is misspeciﬁed and we can question whether alpha
corresponds to the manager skill. If there are too many factor, we can have a problem of
multicollinearity.
To the best of our knowledge, the most accomplished article about Hedge Fund factors was
written by Fung and Hsieh (2004). They showed that their seven factor model strongly explains
variation in Hedge Fund returns and at the same time avoid multicollinearity
19
. Moreover,
they managed to obtain similar results using the Agarwal and Naik (2004) optionbased factor
14
We merged the HedgeFund.Net database to the CISDM and found that they are 925 funds in common.
15
equity Long/Short, Short selling and equity market timing.
16
distressed securities, merger arbitrage and event driven.
17
failure of the proposed transaction.
18
ﬁxed income arbitrage, convertible bond arbitrage, equity market neutral, statistical arbitrage, and relative value
arbitrage.
19
We use the diagnostic technique presented in chap 3 of Regression Diagnostic by Belsley, Kuh, and Welsh (1980).
The diagnostic is capable of determining the number of near linear dependencies in a given data matrix X, and the
diagnostic identiﬁes which variable are involved in each linear dependency. We do not detect any multicollinearity
with these eight factors.
8
model.
Their paper included seven factors and they added an eighth factor on their website
20
that we
have also added.
Therefore we will follow the eight Hedge Fund risk factors deﬁned in Fung and Hsieh’s paper
(2004)
21
.
These factors are :
Three TrendFollowing Factors: Bond, Currency and Commodity
22
which capture a non linear
exposure.
•• 2 Equityoriented Risk Factors: S&P500 minus risk free rate
23
and Size Spread Factors deﬁned
by the Russell 2000 index monthly total return less S&P500 monthly total return.
• 2 Bondoriented Risk Factors: a Bond Market Factor represented by the monthly change in the
10year treasury constant maturity yield, and a Credit Spread Factor formed by the monthly
change in the Moody’s Baa yield less 10year treasury constant maturity yield.
• One Emerging Market Risk Factor, the MSCI Emerging market minus the risk free rate.
5 Structural Change
Few researchers have tested if there has been a structural change. The tests used considered
only one structural change and most of them needed to know when the break point should
happened. Indeed, the classical test for structural change is typically attributed to Chow
(1960) with the weakness that the breakdate must be known `a priori. Quandt (1960) treated
the breakdate as unknown but, in this case, the chisquare critical values are inappropriate.
The problem of critical values was solved simultaneously in the early 1990s by several sets of
authors. Nevertheless, the best known solution was provided by Andrews (1993), and Andrews
and Plobergers (1994). Andrews, Lee and Ploberger (1996) developed the precedent results
into a class of optimal tests for linear models with known variance. This class of optimal
tests allows an arbitrary number of changepoints. However, Bolley and Whaley implemented
this test using just one changepoint. The problem with these tests in the case of multiple
structural change is practical implementation. According to Perron, the AvgW and ExpW
20
http://faculty.fuqua.duke.edu/ dah7/.
21
For more details about the construction of these factors see Fung and Hsieh, 1997, 2001, 2004a.
22
These factors are downloadable on http://faculty.fuqua.duke.edu/ dah7/DataLibrary/TFFAC.xls .
23
3month USD LIBOR
9
tests require the computation of the Wtest over all permissible partitions of the sample. Bai
and Perron (1998, 2004) solved this problem with a very eﬃcient algorithm which is available
on their website
24
. Although Bai and Perron (1998, 2004) proposed three diﬀerent tests
25
, we
have applied just one of them which is, in our opinion, the most signiﬁcant for this analysis.
This is the second test provided by Bai and Perron, called Double Maximum Test. It is a
test of no structural breaks against an unknown number of breaks given some upper bound
(M = 5 in our application). In this category, there are two tests, the UDmax and the WDmax
which diﬀer by their weight methodology
26
.
The results are conclusive.
{please insert Table III here}
The majority of Hedge Funds present structural breaks. By strategy, the minimum percentage
of Hedge Funds with breaks is 31 percent (Other Relative Value) and the maximum is 70
percent for Event Driven Multi Strategy. If we consider the most representative strategy
(Equity Long/Short, 2142 Hedge Funds), 62 percent present some structural changes. All
these precedent results took into consideration track records with more than 30 months. Bai
and Perron’s tests also give the break’s date. Therefore another interesting view is to look at
the frequency of break’s dates in the Hedge Fund’s universe by considering the number of the
breaks at time t; relative to one fund. Indeed, the increase in the amount of Hedge Funds is
considerable and we must take this huge growth into consideration within our analysis. For
this reason, we have suggested to create a ratio called R
breaks
, which is deﬁned as the number
of breaks at time t divided by the number of funds at time t.
{please insert Graph I,II,III,IV here}
Once again, the results are conclusive. Whatever the strategy, we have noted an increase in
the ratio over the period 20022007 as well as an increase in the frequency of breaks.
There are two interesting things to say about the two crisis. In August 1998, the Russian gov
ernment defaulted on the payment of its outstanding bonds. This default caused a worldwide
liquidity crisis with credit spreads expanding rapidly all around the globe. The Russian debt
24
http://people.bu.edu/perron/. This code is a companion to the paper: Estimating and testing structural changes
in multivariate regressions (Econometrica, 2007) (developed by Zhongjun Qu).
25
The primer test considers no break versus ﬁxed number of breaks (up to 5 breaks)This test considers the sup F
type test of no structural break (m = 0) versus the alternative hypothesis that there are m = 1, ..., 5 breaks. It is a
generalization of the sup F test considered by Andrews (1993). The latter is a test of the null hypothesis of l breaks
against the alternative that an additional break exists. This sequential procedure estimates each break one at a time.
It stops when the sup F(l + 1l) test is not signiﬁcant.
26
See Bai and Perron (1998) for a full explanation on the diﬀerent weights.
10
crisis (LTCM) was a crisis that materially aﬀected Hedge Fund returns that was conﬁrmed by
the results from R
breaks
. Furthermore, the precedent results for a majority of strategies were
conﬁrmed.
Surprisingly, we noticed the opposite for the equity bubble crisis which corroborated the con
clusions of Brunnermeier and Nagel (2002).
The conditions for ﬁnancial markets in 1999 were very good, especially for riskier assets.
During this period a bubble developed. According to Brunnermeier and Nagel (2002) most
Hedge Funds, despite irrational levels of valuation, decided to ride the bubble rather than
clear their positions. They explained that Hedge Funds heavily tilted their portfolios towards
technology stocks without oﬀsetting this long exposure by short or derivatives. They concluded
that Hedge Funds deliberately held technology stocks and were able to exploit this opportunity.
These arguments are conﬁrmed by the fact that R
breaks
is very low, for the majority of the
strategies, showing fewer structural changes.
This section has shown that it is not enough to purely take into consideration diﬀerent dis
tributions but also that alpha and beta could be dynamic and consequently depend on time.
Furthermore, we showed that the risk in Hedge Funds increased over the last few years es
sentially due to the use of leverage in order to reach the investors desired return target. The
next section presents an answer to the problem of dynamic for alpha and beta. We suggest a
timevarying coeﬃcient model with the Fung and Hsieh factors which are, in this way, better
suited for Hedge Funds.
6 Our Model
Stone (1977) introduced local linear least squares kernel estimators as a regression estimator
which has been generalized by Cleveland (1979)
27
.
Stone (1980, 1982) used local linear least squares kernel and its generalization to higherorder
polynomials to show the achievement of his bounds on rates of convergence of estimators of a
function m and its derivatives. Fan (1992, 1993) showed in the univariate case that another
important advantage of local linear least squares kernel estimators is that the asymptotic bias
and variance expressions are particularly interesting and appear to be superior to those of
the NadarayaWatson or GasserM¨ uller kernel estimators. Furthermore, Kernel estimators
27
The robust local regression estimators.
11
have the advantage of being simple to understand and globally used by researchers. The
mathematical analyze and the implement into a computer is easy, and they are consistent for
any smooth m, provided the density f of X
i
s satisﬁes certain assumptions.
The varying coeﬃcient model assumes the following conditional linear structure:
Y
t
=
p
j=1
β
j
(t)X
jt
+ε
t
= α(t) +Xβ(t) +ε
t
for a given covariates (t, X
1
, ..., X
p
)
and variable Y . See appendix for more details on the
TimeVarying Coeﬃcient Model.
Thus β
i
(t) depend on t, this hypothesis can signiﬁcantly reduce the modeling bias and espe
cially avoid the “curse”of dimensionality
28
.
To practice statistical inferences such as the construction of conﬁdence interval for β
i
(t) dif
ferent methods have been suggested. Coling and Chiang (2000)
29
suggest a naive bootstrap
procedure that we have applied in this article.
The main advantage of this naive bootstrap procedure is that it does not rely on the asymp
totic distributions of
¯
β
i
(t). Coling and Chiang (2000) recommend another alternative boot
strap procedure suggested by Hoover et al. (1998), which relies on normal approximations
of the critical values
30
. According to the authors, both bootstrap procedures may lead to
good approximations of the actual (1 − α) conﬁdence intervals when the biases of
¯
β
i
(t) are
negligible
31
.
It is wellknown in kernel regression that the selection of bandwidths is more important than
the selection of kernel function. In practice, bandwidth may be selected by examining the plots
of the ﬁtted curves. Moreover, for this study we need an automatic bandwidth selection. Colin,
Wu and Chiang (2000) suggest that we apply the choice of bandwidth “leaveonesubject
out”crossvalidation
32
. We illustrated the performance of our estimator by a simulation study.
We used the bandwidth deﬁned above and a number of data equal to 50 in order to respect
28
Kernel depends, in this context, only on t.
29
Another paper of Galindo, Kauermann, and Carroll (2000) suggest another bootstrap method based on the wild
bootstrap of H¨ardle and Marron (1991)
30
Construct pointwise intervals of the form
β
i
(t) ±z
(1−α/2)
se
∗
B
(t),
where se
∗
B
(t) is the estimated standard error of
β
i
(t) from the B bootstrap estimators and z
(1−α/2)
is the (1 −α/2
th
)
percentile of the standard Gaussian distribution.
31
They point out that theoretical properties of these bootstrap procedures have not yet been developed.
32
Appendix I gives a summary to the algorithm.
12
the average size of Hedge Fund tracks (see appendix).
We regress the netoffee monthly excess return (in excess of the riskfree rate) of a Hedge
Fund on the excess returns earned by traditional buy and hold and primitive trend following
strategies deﬁned above
33
. We have grouped the estimates together into the 23 strategies
deﬁned by the CISDM. In each group obtained, we have formed 2 groups. The ﬁrst is composed
by all tracks record which is superior to 30 months. The second has the ﬁrst requirement (more
than 30 months) but only has funds which were available during the 2 crisis.
{please insert Table I here}
Indeed, we are concerned with the 2 diﬀerent aspects. The ﬁrst focusses on the security
selection ability (´ α(t)). The second aspect treats the ability to anticipate market events or
to cope with them (i.e. robust ability). In order to capture anticipation in alpha, we are
going to look at 2 strong market events: the LTCM crisis and the Equity Bubble crisis. To
properly cover these periods, we have selected 3 consecutive months. For the LTCM period we
will focus on July, August, and September 1998, and for the Equity Bubble period, February,
March and April 2000
35
. Using these 3 months’ data, we have built 2 indicators for α
36
and
2 indicators for β
i
, (i = 1, ..., 8). The 2 primers are the diﬀerence between the second month
and the ﬁrst month and the diﬀerence between the last month and the second month.
37
The 2
indicators for beta
38
have the same diﬀerences but are augmented by 10 percent
39
. By doing
this, we were able to determine whether or not managers were able to react quickly and also
if we can capture change exposure superior to 10%. In each case, we have used the tstatistic,
´
t
i,α
= ´ α
i
/´ σ
α
i
and
´
t
i,β
=
´
β
i
/´ σ
β
i
for every ´ α
j
(t) and
´
β
ij
(t), i = 1, ..., 8. For every indicator
33
Kat and Lu (2002), Brooks and Kat (2002) show that the netoffees monthly returns of the average individual
Hedge Funds exhibit positive ﬁrstorder serial correlation which is due, according to the authors, to markingtomarket
problems. We have removed serial correlation by applying the same methodology as used in Brooks and Kat’s paper
(2001), called the simple Blundellward ﬁlter
34
.
The observed (or smoothed) value V
∗
t
of a Hedge Fund at time t could be expressed as a weighted average of the true
value at time t, V
t
and the smoothed value at time t −1, V
∗
t−1
:
V
∗
t
= αV
t
+ (1 −α)V
∗
t−1
,
r
t
=
r
∗
t
−αr
∗
t−1
1 −α
.
35
End of month of July, Auguste, and September and end of month of February, March, and April.
36
ˆ α
t
− ˆ α
t−1
37
These two indicators were created in order to show a new methodology that this model bring to the analysis of
Hedge Funds. Nevertheless, some others indicators could be more appropriate for a speciﬁc analysis of these market
events.
38
ˆ
β
t
−(1.1) ×
ˆ
β
t−1
39
This methodology is simply a linear relation between 2 independent variables which, under the condition of
normality for
ˆ
β
jt
j = 1, ...N
j
; N
j
being the number of funds, assure that the linear relation follows also a normal
distribution.
13
covering the 2 periods i.e. 4 indicators by fund for
α
j
(t) and 32 indicators by fund for
β
ij
(t).
Once it was done, we calculated the pvalue for every indicator and the precedent means.
40
If we examine just 1 fund, we only have to test a null hypothesis H
0
versus an alternative H
1
based on a statistic (let’s say X). We have 2 possibilities for a given rejection region Γ. Either
we reject H
0
when X ∈ Γ or we accept H
0
when X ∈ Γ. Nevertheless, there is the possibility
that there is an error in the test. On the one hand, the test can reject H
0
whereas H
0
is true,
called type I error. On the other hand, the test can accept H
0
whereas it is H
1
which is true.
This error is called a type II error. Now, if we want to test for numerous Hedge Funds this
problem becomes much more complicated. Testing numerous Hedge Funds is to do, in fact, a
multiplehypothesis testing.
A ﬁrst approach was suggested by Kosowski, Naik and Teo (2005), using a bootstrap proce
dure; They analyzed whether or not Hedge Fund performance can be explained by luck and if
Hedge Fund performance persists at annual horizons. Their methodology tested the skills of
a single fund that was chosen from the universe of alpharanked funds. Barras, Scaillet, and
Wermers (2008) suggested another approach which is notably informative with regards to the
prevalence of outstanding managers in the whole fund population. Their approach simultane
ously estimated the prevalence and location of multiple outperforming fund within a group;
examining fund performance, in this manner, from a more general perspective. Therefore we
have used their methodology called False Discovery Rate (FDR hereafter) in our multiple
hypothesis test.
Step 1, for each strategy, we have applied the FDR on ´ α (security selection ability) in order to
determine the proportion of zeroalpha funds and skilled and unskilled funds; but also for the
indicator deﬁned above ˆ α
t
− ˆ α
t−1
(timing ability) for the 2 crisis. In this way, we were able
to determine, for each strategy, the proportion of the security selection ability as well as the
proportion of the timing ability.
Step 2, we applied the FDR on the “beta” indicators deﬁned above during the 2 crisis in order
to determine if we could observe a common increasing trend for change in market exposure i.e.
a common increase in beta value of more than 10%. Moreover, we calculated the median of the
percentage change for each beta of each strategy in order to give an “I.D.” of the possibility
change in market exposure.
40
The tstatistic distributions for individual Hedge Funds are generally nonnormal. In order to overcome the non
normality, we use the same approach as Barras, Scaillet, and Wermers (2008), consisting of the use of a bootstrap to
more accurately estimate the distribution of tstatistics for each Hedge Funds (and their associated pvalues).
14
With these results, we brought 2 diﬀerent views about exposure change. Firstly, a view about
common trend change and secondly about what are the highest impacted beta.
7 Results
Generally speaking, Mutual Fund managers use a buy and hold strategy consisting of buying
a range of ﬁnancial products according to their investment strategy and then they hold them
according to the time horizon (or investment horizon)
41
. Therefore Mutual Funds are often
associated with Funds that have a relative performance. Unsurprisingly, Barras, Scaillet and
Wermers (2008) have shown that only 0.6 percent of Mutual Funds generate positive alpha
and a majority of them can also be considered as zeroalpha funds. So the legitimate question
is: Are the results relatively the same for Hedge Funds or are they completely diﬀerent?
On the one hand, if we apply a static linear factor model such as the NeweyWest (1987)
heteroscedasticity and autocorrelation consistent estimator, we determine a “static”alpha that
does not capture the particularities of Hedge Fund Strategies. And therefore we have showed
that, whatever the strategies, the majority of Hedge Funds are zeroalpha funds.
On the other hand, when we applied our timevarying coeﬃcient model, we were able to
capture other skills from Hedge Fund managers, resulting in obtaining us a non negligible
increase of positive alpha funds.
{please insert Table V here}
Concerning the risk behavior, we showed that the majority of Hedge Funds have a tendency
to get an increase in credit spread as well as bond market risk factors. We are going to look
into the results for seven out of the twenty four strategies in the following part of this paper
42
.
Results for the CTA
We observed a strong diﬀerence for estimated alpha between the static factors model (SFM
hereafter) and our timevarying coeﬃcient model (TVCM hereafter). The SFM gave a very
small percentage of positive and negative alpha funds with two percent and one percent respec
tively. Whereas the TVCM roughly gave nineteen and twenty eight percent. The estimated
alpha from the two diﬀerent crisis are interesting to point out. Although The CTA coped
41
refer to the time between making an investment and needing the funds.
42
The results and the graphs for the seventeen remaining strategies are available upon request.
15
during the two crisis it obtained a strong percentage of positive alpha funds during the Equity
Bubble Crisis where approximately twenty seven percent were positive alpha funds.
During the Summer of 1998, CTAs had one of its best performances
43
while all other Hedge
Fund strategies were struggling.
Unsurprisingly, during the two events, the CTA had a tendency to show a slight increase in
the credit spread risk factor and the emerging market risk factor. Nevertheless this sensitivity
concerned only a small percentage of our population. The majority of CTAs had a relatively
stable exposure.
{please insert Graph XI here}
Emerging Markets
The emerging market strategy showed a good proportion of stockpicker skilled funds where
approximately twelve percent generated positivealpha which pointed out that the majority of
managers were fundamental bottomup stockpickers. We obtained the same results using the
SFM. It indicated that the estimated alpha was more “static”than with other strategies. The
proportion of dynamic skilled funds was very good during the two crisis with a huge number of
roughly forty percent, of positive alpha funds. These results conﬁrmed that emerging market
equity hedge fund managers perceived the high volatility of emerging markets as an asset.
During the Equity Bubble and LTCM we noticed a greater dynamic strategy than previously
seen where approximately twentyﬁve and ﬁfty percent of our population showed an increase
in two risk factors: the credit spread risk factor and the bond market risk factor. The credit
spread risk factor was the most sensitive factor during LTCM, whereas, four out of the eight
factors showed a sensitivity during the Equity Bubble crisis.
{please insert Graph V here}
Equity Long/Short
Equity Long/Short obtained approximately the same results as the CTA apart from the Equity
Bubble crisis. It obtained a better percentage with twenty four percent using the TVCM. The
SFM gave a small four percent and three percent of negative alpha funds. Certain Equity
Long/Short specialize in a speciﬁc sector like technology, and, unsurprisingly, the timer ability
had been more impact during the Equity Bubble than during LTCM. Generally speaking, the
43
Approximately 10 percent in August and 7.5 percent in September according to CSFB/Tremont Managed Futures
16
proportion stayed relatively consistent pointingout their ability to switch from the short to
the long position and visa versa.
A small percentage of the population showed an increase or a decrease in exposure during the
two crisis. We still noticed a sensitivity to the credit spread risk factor and to the commodities
factors during LTCM. Whereas emerging risk factor was the most sensitive during the Equity
Bubble crisis.
{please insert Graph VI here}
Equity Market Neutral
This strategy produced a very interesting result where the proportion of stockpicker skilled
funds was three percent higher than the estimated proportion using the SFM. Therefore the
two results were relatively closer than with the other strategies. Furthermore this result was
conﬁrmed by the obtained percentage of estimated alpha during LTCM with zero percent. It
didn’t cope as well during the equity bubble. We noticed a tiny four percent during the ﬁrst
period which reached a strong sixteen percent in the second period.
{please insert Graph VII here}
It was the most robust strategy in terms of change in exposure. Nevertheless, for a minority
of the population we noticed that the credit spread and the commodity showed the biggest
sensitivity during LTCM and the emerging market factor during the Equity Bubble.
Event Driven Multi Strategy
Event Driven Multi Strategy obtained the worst percentage of skilledfunds with zero percent
for the SFM and a small two point ﬁve percent for the TVCM. The result was conﬁrmed
during the two crisis with zero percent during LTCM. This was surprising because the two
crisis created several opportunities
44
, and only the second crisis was more proﬁtable. On the
other hand it was the strategy which obtained the smallest proportion of unskilled funds.
Therefore we can deﬁne it as a zeroalpha fund strategy.
{please insert Graph VIII here}
Another strength, for this strategy, concerned the percentage of the population to show a
variation in factor exposure. A negligible part showed an instability during the crisis. The
44
Invests in mergers, spinoﬀs, reorganizations, and other announced events.
17
factors concerned were the credit spread the emerging market factor and the commodity factor
for LTCM and the emerging risk factor for the equity bubble.
Global Macro
Global Macro showed a proportion of stockpicker skilled funds equal to ﬁve percent using
TVCM and one percent using the SFM. It obtained one of the bigger groups of unskilled
funds with roughly twenty eight percent. Unsurprisingly, the percentage of positive alpha
during LTCM increased up to twenty six percent. These results conﬁrmed that global macro
managers have the most extensive investment universe and that they were able to ﬁnd some
arbitrages. The Equity Bubble crisis also gave a good percentage of positive alpha funds with
eighteen percent.
Less than ten percent of our population showed an increase or decrease in our general expo
sure. Credit Spread risk factor stayed the most sensitive during LTCM. Whereas, the size
spread factor, the emerging market risk factor and the credit spread risk factor (slightly) were
concerned about the change in exposure.
{please insert Graph IX here}
Short Bias
Step one, it is important to say that the number of Hedge Funds following this strategy was
relatively small, so the result, in our opinion, could be debatable.
The SFM gave ﬁve percent of skilled funds whereas the TVCM gave thirty ﬁve percent.
Furthermore, this strong percentage was less than the seventy one percent of positive alpha
funds found during LTCM. The Equity Bubble obtained a tiny twenty seven percent: This
impressive result conﬁrmed the strong dynamic within the strategy.
Moreover, the short bias produced the best percentage of variation exposure in. More than
sixty percent of our population showed an increase in diﬀerent market risk factors. Credit
spread was still present during LTCM while Sizespread risk factor had an non negligible
sensitivity to bond, commodity and emerging market risk factors during the Equity Bubble.
{please insert Graph X here}
18
Robustness
The goal of this section is to demonstrate that our model is robust to structural change and
that we succeeded to capture a jump in estimated alpha or beta. For this, we created a track
of return where the two betas had a structural change. We allowed for three diﬀerent sizes:
ﬁfty, one hundred and one hundred and ﬁfty months which akin the size of Hedge Funds that
we can ﬁnd in diﬀerent databases.
The main problem with structural change concerns some technical conditions. Mainly that the
second derivative is required to be continuous in a neighborhood of t. As showed in appendix
I, we managed to obtain some very good results; even in the short sample.
Furthermore, in order to show that our results were independent to our model we created
another timevarying coeﬃcient model which was also based on the work of Fan and Zhang
(1999). Nevertheless in second step to our estimator we used a Bspline smoothing instead of
a local regression. We retested the ability to capture a structural break (see appendix I). The
results were also very good.
We used one of the same example as in Zhang, Lee and Song (2002). Our result are less
accurate than Zhang, Lee and Song due to three reasons.
Firstly, our sample contains between 50 and 150 values in opposite to the range in 2501000.
Secondly, we did not apply a monte carlo study in order to ﬁnd the optimal bandwidth (or
knot for the second model). Because of the size of our database, the use of a second bootstrap
algorithm for each fund (the ﬁrst algorithm calculates the pvalue) would have made the
estimation process too costly.
Thirdly, we used the S&P500 and the monthly change in the 10year treasury constant maturity
yield as factors which are not a perfect factors set. And therefore these diﬀerences slightly
alter the performance of our estimator. For these reasons the diﬀerent graphics truly represent
the obtained result for each fund.
In a second step, we applied the false discovery rate to the estimated alpha and betas obtained
by Bspline. The results gave approximately the same percentage of unskilled, zero and skilled
funds for our population.
19
8 Conclusion
Hedge Funds cover many diﬀerent strategies which radically vary in terms of market exposure
and risk. However there are some common characteristics. Hedge Fund managers try to focus
on positive return (whatever the market conditions), the use of leverage and their structural
fees. If we want to model Hedge Funds, these characteristics imply a model which takes
some new assumptions into consideration. No assumption about statistical distribution, the
dynamic in beta as well as non linearity exposure to the market.
In order to overcome these drawbacks, we used a timevarying coeﬃcient model as well as
the whole factors; deﬁned in Fung and Hsieh’s paper (1997, 2001, 2004). The model allowed
us to postulate that alpha and beta are a function which depended on time and avoided
parametric distribution. The use of the factors from Fung and Hsieh allowed us to capture
the nonlinearity in beta and therefore gave the best overall risk factors.
This model allowed us to separate manager skill into two components pointedout by the (stock
or bond or funds)picking and the ability to anticipate market events. It also allowed us to see
what were the changes in beta exposure or the managers’ reactions according to the market’s
conditions.
Whereas Barras, Scaillet and Wermers (2008) showed us that only 0.6 percent of Mutual
Funds generated positive alpha and therefore a majority of them can be considered as zero
alpha funds. We found a diﬀerent result for Hedge Funds. Hedge Fund managers seek absolute
returns and try to outperform the market whatever the market conditions are. We showed
that this dynamic was not wellcaptured by a static factor model where the results considered
Hedge Funds as zeroalpha funds. Whereas our model found a better proportion of positive
alpha funds but also, unfortunately, a proportion of negative alpha funds.
Another advantage of this model was its capacity to analyze change in factorial exposure. We
saw three main results. Firstly, that a majority of Hedge Funds with a track record superior
to 30 months had a minimum of one structural break. Furthermore we noticed that the
frequency of breaks has increased over the last few years. Secondly, by strategy, we examined
the percentage change between our eight factors and saw if there was a persistent for beta
parameters. Our results suggest that the common increasing percentage change of diﬀerent
Hedge Funds in the downstate of the market is strongly represented by one exposure. it
was the credit spread risk factor. Last but not least, we applied the FDR to determine the
20
proportion of the fund’s population which had an increase, a decrease, or stayed constant
in the market’s exposure during the two crisis (LTCM and the Equity Bubble). We show
that within a strategy, Hedge funds have a tendency to have an increase in market exposure.
The changes in factorial exposure and the proportion of funds give us a strong tool for risk
managers and speciﬁcally for stresstesting.
21
References
[1] Ackermann C., McEnally R., and Ravenscraft D. The Performance of Hedge Funds:
Risk, Return, and Incentives. The journal of ﬁnance, 54:833–874, 1999.
[2] Agarwal V., Bakshi G., and Huij J. Dynamic Investment Opportunities and the Cross
Section of HedgeFund Returns: Implications of HigherMoment Risks for Performance. Work
ing Paper, May 2008.
[3] Agarwal V., Boyson N. M., and Naik N. Y. Poor Man’s Hedge Funds?, Performance and
Risktaking of Hedged Mutual Funds. Working paper, 2006.
[4] Agarwal V., Daniel N., and Naik N. Y. Do Hedge Funds Manage their Reported Returns.
Working Paper, 2007.
[5] Agarwal V. and Daniel N. D.and Naik N. Y. Flows, Performance, and Managerial Incen
tives in Hedge Funds. Working paper, 2004.
[6] Agarwal V. and Fung W. H.and Loon Y. C. Risk and Return in Convertible Arbitrage:
Evidence from the Convertible Bond Market. Working paper, 2006.
[7] Agarwal V. and Naik N. Y. MultiPeriod Performance Persistence Analysis of Hedge Funds.
The Journal of Financial and Quantitative Analysis, 35:327–342, 2000.
[8] Agarwal V. and Naik N. Y. Performance Evaluation of Hedge Funds with Optionbased
and BuyandHold Strategies. Working Paper, 2000.
[9] Agarwal V. and Naik N. Y. Risks and Portfolio Decisions Involving Hedge Funds. Working
Parer, 2002.
[10] Agarwal V., Naik N. Y., and Daniel N. D. Role of Managerial Incentives and Discretion
in Hedge Fund Performance. Working Paper, 2003.
[11] Ahmad I., Leelahanon S., and li Q. Eﬃcient Estimation of a Semiparametric Partially
Linear Varying Coeﬃcient Model. The annals of Statistics, 33(1):258–283, 2005.
[12] Amenc N., El Bied S., and Martellini L. Evidence of Predictability in Hedge Fund Returns
and MultiStyle MultiClass Tactical Style Allocation Decisions. Working paper, 2002.
[13] Amenc N. and Martellini L. Portfolio Optimization and Hedge Fund Style Allocation
Decisions. Working paper, March 2002.
[14] Amin G. S. and Kat H. M. Portfolio of Hedge Funds. Working Paper, 2002.
[15] Amin G. S. and Kat H. M. Hedge Fund Performance 19902000. Working Paper, January 4,
2002.
[16] Ang A. and Bollen N. P.B. Locked Up by a Lockup: Valuing Liquidity as a Real Option.
Working Paper, November 2008.
[17] Avramov D., Kosowski R., Naik N. Y., and Teo M. investing in Hedge Funds when
Returns are Predictable. Working Paper, 2008.
[18] Bacmann J.F. and Jeanneret A. Funds of Funds Are Still Providing Alpha. Hedgequest,
1(1):22–26, 2005.
[19] Bailey W., Li H., and Zhang X. Hedge Fund Performance Evaluation: A Stochastic Dis
count Approach. Working Paper, December 2004.
22
[20] Bansal R. and Viswanathan S. No arbitrage and arbitrage pricing: A new approach. The
Journal of Finance, 48:1231–1262, 1993.
[21] Bartholdy J. and Peare P. Estimation of Expected Return: Capm vs Fama and French.
Working Paper, 2004.
[22] Beran R. and Hall P. estimating Coeﬃcient Distributions in Random Coeﬃcient Regres
sions. The annals of Statistics, 20(4):1970–1984, Dec., 1992.
[23] Billio M., Getmansky M., and Pelizzon L. NonParametric Analysis of Hedge Fund
Returns: New Insights from High Frequency Data. Working Paper, 2008.
[24] Bollen N. P.B. and Whaley R. E. Hedge Fund Risk Dynamics: Implications for Perfor
mance Appraisal. Working Paper, 2008.
[25] Brealey R.A. and Kaplanis E. Hedge Funds and Financial Stability: an Analysis of Factor
Exposures. International Finance, 4(2):161–187, 2001.
[26] Brooks C. and Kat H. M. The Statistical Properties of Hedge Fund Index Returns and their
Implications for Investors. Journal of Alternative investments, 2002.
[27] brown S., Goetzmann W.and Liang B., and Schwarz C. estimating Operational Risk for
Hedge Funds the ωScore. Working Paper, May 2008.
[28] Brown S. J. and Goetzmann W. N. Hegde Funds and Style. Journal of Portfolio Manage
ment, 29(2):101–112, 2003.
[29] Brown S. J., Goetzmann W. N., and Ibbotson R. G. Oﬀshore Hedge Funds: Survival and
Performance, 198995. The Journal of Business, 72:91–117, 1999.
[30] Brown S. J., Goetzmann W. N., and Park J. Careers and Survival: Competition and Risk
in the Hedge Fund and CTA Industry. The journal of Finance, 56:1869–1886, 2001.
[31] Brown S. J.and Goetzmann W. N.and Liang B. Fees on Fees in Funds of Funds. Working
paper, 2004.
[32] Brunnermeier M. K.and Nagel S. Arbitrage at its Limits: Hedge Funds and the Technology
Bubble. Working Paper, 2002.
[33] Carol A. and Dimitriu A. A Comparison of Cointegration and Tracking Error Models for
Mutual Funds and Hedge Funds. Working paper, 2004.
[34] Chambers J. M. and Hastie T. J. Statistical Models in S. Wadsworth, Paciﬁc Grove, 1992.
[35] Chan K. S. Consistency and Limiting Distribution of the Least Squares Estimator of a
Treshold Autoregressive Model. The annals of Statistics, 21(1):520–533, 1993.
[36] Chan N., Getmansky M., Haas S. M., and Lo A. W. Systemic Risk and Hedge Funds.
Working paper, 2005.
[37] Christiansen C. B., Madsen P. B., and Christensen M. A Quantitative Analysis of Hedge
Fund Style and Performance. Working paper, 2004.
[38] Colin O. and ChinTsang C. Kernel Smoothing on Varying Coeﬃcient Models with Longi
tudinal Dependent Variable. Statistica Sinica, 10:433–456, 2000.
[39] Darolles S. and Florens J. P.and Simon G. Hedge Fund Duration: Endogeneity of Per
formance and Assets Under Management. Working Paper, 2009.
23
[40] Darolles S., Gourieroux C., and Jasiak J. LPerformance with an Application to Hedge
Funds. Working Paper, July 2008.
[41] Davies R. J., Kat H. M., and Lu S. Fund of Hedge Funds Portfolio Selection: A Multiple
Objective Approach. Working Paper, August,2005.
[42] Detemple J.B., Garcia R., and Rindisbacher M. A Monte Carlo Method for Optimal
Portfolios. The journal of Finance, 58:401–446, 2003.
[43] Eberlein E. and Madan D. B. Hedge Fund Performance: Sources and Measures. Working
Paper, 2008.
[44] Fan J. DesignAdaptive Nonparametric Regression. J. Amer. Statist. Assoc., 87:998–1004,
1992.
[45] Fan J. Local Linear Regression Smoothers and their Minimax Eﬃciencies. Ann. Statist,
21:196–216, 1993.
[46] Fan J. and Gijbels I. Variable Bandwidth and Local Linear Regression Smoothers. Ann.
Statist, 20:2008–2036, 1992.
[47] Fan J. and Huang T. Proﬁle Likelihood Inferences on Semiparametric VaryingCoeﬃcient
Partially Linear Models. Bernoulli, 11(6):1031–1057, 2005.
[48] Fan J. and Zhang W. Statistical Estimation in Varying Coeﬃcient Models. The annals of
Statistics, 27(5):1491, Oct., 1999.
[49] Favre L. and Galeano J.A. MeanModiﬁed ValueatRisk Optimization with Hedge Funds.
The Journal of Alternative Investments, 5:21–25, 2002.
[50] Favre L. and Signer A. The Diﬃculties of Measuring the beneﬁts of Hedge Funds. Journal
of alternative investments, 2002.
[51] Fung W. and Hsieh D. A. Empirical Characteristics of Dynamic Trading Strategies: The
case of Hedge Funds. The Review of Financial Studies, 10:275–302, 1997.
[52] Fung W. and Hsieh D. A. Is MeanVariance Analysis Applicable to Hedge Funds. Economics
Letters, 62:53–58, 1999.
[53] Fung W. and Hsieh D. A. A Primer of Hedge Funds. Working paper, 1999.
[54] Fung W. and Hsieh D. A. Performance Characteristics of Hedge Funds and Commodity
Funds: Natural vs. Spurious Biases. The Journal of Financial and Quantitative Analysis,
35:291–307, 2000.
[55] Fung W. and Hsieh D. A. The Risk in Hedge Fund Strategies: Theory and Evidence from
Trend Followers. The Review of Financial Studies, 14:313–341, 2001.
[56] Fung W. and Hsieh D. A. AssetBased Style Factors for Hedge Funds. Analyst Journal,
58(5):16–27, 2002.
[57] Fung W. and Hsieh D. A. The Risk in Fixed Income Hedge Fund Styles. Journal of Fixed
Income, 2002.
[58] Fung W. and Hsieh D. A. Extracting Portable Alphas from Equity Long/Short Hedge Funds.
Journal of Investment Management, 2:1–19, 2004.
[59] Fung W. and Hsieh D. A. Hedge Fund Benchmarks: A Risk Approach. Financial Analyst
Journal, pages 65–80, 2004.
24
[60] Fung W. and Hsieh D. A. Hedge Funds: An Industry in its Adolescence. Economic Review,
Fourth Quarter 2006.
[61] Fung W., Hsieh D. A., Naik N. Y., and Ramadorai T. Hedge Funds: Performance, Risk
and Capital Formation. Working paper, 2006.
[62] Galindo C. D., liang H. andKauermann G., and Carroll R. J. Bootstrap Conﬁdence
Intervals for Local Likelihood, Local Estimating Equations and Varying Coeﬃcient Models.
working paper, sept, 2000.
[63] Gayer G., Gilboa I., and Lieberman O. RuleBased and CaseBased Reasoning in Housing
Prices. Working Paper, 2004.
[64] Geltner D. Smoothing in AppraisalBased Returns. journal of Real Estate Finance and
Economics, 4:327–345, 1991.
[65] Geltner D. Estimating Market Values from Appraisal Values without Assuming an Eﬃcient
Market. Journal of Real Estate Research, 8:325–345, 1993.
[66] Getmansky M., Lo A. W., and Makarov I. An Econometric Model of Serial Correlation
and Illiquidity in Hedge Fund Returns. Working paper, 2003.
[67] Glosten L.R.and Jagannathan R. A Contingent Claim Approach to Performance Evalua
tion. Journal of Empirical Finance, 1:133–160, 1994.
[68] Goodworth T. R. and Jones C. M. FactorBased, Nonparametric Risk Measurement
Framework for Hedge Funds and FundsofFunds. The European Journal of Finance, 13:7:645–
655, October 2007.
[69] H¨ ardle W. and Marron J. S. Bootstrap Simultaneous Error Bars for Nonparametric Re
gression. The annals of Statistics, 19:778–796, 1991.
[70] Hastie T. and Tibshirani R. VaryingCoeﬃcient Models. Journal of the Royal Statistical
Society. Series B(methodological), 55(4):757–796, 1993.
[71] Hildreth C. and houck J. P. Some Estimators for a Linear Model with Random Coeﬃcients.
Journal of the american Statistical Association, 63(322):584–595, Jun., 1968.
[72] Hodder J. E. and Jackwerth J. C. Incentive Contracts and Hedge Fund Management.
Working paper, 2005.
[73] Hsiao C. Some Estimation Methods for a Random Coeﬃcient Model. Econometrica,
43(2):305–325, Mar., 1975.
[74] Jaeger L. and Wagner C. Factor Modelling and Benchmarking of Hedge Funds: Can
Passive Investments in Hedge Fund Strategies Deliver. Working Paper, 2005.
[75] Kat H. Integrating Hedge Funds into the Traditional Portfolio. Working paper 0022, 2005
january.
[76] Kat H. and Palaro H. Replication Evaluation of Fund of Hedge Funds Returns. Working
Paper n28, January 2006.
[77] Kat H. M. Managed Futures and Hedge Funds: A Match Made in Heaven. Working Paper,
November,2002.
[78] Kat H. M. The Dangers of Mechanical Investment DecisionMaking: The Case of Hedge
Funds. Working Paper, November,2003.
25
[79] Kat H. M. In Search of the Optimal Fund of Hedge Funds. Working Paper, October,2002.
[80] Kat H. M. Taking the Sting Out of Hedge Funds. Working Paper, October,2002.
[81] Kat H. M. and Lu S. An Excursion into the Statistical Properties of Hedge Funds. Working
paper, 2002.
[82] Kat H. M. and Miffre J. Performance Evaluation and Conditioning Information: The Case
of Hedge Funds. Working paper, 2002.
[83] Kat H. M. and Miffre J. The Impact of NonNormality Risks and Tactical Trading on
Hedge Fund Alphas. Working paper, May 2006.
[84] Kat H. M. and Palaro H. P. Who Needs Hedge Funds ? Working Paper 0027, 2005.
[85] Kosowski R., Naik N. Y., and Teo M. Do Hedge Fund Deliver Alpha ? A Bayesian and
Bootstrap Analysis. Journal of Financial Economics, 2005.
[86] Li S. and Patton A. J. TimeVarying liquidity in Hedge Fund Returns. Working Paper,
London school of Economics, 2006.
[87] Li S. and Patton A. J. TimeVarying Liquidity in Hedge Fund Returns. Working Paper,
May 2007.
[88] Liang B. Hedge Funds: The Living and the Dead. Journal of Financial and Quantitative
Analysis, 35:309–336, 2000.
[89] Linton O. and Li S. Evaluating Hedge Fund Performance: a Stochastic Dominance Approach.
Working paper, 2007.
[90] Malkiel B. G. and Saha A. Hedge Funds: Risk and Return. Working Paper, 2004.
[91] McFall Lamm R. Asymmetric Returns and Optimal Hedge Fund Portfolios. The journal of
Alternative Investments, 2003.
[92] Mitchell M. and Pulvino T. characteristics of Risk and Return in Risk Arbitrage. The
journal of Finance, 56:2135–2175, 2001.
[93] Pezier J. and White A. The Relative Merits of Investable Hedge Fund Indices and of Funds
of Hedge Funds in Optimal Passsive Portfolios. Working Paper, 2006.
[94] Popova I., Morton D., Popova E., and Yau J. Optimal Hedge Fund Allocation with
Asymmetric Preferences and Distributions. Working paper.
[95] Ruppert D. and Wand M. P. Multivariate Locally Weighted Least Squares Regression. The
annals of Statistics, 22(3):1346–1370, 1994.
[96] Schneeweis T. and Spurgin R. Multifactor Analysis of Hedge Fund and Managed Futures,
Return and Risk Characteristics. Journal of Alternative Investments, 1:1–24, 1998.
[97] Severini Thomas A. and H. W. Wong. Proﬁle Likelihood and Conditionally Parametric
Models. The Annals of Statistics, 20(4):1768–1802, Dec., 1992.
[98] Stone C. J. optimal Global Rates of Convergence for Nonparametric Regression. Ann. Statist,
10:1040–1053.
[99] Stone C. J. Consistent Nonparametric Regression. Ann. Statist, 5:595–620, 1977.
[100] Stone C. J. Optimal Rates of Convergence for Nonparametric Estimator. Ann. Statist,
8:1348–1360, 1980.
26
[101] Swamy P. A. V. and Mehta J. S. Estimation of Linear Models with Time and Cross
Sectionally Varying Coeﬃcients. Journal of the American Statistical Association, 72(360):890–
898, Dec., 1977.
[102] zhang W., Lee S.Y., and Song X. Local Polynomial Fitting in Semivarying Coeﬃcent
Model. journal of Multivariate Analysis, 82:166–188, jan. 2002.
27
Table I
See Appendix I for deﬁnitions of fund types. The funds used at minimum cover the LTCM and bubble period which
represent a track record of a minimum of 30 months. Certain of the funds used are considered as dead funds i.e. they
stopped their activities. The database marked by a asterisk give the number of funds covering the 2 speciﬁc periods
i.e, LTCM period (July, Auguste, and September 1998) and the Equity Bubble years (February, March, and April 2000)
Strategies Number of Hedge Funds
CISDM merge CISDM* merge*
Equity Long/Short 2141 3519 1561 2525
Multi Strategy 251 734 142 462
Emerging Markets 445 635 331 461
Sector 387 630 296 468
Equity Market Neutral 322 698 227 519
Event Driven Multi Strategy 224 384 168 293
Global Macro 303 534 205 377
Equity Long Only 148 276 99 174
Single Strategy 114 112 50 50
Fixed Income 153 326 109 260
Distressed Securities 162 268 131 232
Fixed Income Arbitrage 190 314 143 237
Convertible Arbitrage 207 271 181 227
Relative Value Multi Strategy 89 179 74 137
Fixed Income  MBS 74 98 59 69
Option Arbitrage 29 126 17 76
Merger Arbitrage 126 139 111 122
Other relative Value 16 32 7 17
Short bias 51 74 32 56
Regulation D 16 56 13 44
Capital Structure Arbitrage 21 30 13 21
Market Timing 2 3 1 1
Unclassiﬁed 58 521 37 369
CTAs (systematic) 1003 759
CTAs (discretionary) 283 1915 202 1394
FoHFs (multi strategy) 1837 1390
28
Table II
See Appendix I for deﬁnitions of fund types. This table shows how we have grouped the 31 strategies from Hedge
Fund.Net and the 23 strategies from CISDM .
CISDM database HedgeFund.Net database
23 Strategies 31 Strategies
Multi Strategy
Multi Strategy
Statistical Arbitrage
Equity Long/Short Equity Long/Short
Short bias Short bias
Event Driven Multi Strategy Event Driven
Emerging Markets Emerging Markets
Merger Arbitrage Merger (risk) Arbitrage
Fixed Income Fixed Income (non arbitrage)
Equity Market Neutral Market Neutral Equity
Global Macro Macro
Relative Value Multi Strategy Value
Sector
Small/Micro Cap
Finance Sector
Technology Sector
Energy Sector
Healthare Sector
Equity Long Only Long Only
Distressed Securities Distressed
Single Strategy FoF Market Neutral
Unclassiﬁed
Asset Based Lending
country speciﬁc
Special situations
shortterm trading
Fixed Income  MBS Mortgage
Convertible Arbitrage Convertible Arbitrage
Fixed Income Arbitrage Fixed Income Arbitrage
Other relative Value Other Arbitrage
Market Timing Market Timer
Option Arbitrage Option Strategies
Regulation D Regulation D
Capital Structure Arbitrage Capital Structure Arbitrage
29
Table III: Test of Multiple Structural Changes
See Appendix I for deﬁnitions of fund types. The funds used at minimum cover the LTCM and bubble period which
represent a track record a minimum of 30 months. Listed is a test provided by Bai and Perron (1998) to analyze
whether Hedge Funds have some structural. We note the percentage of Hedge Funds signiﬁcant to the test after
applying the False Discovery Rate methodology . The test considers tests of no structural break against an unknown
number of breaks given some upper bound (m = 5).
Strategy Double Maximum test
UDmax WDmax
Equity Long/Short 38% 38%
Multi Strategy 46% 46%
Emerging Markets 34% 34%
Sector 41% 41%
Equity Market Neutral 49% 49%
Event Driven Multi Strategy 35% 35%
Global Macro 42% 42%
Equity Long Only 53% 53%
Single Strategy 71% 71%
Fixed Income 40% 40%
Distressed Securities 20% 20%
Fixed Income Arbitrage 41% 41%
Convertible Arbitrage 41% 41%
Relative Value Multi Strategy 30% 30%
Fixed Income  MBS 58% 58%
Option Arbitrage 57% 57%
Merger Arbitrage 27% 27%
Other relative Value 75% 75%
Short bias 51% 51%
Regulation D 29% 29%
Capital Structure Arbitrage 53% 53%
Market Timing 100% 100%
Unclassiﬁed 38% 38%
Fund of Hedge Funds 65% 66%
CTA Systematic 70% 70%
CTA Discretionary 66% 66%
30
T
a
b
l
e
V
:
e
s
t
i
m
a
t
e
d
p
r
o
p
o
r
t
i
o
n
s
o
f
z
e
r
o

a
l
p
h
a
,
u
n
s
k
i
l
l
e
d
,
a
n
d
s
k
i
l
l
e
d
f
u
n
d
s
f
r
o
m
m
e
r
g
e
d
a
t
a
b
a
s
e
S
e
e
A
p
p
e
n
d
i
x
I
f
o
r
d
e
ﬁ
n
i
t
i
o
n
s
o
f
f
u
n
d
t
y
p
e
s
.
T
h
i
s
t
a
b
l
e
d
i
s
p
l
a
y
s
t
h
e
e
s
t
i
m
a
t
e
d
p
r
o
p
o
r
t
i
o
n
s
o
f
z
e
r
o

a
l
p
h
a
,
u
n
s
k
i
l
l
e
d
,
a
n
d
s
k
i
l
l
e
d
f
u
n
d
f
o
r
e
a
c
h
s
t
r
a
t
e
g
y
a
f
t
e
r
a
p
p
l
y
i
n
g
t
h
e
f
a
l
s
e
D
i
s
c
o
v
e
r
y
r
a
t
e
m
e
t
h
o
d
o
l
o
g
y
d
e
v
e
l
o
p
e
d
b
y
B
a
r
r
a
s
,
S
c
a
i
l
l
e
t
a
n
d
W
e
r
m
e
r
s
(
2
0
0
8
)
.
T
h
e
s
e
r
e
s
u
l
t
s
c
o
m
e
f
r
o
m
t
h
e
m
e
r
g
e
b
e
t
w
e
e
n
t
h
e
C
I
S
D
M
a
n
d
t
h
e
H
e
g
d
e
F
u
n
d
.
n
e
t
d
a
t
a
b
a
s
e
s
.
T
h
e
f
u
n
d
s
u
s
e
d
a
t
m
i
n
i
m
u
m
c
o
v
e
r
t
h
e
L
T
C
M
a
n
d
b
u
b
b
l
e
p
e
r
i
o
d
w
h
i
c
h
r
e
p
r
e
s
e
n
t
a
t
r
a
c
k
r
e
c
o
r
d
a
m
i
n
i
m
u
m
o
f
3
0
m
o
n
t
h
s
.
W
e
e
s
t
i
m
a
t
e
a
l
p
h
a
w
i
t
h
t
h
e
t
i
m
e

v
a
r
y
i
n
g
c
o
e
ﬃ
c
i
e
n
t
m
o
d
e
l
d
e
ﬁ
n
e
d
i
n
s
e
c
t
i
o
n
6
.
W
i
t
h
t
h
e
e
s
t
i
m
a
t
e
s
,
w
e
c
r
e
a
t
e
t
h
e
m
e
a
n
(
α
)
w
h
i
c
h
r
e
p
r
e
s
e
n
t
s
t
h
e
s
t
o
c
k

p
i
c
k
e
r
a
b
i
l
i
t
y
a
n
d
t
h
e
i
n
d
i
c
a
t
o
r
s
d
e
ﬁ
n
e
d
i
n
s
e
c
t
i
o
n
7
d
u
r
i
n
g
t
h
e
2
c
r
i
s
i
s
:
α
L
1
a
n
d
α
L
2
f
o
r
L
T
C
M
a
n
d
α
B
1
a
n
d
α
B
2
f
o
r
t
h
e
E
q
u
i
t
y
B
u
b
b
l
e
w
h
i
c
h
r
e
p
r
e
s
e
n
t
s
t
h
e
d
i
ﬀ
e
r
e
n
t
m
a
r
k
e
t
t
i
m
e
r
a
b
i
l
i
t
i
e
s
.
W
e
a
l
s
o
g
a
v
e
t
h
e
r
e
s
u
l
t
u
s
i
n
g
a
l
i
n
e
a
r
f
a
c
t
o
r
m
o
d
e
l
:
t
h
e
N
e
w
e
y

W
e
s
t
(
1
9
8
7
)
h
e
t
e
r
o
s
c
e
d
a
s
t
i
c
i
t
y
a
n
d
a
u
t
o
c
o
r
r
e
l
a
t
i
o
n
c
o
n
s
i
s
t
e
n
t
e
s
t
i
m
a
t
o
r
.
W
e
d
o
n
o
t
g
i
v
e
s
o
m
e
r
e
s
u
l
t
s
f
o
r
S
i
n
g
l
e
S
t
r
a
t
e
g
y
b
e
c
a
u
s
e
o
f
t
o
o
f
e
w
d
a
t
a
.
F
u
r
t
h
e
r
m
o
r
e
t
h
e
r
e
s
u
l
t
s
a
b
o
u
t
t
h
e
o
t
h
e
r
s
t
r
a
t
e
g
i
e
s
a
r
e
a
v
a
i
l
a
b
l
e
u
p
o
n
r
e
q
u
e
s
t
.
S
t
r
a
t
e
g
y
M
o
d
e
l
α
α
L
1
α
L
2
α
B
1
α
B
2
u
s
e
d
π
+ A
π
0
π
− A
π
+ A
π
0
π
− A
π
+ A
π
0
π
− A
π
+ A
π
0
π
− A
π
+ A
π
0
π
− A
E
q
.
L
o
n
g
/
S
h
o
r
t
T
V
C
M
1
8
.
6
%
5
8
.
8
%
2
2
.
6
%
1
6
.
0
%
6
1
.
9
%
2
2
.
1
%
2
2
.
8
%
6
2
.
2
%
1
5
.
0
%
1
7
.
0
%
5
8
.
8
%
2
4
.
2
%
1
7
.
0
%
5
8
.
8
%
2
4
.
2
%
N
w
e
s
t
3
.
8
%
%
9
3
.
5
%
2
.
7
%
%
E
m
e
r
g
i
n
g
M
a
r
k
e
t
s
T
V
C
M
1
2
.
6
%
4
1
.
5
%
4
5
.
9
%
4
4
.
2
%
4
2
.
6
%
1
3
.
2
%
3
8
.
9
%
4
3
.
8
%
1
7
.
3
%
4
4
.
4
%
3
9
.
2
%
1
6
.
4
%
4
3
.
6
%
3
9
.
2
%
1
7
.
3
%
N
w
e
s
t
1
2
.
3
%
8
7
.
5
%
0
.
2
%
E
q
.
M
a
r
k
e
t
N
e
u
t
r
a
l
T
V
C
M
1
1
.
6
%
7
4
.
1
%
1
4
.
2
%
0
.
0
%
9
0
.
4
%
9
.
6
%
0
.
0
%
9
2
.
2
%
7
.
8
%
4
.
3
%
8
3
.
2
%
1
2
.
5
%
1
6
.
4
%
8
3
.
2
%
0
.
5
%
N
w
e
s
t
8
.
6
%
8
9
%
2
.
4
%
E
v
e
n
t
D
r
i
v
e
n
M
.
S
.
T
V
C
M
2
.
5
%
9
7
.
5
%
0
.
0
%
0
.
0
%
9
7
.
5
%
2
.
5
%
0
.
0
%
9
9
.
8
%
0
.
2
%
0
.
9
%
9
0
.
7
%
8
.
4
%
0
.
9
%
9
0
.
7
%
8
.
4
%
N
w
e
s
t
0
%
1
0
0
%
0
%
G
l
o
b
a
l
M
a
c
r
o
T
V
C
M
5
.
2
%
6
7
.
0
%
2
7
.
8
%
2
6
.
1
%
6
9
.
1
%
4
.
8
%
2
6
.
2
%
6
6
.
8
%
7
.
0
%
1
8
.
5
%
5
9
.
9
%
2
1
.
6
%
1
8
.
5
%
5
9
.
9
%
2
1
.
6
%
N
w
e
s
t
1
%
9
6
%
3
%
S
h
o
r
t
b
i
a
s
T
V
C
M
3
5
.
2
%
9
.
5
%
5
5
.
2
%
7
1
.
4
%
0
.
0
%
2
8
.
6
%
7
1
.
4
%
0
.
0
%
2
8
.
6
%
4
9
.
0
%
1
9
.
0
%
3
1
.
9
%
4
9
.
0
%
1
9
.
0
%
3
1
.
9
%
N
w
e
s
t
5
.
3
%
6
3
.
6
%
3
1
.
2
%
C
T
A
T
V
C
M
1
8
.
8
%
5
3
.
1
%
2
8
.
1
%
1
7
.
0
%
6
0
.
2
%
2
2
.
8
%
1
7
.
0
%
6
0
.
2
%
2
2
.
8
%
2
7
.
2
%
5
5
.
3
%
1
7
.
5
%
2
7
.
2
%
5
5
.
3
%
1
7
.
5
%
N
w
e
s
t
2
%
9
7
%
1
%
31
Figure I:
31−Mar−1997 30−Nov−1998 31−Jul−2000 31−Mar−2002 30−Nov−2003 31−Jul−2005 31−Mar−2007
0
0.2
0.4
0.6
0.8
1
LTCM Equity Bubble
R
breaks
Capital Structure Arbitrage
31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007
0
0.05
0.1
0.15
0.2
LTCM Equity Bubble
R
breaks
Convertible Arbitrage
31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007
0
0.05
0.1
0.15
0.2
LTCM Equity Bubble
R
breaks
CTA
31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007
0
0.05
0.1
0.15
0.2
LTCM Equity Bubble
R
breaks
Distressed Securities
31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007
0
0.05
0.1
0.15
0.2
0.25
LTCM Equity Bubble
R
breaks
Emerging Markets
31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007
0
0.1
0.2
0.3
0.4
LTCM Equity Bubble
R
breaks
Equity Long Only
See Appendix I for deﬁnitions of fund types. The number of hedge Funds has increased drastically since 2000 therefore
the number of breaks in 1995, for instance, has not the same signiﬁcance as the number of breaks in 2002. In order
to overcome this problem and to study correctly the dynamic in beta we deﬁne a ratio: R
breaks
which is equal to
the number of funds at time t divided by the number of break at time t. In this way, we are able to study how the
dynamic in hedge Funds behaves during a long period.
The bar ﬁgures illustrate the number of breakdates relative to one fund.
32
Figure II:
31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007
0
0.05
0.1
0.15
0.2
LTCM Equity Bubble
R
breaks
Equity Long/Short
31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007
0
0.05
0.1
0.15
0.2
LTCM Equity Bubble
R
breaks
Equity Market Neutral
31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007
0
0.05
0.1
0.15
0.2
0.25
LTCM Equity Bubble
R
breaks
Event Driven Multi Strategy
31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007
0
0.1
0.2
0.3
0.4
0.5
LTCM Equity Bubble
R
breaks
Fixed Income − MBS
31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007
0
0.1
0.2
0.3
0.4
LTCM Equity Bubble
R
breaks
Fixed Income Arbitrage
31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007
0
0.05
0.1
0.15
0.2
0.25
LTCM Equity Bubble
R
breaks
Fixed Income
See Appendix I for deﬁnitions of fund types. The number of hedge Funds has increased drastically since 2000 therefore
the number of breaks in 1995, for instance, has not the same signiﬁcance as the number of breaks in 2002. In order
to overcome this problem and to study correctly the dynamic in beta we deﬁne a ratio: R
breaks
which is equal to
the number of funds at time t divided by the number of break at time t. In this way, we are able to study how the
dynamic in hedge Funds behaves during a long period.
The bar ﬁgures illustrate the number of breakdates relative to one fund.
33
Figure III:
31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007
0
0.05
0.1
0.15
0.2
LTCM Equity Bubble
R
breaks
Global Macro
31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007
0
0.1
0.2
0.3
0.4
LTCM Equity Bubble
R
breaks
Merger Arbitrage
31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007
0
0.1
0.2
0.3
0.4
0.5
LTCM Equity Bubble
R
breaks
Multi Strategy
31−Dec−1993 30−Nov−1995 30−Sep−1997 31−Aug−1999 30−Jun−2001 31−May−2003 31−Mar−2005 28−Feb−2007 31−Dec−1993
LTCM Equity Bubble
Option Arbitrage
31−Dec−1993 30−Nov−1995 30−Sep−1997 31−Aug−1999 30−Jun−2001 31−May−2003 31−Mar−2005 28−Feb−2007 31−Dec−1993
LTCM Equity Bubble
Other Relative Value
31−Dec−1993 30−Nov−1995 30−Sep−1997 31−Aug−1999 30−Jun−2001 31−May−2003 31−Mar−2005 28−Feb−2007 31−Dec−1993
LTCM Equity Bubble
Regulation D
See Appendix I for deﬁnitions of fund types. The number of hedge Funds has increased drastically since 2000 therefore
the number of breaks in 1995, for instance, has not the same signiﬁcance as the number of breaks in 2002. In order
to overcome this problem and to study correctly the dynamic in beta we deﬁne a ratio: R
breaks
which is equal to
the number of funds at time t divided by the number of break at time t. In this way, we are able to study how the
dynamic in hedge Funds behaves during a long period.
The bar ﬁgures illustrate the number of breakdates relative to one fund.
34
Figure IV:
31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007
0
0.05
0.1
0.15
0.2
0.25
LTCM Equity Bubble
R
breaks
Relative Value Multi Strategy
31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007
0
0.05
0.1
0.15
0.2
0.25
LTCM Equity Bubble
R
breaks
Sector
31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007
0
0.1
0.2
0.3
0.4
0.5
LTCM Equity Bubble
R
breaks
Short Bias
28−Feb−1994 31−Jan−1996 31−Dec−1997 30−Nov−1999 30−Sep−2001 31−Aug−2003 31−Jul−2005 30−Jun−2007 28−Feb−1994
0
0.2
0.4
0.6
0.8
1
LTCM Equity Bubble
R
breaks
Single Strategy
31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007
0
0.05
0.1
0.15
0.2
LTCM Equity Bubble
R
breaks
unclassified
See Appendix I for deﬁnitions of fund types. The number of hedge Funds has increased drastically since 2000 therefore
the number of breaks in 1995, for instance, has not the same signiﬁcance as the number of breaks in 2002. In order
to overcome this problem and to study correctly the dynamic in beta we deﬁne a ratio: R
breaks
which is equal to
the number of funds at time t divided by the number of break at time t. In this way, we are able to study how the
dynamic in hedge Funds behaves during a long period.
The bar ﬁgures illustrate the number of breakdates relative to one fund.
35
F
i
g
u
r
e
V
:
L
T
C
M
f
i
r
s
t
P
e
r
i
o
d
0
.
0
0
9
6
0
.
0
1
9
0
.
0
2
9
0
.
0
3
8
0
.
0
4
8
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
L
T
C
M
S
e
c
o
n
d
P
e
r
i
o
d
0
.
0
0
9
9
0
.
0
2
0
.
0
3
0
.
0
4
0
.
0
4
9
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
E
q
u
i
t
y
B
u
b
b
l
e
F
i
r
s
t
P
e
r
i
o
d
0
.
0
0
5
0
.
0
1
0
.
0
1
5
0
.
0
2
0
.
0
2
5
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
E
q
u
i
t
y
B
u
b
b
l
e
S
e
c
o
n
d
P
e
r
i
o
d
0
.
0
0
4
1
0
.
0
0
8
2
0
.
0
1
2
0
.
0
1
6
0
.
0
2
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
0
%
5
0
%
1
0
0
%
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
L
T
C
M
f
i
r
s
t
p
e
r
i
o
d
f a c t o r
0
%
5
0
%
1
0
0
%
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
L
T
C
M
s
e
c
o
n
d
p
e
r
i
o
d
f a c t o r
0
%
5
0
%
1
0
0
%
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
E
q
u
i
t
y
B
u
b
b
l
e
f
i
r
s
t
p
e
r
i
o
d
f a c t o r
0
%
5
0
%
1
0
0
%
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M E
q
u
i
t
y
B
u
b
b
l
e
s
e
c
o
n
d
p
e
r
i
o
d
f a c t o r
π
A −
π
0
π
A +
E
m
e
r
g
i
n
g
M
a
r
k
e
t
s
L
e
g
e
n
d
:
1
)
T
F
B
:
B
o
n
d
T
r
e
n
d
−
F
o
l
l
o
w
i
n
g
f
a
c
t
o
r
2
)
B
T
C
o
:
C
o
m
m
o
d
i
t
y
T
r
e
n
d
−
F
o
l
l
o
w
i
n
g
f
a
c
t
o
r
3
)
T
F
C
u
:
C
u
r
r
e
n
c
y
T
r
e
n
d
−
F
o
l
l
o
w
i
n
g
f
a
c
t
o
r
4
)
S
P
5
0
0
:
S
&
P
5
0
0
5
)
S
S
d
:
S
i
z
e
S
p
r
e
a
d
f
a
c
t
o
r
6
)
B
M
:
B
o
n
d
m
a
r
k
e
t
f
a
c
t
o
r
7
)
C
S
:
C
r
e
d
i
t
S
p
r
e
a
d
f
a
c
t
o
r
8
)
E
M
:
E
m
e
r
g
i
n
g
M
a
r
k
e
t
f
a
c
t
o
r
S
e
e
A
p
p
e
n
d
i
x
I
f
o
r
d
e
ﬁ
n
i
t
i
o
n
s
o
f
f
u
n
d
t
y
p
e
s
.
T
h
e
n
u
m
b
e
r
o
f
f
u
n
d
s
c
o
v
e
r
i
n
g
t
h
e
p
e
r
i
o
d
i
s
e
q
u
a
l
t
o
4
6
1
.
T
h
e
b
a
r
ﬁ
g
u
r
e
s
i
l
l
u
s
t
r
a
t
e
t
h
e
p
r
o
p
o
r
t
i
o
n
o
f
f
u
n
d
s
h
a
v
i
n
g
m
o
r
e
t
h
a
n
1
0
p
e
r
c
e
n
t
d
e
c
r
e
a
s
e
(
b
l
u
e
o
r
π
− A
)
,
c
o
n
s
t
a
n
t
(
w
h
i
t
e
o
r
π
0
)
,
a
n
d
i
n
c
r
e
a
s
e
(
r
e
d
o
r
π
+ A
)
m
a
r
k
e
t
e
x
p
o
s
u
r
e
d
u
r
i
n
g
t
h
e
2
c
r
i
s
i
s
.
E
a
c
h
c
r
i
s
i
s
i
s
d
i
v
i
d
e
d
i
n
2
p
e
r
i
o
d
(
s
e
e
s
e
c
t
i
o
n
7
:
m
e
t
h
o
d
o
l
o
g
y
)
.
.
T
h
e
s
e
c
o
n
d
ﬁ
g
u
r
e
s
(
F
a
c
t
o
r
R
a
d
a
r
C
h
a
r
t
)
i
n
d
i
c
a
t
e
s
t
h
e
s
t
r
a
t
e
g
y
’
s
s
e
n
s
i
t
i
v
i
t
i
e
s
(
p
e
r
c
e
n
t
a
g
e
c
h
a
n
g
e
)
t
o
v
a
r
i
o
u
s
f
a
c
t
o
r
s
i
n
r
e
g
a
r
d
s
t
o
t
h
e
2
c
r
i
s
i
s
.
36
F
i
g
u
r
e
V
I
:
L
T
C
M
f
i
r
s
t
P
e
r
i
o
d
0
.
0
0
4
3
0
.
0
0
8
7
0
.
0
1
3
0
.
0
1
7
0
.
0
2
2
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
L
T
C
M
S
e
c
o
n
d
P
e
r
i
o
d
0
.
0
0
4
1
0
.
0
0
8
2
0
.
0
1
2
0
.
0
1
6
0
.
0
2
1
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
E
q
u
i
t
y
B
u
b
b
l
e
F
i
r
s
t
P
e
r
i
o
d
0
.
0
0
5
7
0
.
0
1
1
0
.
0
1
7
0
.
0
2
3
0
.
0
2
8
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
E
q
u
i
t
y
B
u
b
b
l
e
S
e
c
o
n
d
P
e
r
i
o
d
0
.
0
0
5
8
0
.
0
1
2
0
.
0
1
7
0
.
0
2
3
0
.
0
2
9
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
0
%
5
0
%
1
0
0
%
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
L
T
C
M
f
i
r
s
t
p
e
r
i
o
d
f a c t o r
0
%
5
0
%
1
0
0
%
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
L
T
C
M
s
e
c
o
n
d
p
e
r
i
o
d
f a c t o r
0
%
5
0
%
1
0
0
%
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
E
q
u
i
t
y
B
u
b
b
l
e
f
i
r
s
t
p
e
r
i
o
d
f a c t o r
0
%
5
0
%
1
0
0
%
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M E
q
u
i
t
y
B
u
b
b
l
e
s
e
c
o
n
d
p
e
r
i
o
d
f a c t o r
π
A −
π
0
π
A +
E
q
u
i
t
y
L
o
n
g
/
S
h
o
r
t
L
e
g
e
n
d
:
1
)
T
F
B
:
B
o
n
d
T
r
e
n
d
−
F
o
l
l
o
w
i
n
g
f
a
c
t
o
r
2
)
B
T
C
o
:
C
o
m
m
o
d
i
t
y
T
r
e
n
d
−
F
o
l
l
o
w
i
n
g
f
a
c
t
o
r
3
)
T
F
C
u
:
C
u
r
r
e
n
c
y
T
r
e
n
d
−
F
o
l
l
o
w
i
n
g
f
a
c
t
o
r
4
)
S
P
5
0
0
:
S
&
P
5
0
0
5
)
S
S
d
:
S
i
z
e
S
p
r
e
a
d
f
a
c
t
o
r
6
)
B
M
:
B
o
n
d
m
a
r
k
e
t
f
a
c
t
o
r
7
)
C
S
:
C
r
e
d
i
t
S
p
r
e
a
d
f
a
c
t
o
r
8
)
E
M
:
E
m
e
r
g
i
n
g
M
a
r
k
e
t
f
a
c
t
o
r
S
e
e
A
p
p
e
n
d
i
x
I
f
o
r
d
e
ﬁ
n
i
t
i
o
n
s
o
f
f
u
n
d
t
y
p
e
s
.
T
h
e
n
u
m
b
e
r
o
f
f
u
n
d
s
c
o
v
e
r
i
n
g
t
h
e
p
e
r
i
o
d
i
s
e
q
u
a
l
t
o
2
5
2
5
.
T
h
e
b
a
r
ﬁ
g
u
r
e
s
i
l
l
u
s
t
r
a
t
e
t
h
e
p
r
o
p
o
r
t
i
o
n
o
f
f
u
n
d
s
h
a
v
i
n
g
m
o
r
e
t
h
a
n
1
0
p
e
r
c
e
n
t
d
e
c
r
e
a
s
e
(
b
l
u
e
o
r
π
− A
)
,
c
o
n
s
t
a
n
t
(
w
h
i
t
e
o
r
π
0
)
,
a
n
d
i
n
c
r
e
a
s
e
(
r
e
d
o
r
π
+ A
)
m
a
r
k
e
t
e
x
p
o
s
u
r
e
d
u
r
i
n
g
t
h
e
2
c
r
i
s
i
s
.
E
a
c
h
c
r
i
s
i
s
i
s
d
i
v
i
d
e
d
i
n
2
p
e
r
i
o
d
(
s
e
e
s
e
c
t
i
o
n
7
:
m
e
t
h
o
d
o
l
o
g
y
)
.
.
T
h
e
s
e
c
o
n
d
ﬁ
g
u
r
e
s
(
F
a
c
t
o
r
R
a
d
a
r
C
h
a
r
t
)
i
n
d
i
c
a
t
e
s
t
h
e
s
t
r
a
t
e
g
y
’
s
s
e
n
s
i
t
i
v
i
t
i
e
s
(
p
e
r
c
e
n
t
a
g
e
c
h
a
n
g
e
)
t
o
v
a
r
i
o
u
s
f
a
c
t
o
r
s
i
n
r
e
g
a
r
d
s
t
o
t
h
e
2
c
r
i
s
i
s
.
37
F
i
g
u
r
e
V
I
I
:
L
T
C
M
f
i
r
s
t
P
e
r
i
o
d
0
.
0
0
9
1
0
.
0
1
8
0
.
0
2
7
0
.
0
3
6
0
.
0
4
5
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
L
T
C
M
S
e
c
o
n
d
P
e
r
i
o
d
0
.
0
0
7
9
0
.
0
1
6
0
.
0
2
4
0
.
0
3
2
0
.
0
3
9
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
E
q
u
i
t
y
B
u
b
b
l
e
F
i
r
s
t
P
e
r
i
o
d
0
.
0
0
4
1
0
.
0
0
8
1
0
.
0
1
2
0
.
0
1
6
0
.
0
2
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
E
q
u
i
t
y
B
u
b
b
l
e
S
e
c
o
n
d
P
e
r
i
o
d
0
.
0
0
4
3
0
.
0
0
8
5
0
.
0
1
3
0
.
0
1
7
0
.
0
2
1
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
0
%
5
0
%
1
0
0
%
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
L
T
C
M
f
i
r
s
t
p
e
r
i
o
d
f a c t o r
0
%
5
0
%
1
0
0
%
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
L
T
C
M
s
e
c
o
n
d
p
e
r
i
o
d
f a c t o r
0
%
5
0
%
1
0
0
%
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
E
q
u
i
t
y
B
u
b
b
l
e
f
i
r
s
t
p
e
r
i
o
d
f a c t o r
0
%
5
0
%
1
0
0
%
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M E
q
u
i
t
y
B
u
b
b
l
e
s
e
c
o
n
d
p
e
r
i
o
d
f a c t o r
π
A −
π
0
π
A +
E
q
u
i
t
y
M
a
r
k
e
t
N
e
u
t
r
a
l
L
e
g
e
n
d
:
1
)
T
F
B
:
B
o
n
d
T
r
e
n
d
−
F
o
l
l
o
w
i
n
g
f
a
c
t
o
r
2
)
B
T
C
o
:
C
o
m
m
o
d
i
t
y
T
r
e
n
d
−
F
o
l
l
o
w
i
n
g
f
a
c
t
o
r
3
)
T
F
C
u
:
C
u
r
r
e
n
c
y
T
r
e
n
d
−
F
o
l
l
o
w
i
n
g
f
a
c
t
o
r
4
)
S
P
5
0
0
:
S
&
P
5
0
0
5
)
S
S
d
:
S
i
z
e
S
p
r
e
a
d
f
a
c
t
o
r
6
)
B
M
:
B
o
n
d
m
a
r
k
e
t
f
a
c
t
o
r
7
)
C
S
:
C
r
e
d
i
t
S
p
r
e
a
d
f
a
c
t
o
r
8
)
E
M
:
E
m
e
r
g
i
n
g
M
a
r
k
e
t
f
a
c
t
o
r
S
e
e
A
p
p
e
n
d
i
x
I
f
o
r
d
e
ﬁ
n
i
t
i
o
n
s
o
f
f
u
n
d
t
y
p
e
s
.
T
h
e
n
u
m
b
e
r
o
f
f
u
n
d
s
c
o
v
e
r
i
n
g
t
h
e
p
e
r
i
o
d
i
s
e
q
u
a
l
t
o
5
1
9
.
T
h
e
b
a
r
ﬁ
g
u
r
e
s
i
l
l
u
s
t
r
a
t
e
t
h
e
p
r
o
p
o
r
t
i
o
n
o
f
f
u
n
d
s
h
a
v
i
n
g
m
o
r
e
t
h
a
n
1
0
p
e
r
c
e
n
t
d
e
c
r
e
a
s
e
(
b
l
u
e
o
r
π
− A
)
,
c
o
n
s
t
a
n
t
(
w
h
i
t
e
o
r
π
0
)
,
a
n
d
i
n
c
r
e
a
s
e
(
r
e
d
o
r
π
+ A
)
m
a
r
k
e
t
e
x
p
o
s
u
r
e
d
u
r
i
n
g
t
h
e
2
c
r
i
s
i
s
.
E
a
c
h
c
r
i
s
i
s
i
s
d
i
v
i
d
e
d
i
n
2
p
e
r
i
o
d
(
s
e
e
s
e
c
t
i
o
n
7
:
m
e
t
h
o
d
o
l
o
g
y
)
.
.
T
h
e
s
e
c
o
n
d
ﬁ
g
u
r
e
s
(
F
a
c
t
o
r
R
a
d
a
r
C
h
a
r
t
)
i
n
d
i
c
a
t
e
s
t
h
e
s
t
r
a
t
e
g
y
’
s
s
e
n
s
i
t
i
v
i
t
i
e
s
(
p
e
r
c
e
n
t
a
g
e
c
h
a
n
g
e
)
t
o
v
a
r
i
o
u
s
f
a
c
t
o
r
s
i
n
r
e
g
a
r
d
s
t
o
t
h
e
2
c
r
i
s
i
s
.
38
F
i
g
u
r
e
V
I
I
I
:
L
T
C
M
f
i
r
s
t
P
e
r
i
o
d
0
.
0
0
5
7
0
.
0
1
1
0
.
0
1
7
0
.
0
2
3
0
.
0
2
9
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
L
T
C
M
S
e
c
o
n
d
P
e
r
i
o
d
0
.
0
0
4
2
0
.
0
0
8
5
0
.
0
1
3
0
.
0
1
7
0
.
0
2
1
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
E
q
u
i
t
y
B
u
b
b
l
e
F
i
r
s
t
P
e
r
i
o
d
0
.
0
0
8
4
0
.
0
1
7
0
.
0
2
5
0
.
0
3
4
0
.
0
4
2
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
E
q
u
i
t
y
B
u
b
b
l
e
S
e
c
o
n
d
P
e
r
i
o
d
0
.
0
0
8
4
0
.
0
1
7
0
.
0
2
5
0
.
0
3
4
0
.
0
4
2
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
0
%
5
0
%
1
0
0
%
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
L
T
C
M
f
i
r
s
t
p
e
r
i
o
d
f a c t o r
0
%
5
0
%
1
0
0
%
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
L
T
C
M
s
e
c
o
n
d
p
e
r
i
o
d
f a c t o r
0
%
5
0
%
1
0
0
%
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
E
q
u
i
t
y
B
u
b
b
l
e
f
i
r
s
t
p
e
r
i
o
d
f a c t o r
0
%
5
0
%
1
0
0
%
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M E
q
u
i
t
y
B
u
b
b
l
e
s
e
c
o
n
d
p
e
r
i
o
d
f a c t o r
π
A −
π
0
π
A +
E
v
e
n
t
D
r
i
v
e
n
M
u
l
t
i
S
t
r
a
t
e
g
y
L
e
g
e
n
d
:
1
)
T
F
B
:
B
o
n
d
T
r
e
n
d
−
F
o
l
l
o
w
i
n
g
f
a
c
t
o
r
2
)
B
T
C
o
:
C
o
m
m
o
d
i
t
y
T
r
e
n
d
−
F
o
l
l
o
w
i
n
g
f
a
c
t
o
r
3
)
T
F
C
u
:
C
u
r
r
e
n
c
y
T
r
e
n
d
−
F
o
l
l
o
w
i
n
g
f
a
c
t
o
r
4
)
S
P
5
0
0
:
S
&
P
5
0
0
5
)
S
S
d
:
S
i
z
e
S
p
r
e
a
d
f
a
c
t
o
r
6
)
B
M
:
B
o
n
d
m
a
r
k
e
t
f
a
c
t
o
r
7
)
C
S
:
C
r
e
d
i
t
S
p
r
e
a
d
f
a
c
t
o
r
8
)
E
M
:
E
m
e
r
g
i
n
g
M
a
r
k
e
t
f
a
c
t
o
r
S
e
e
A
p
p
e
n
d
i
x
I
f
o
r
d
e
ﬁ
n
i
t
i
o
n
s
o
f
f
u
n
d
t
y
p
e
s
.
T
h
e
n
u
m
b
e
r
o
f
f
u
n
d
s
c
o
v
e
r
i
n
g
t
h
e
p
e
r
i
o
d
i
s
e
q
u
a
l
t
o
2
9
3
.
T
h
e
b
a
r
ﬁ
g
u
r
e
s
i
l
l
u
s
t
r
a
t
e
t
h
e
p
r
o
p
o
r
t
i
o
n
o
f
f
u
n
d
s
h
a
v
i
n
g
m
o
r
e
t
h
a
n
1
0
p
e
r
c
e
n
t
d
e
c
r
e
a
s
e
(
b
l
u
e
o
r
π
− A
)
,
c
o
n
s
t
a
n
t
(
w
h
i
t
e
o
r
π
0
)
,
a
n
d
i
n
c
r
e
a
s
e
(
r
e
d
o
r
π
+ A
)
m
a
r
k
e
t
e
x
p
o
s
u
r
e
d
u
r
i
n
g
t
h
e
2
c
r
i
s
i
s
.
E
a
c
h
c
r
i
s
i
s
i
s
d
i
v
i
d
e
d
i
n
2
p
e
r
i
o
d
(
s
e
e
s
e
c
t
i
o
n
7
:
m
e
t
h
o
d
o
l
o
g
y
)
.
.
T
h
e
s
e
c
o
n
d
ﬁ
g
u
r
e
s
(
F
a
c
t
o
r
R
a
d
a
r
C
h
a
r
t
)
i
n
d
i
c
a
t
e
s
t
h
e
s
t
r
a
t
e
g
y
’
s
s
e
n
s
i
t
i
v
i
t
i
e
s
(
p
e
r
c
e
n
t
a
g
e
c
h
a
n
g
e
)
t
o
v
a
r
i
o
u
s
f
a
c
t
o
r
s
i
n
r
e
g
a
r
d
s
t
o
t
h
e
2
c
r
i
s
i
s
.
39
F
i
g
u
r
e
I
X
:
L
T
C
M
f
i
r
s
t
P
e
r
i
o
d
0
.
0
0
6
7
0
.
0
1
3
0
.
0
2
0
.
0
2
7
0
.
0
3
4
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
L
T
C
M
S
e
c
o
n
d
P
e
r
i
o
d
0
.
0
0
6
0
.
0
1
2
0
.
0
1
8
0
.
0
2
4
0
.
0
3
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
E
q
u
i
t
y
B
u
b
b
l
e
F
i
r
s
t
P
e
r
i
o
d
0
.
0
0
4
3
0
.
0
0
8
5
0
.
0
1
3
0
.
0
1
7
0
.
0
2
1
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
E
q
u
i
t
y
B
u
b
b
l
e
S
e
c
o
n
d
P
e
r
i
o
d
0
.
0
0
3
4
0
.
0
0
6
9
0
.
0
1
0
.
0
1
4
0
.
0
1
7
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
0
%
5
0
%
1
0
0
%
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
L
T
C
M
f
i
r
s
t
p
e
r
i
o
d
f a c t o r
0
%
5
0
%
1
0
0
%
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
L
T
C
M
s
e
c
o
n
d
p
e
r
i
o
d
f a c t o r
0
%
5
0
%
1
0
0
%
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
E
q
u
i
t
y
B
u
b
b
l
e
f
i
r
s
t
p
e
r
i
o
d
f a c t o r
0
%
5
0
%
1
0
0
%
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M E
q
u
i
t
y
B
u
b
b
l
e
s
e
c
o
n
d
p
e
r
i
o
d
f a c t o r
π
A −
π
0
π
A +
G
l
o
b
a
l
M
a
c
r
o
L
e
g
e
n
d
:
1
)
T
F
B
:
B
o
n
d
T
r
e
n
d
−
F
o
l
l
o
w
i
n
g
f
a
c
t
o
r
2
)
B
T
C
o
:
C
o
m
m
o
d
i
t
y
T
r
e
n
d
−
F
o
l
l
o
w
i
n
g
f
a
c
t
o
r
3
)
T
F
C
u
:
C
u
r
r
e
n
c
y
T
r
e
n
d
−
F
o
l
l
o
w
i
n
g
f
a
c
t
o
r
4
)
S
P
5
0
0
:
S
&
P
5
0
0
5
)
S
S
d
:
S
i
z
e
S
p
r
e
a
d
f
a
c
t
o
r
6
)
B
M
:
B
o
n
d
m
a
r
k
e
t
f
a
c
t
o
r
7
)
C
S
:
C
r
e
d
i
t
S
p
r
e
a
d
f
a
c
t
o
r
8
)
E
M
:
E
m
e
r
g
i
n
g
M
a
r
k
e
t
f
a
c
t
o
r
S
e
e
A
p
p
e
n
d
i
x
I
f
o
r
d
e
ﬁ
n
i
t
i
o
n
s
o
f
f
u
n
d
t
y
p
e
s
.
T
h
e
n
u
m
b
e
r
o
f
f
u
n
d
s
c
o
v
e
r
i
n
g
t
h
e
p
e
r
i
o
d
i
s
e
q
u
a
l
t
o
3
7
7
.
T
h
e
b
a
r
ﬁ
g
u
r
e
s
i
l
l
u
s
t
r
a
t
e
t
h
e
p
r
o
p
o
r
t
i
o
n
o
f
f
u
n
d
s
h
a
v
i
n
g
m
o
r
e
t
h
a
n
1
0
p
e
r
c
e
n
t
d
e
c
r
e
a
s
e
(
b
l
u
e
o
r
π
− A
)
,
c
o
n
s
t
a
n
t
(
w
h
i
t
e
o
r
π
0
)
,
a
n
d
i
n
c
r
e
a
s
e
(
r
e
d
o
r
π
+ A
)
m
a
r
k
e
t
e
x
p
o
s
u
r
e
d
u
r
i
n
g
t
h
e
2
c
r
i
s
i
s
.
E
a
c
h
c
r
i
s
i
s
i
s
d
i
v
i
d
e
d
i
n
2
p
e
r
i
o
d
(
s
e
e
s
e
c
t
i
o
n
7
:
m
e
t
h
o
d
o
l
o
g
y
)
.
.
T
h
e
s
e
c
o
n
d
ﬁ
g
u
r
e
s
(
F
a
c
t
o
r
R
a
d
a
r
C
h
a
r
t
)
i
n
d
i
c
a
t
e
s
t
h
e
s
t
r
a
t
e
g
y
’
s
s
e
n
s
i
t
i
v
i
t
i
e
s
(
p
e
r
c
e
n
t
a
g
e
c
h
a
n
g
e
)
t
o
v
a
r
i
o
u
s
f
a
c
t
o
r
s
i
n
r
e
g
a
r
d
s
t
o
t
h
e
2
c
r
i
s
i
s
.
40
F
i
g
u
r
e
X
:
L
T
C
M
f
i
r
s
t
P
e
r
i
o
d
0
.
0
1
0
.
0
2
0
.
0
3
0
.
0
4
0
.
0
5
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
L
T
C
M
S
e
c
o
n
d
P
e
r
i
o
d
0
.
0
0
6
7
0
.
0
1
3
0
.
0
2
0
.
0
2
7
0
.
0
3
3
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
E
q
u
i
t
y
B
u
b
b
l
e
F
i
r
s
t
P
e
r
i
o
d
0
.
0
0
5
1
0
.
0
1
0
.
0
1
5
0
.
0
2
1
0
.
0
2
6
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
E
q
u
i
t
y
B
u
b
b
l
e
S
e
c
o
n
d
P
e
r
i
o
d
0
.
0
0
5
4
0
.
0
1
1
0
.
0
1
6
0
.
0
2
2
0
.
0
2
7
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
0
%
5
0
%
1
0
0
%
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
L
T
C
M
f
i
r
s
t
p
e
r
i
o
d
f a c t o r
0
%
5
0
%
1
0
0
%
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
L
T
C
M
s
e
c
o
n
d
p
e
r
i
o
d
f a c t o r
0
%
5
0
%
1
0
0
%
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
E
q
u
i
t
y
B
u
b
b
l
e
f
i
r
s
t
p
e
r
i
o
d
f a c t o r
0
%
5
0
%
1
0
0
%
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M E
q
u
i
t
y
B
u
b
b
l
e
s
e
c
o
n
d
p
e
r
i
o
d
f a c t o r
π
A −
π
0
π
A +
S
h
o
r
t
B
i
a
s
L
e
g
e
n
d
:
1
)
T
F
B
:
B
o
n
d
T
r
e
n
d
−
F
o
l
l
o
w
i
n
g
f
a
c
t
o
r
2
)
B
T
C
o
:
C
o
m
m
o
d
i
t
y
T
r
e
n
d
−
F
o
l
l
o
w
i
n
g
f
a
c
t
o
r
3
)
T
F
C
u
:
C
u
r
r
e
n
c
y
T
r
e
n
d
−
F
o
l
l
o
w
i
n
g
f
a
c
t
o
r
4
)
S
P
5
0
0
:
S
&
P
5
0
0
5
)
S
S
d
:
S
i
z
e
S
p
r
e
a
d
f
a
c
t
o
r
6
)
B
M
:
B
o
n
d
m
a
r
k
e
t
f
a
c
t
o
r
7
)
C
S
:
C
r
e
d
i
t
S
p
r
e
a
d
f
a
c
t
o
r
8
)
E
M
:
E
m
e
r
g
i
n
g
M
a
r
k
e
t
f
a
c
t
o
r
S
e
e
A
p
p
e
n
d
i
x
I
f
o
r
d
e
ﬁ
n
i
t
i
o
n
s
o
f
f
u
n
d
t
y
p
e
s
.
T
h
e
n
u
m
b
e
r
o
f
f
u
n
d
s
c
o
v
e
r
i
n
g
t
h
e
p
e
r
i
o
d
i
s
e
q
u
a
l
t
o
5
6
.
T
h
e
b
a
r
ﬁ
g
u
r
e
s
i
l
l
u
s
t
r
a
t
e
t
h
e
p
r
o
p
o
r
t
i
o
n
o
f
f
u
n
d
s
h
a
v
i
n
g
m
o
r
e
t
h
a
n
1
0
p
e
r
c
e
n
t
d
e
c
r
e
a
s
e
(
b
l
u
e
o
r
π
− A
)
,
c
o
n
s
t
a
n
t
(
w
h
i
t
e
o
r
π
0
)
,
a
n
d
i
n
c
r
e
a
s
e
(
r
e
d
o
r
π
+ A
)
m
a
r
k
e
t
e
x
p
o
s
u
r
e
d
u
r
i
n
g
t
h
e
2
c
r
i
s
i
s
.
E
a
c
h
c
r
i
s
i
s
i
s
d
i
v
i
d
e
d
i
n
2
p
e
r
i
o
d
(
s
e
e
s
e
c
t
i
o
n
7
:
m
e
t
h
o
d
o
l
o
g
y
)
.
.
T
h
e
s
e
c
o
n
d
ﬁ
g
u
r
e
s
(
F
a
c
t
o
r
R
a
d
a
r
C
h
a
r
t
)
i
n
d
i
c
a
t
e
s
t
h
e
s
t
r
a
t
e
g
y
’
s
s
e
n
s
i
t
i
v
i
t
i
e
s
(
p
e
r
c
e
n
t
a
g
e
c
h
a
n
g
e
)
t
o
v
a
r
i
o
u
s
f
a
c
t
o
r
s
i
n
r
e
g
a
r
d
s
t
o
t
h
e
2
c
r
i
s
i
s
.
41
F
i
g
u
r
e
X
I
:
L
T
C
M
f
i
r
s
t
P
e
r
i
o
d
0
.
0
0
3
8
0
.
0
0
7
6
0
.
0
1
1
0
.
0
1
5
0
.
0
1
9
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
L
T
C
M
S
e
c
o
n
d
P
e
r
i
o
d
0
.
0
0
3
9
0
.
0
0
7
7
0
.
0
1
2
0
.
0
1
5
0
.
0
1
9
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
E
q
u
i
t
y
B
u
b
b
l
e
F
i
r
s
t
P
e
r
i
o
d
0
.
0
0
6
4
0
.
0
1
3
0
.
0
1
9
0
.
0
2
6
0
.
0
3
2
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
E
q
u
i
t
y
B
u
b
b
l
e
S
e
c
o
n
d
P
e
r
i
o
d
0
.
0
0
6
2
0
.
0
1
2
0
.
0
1
8
0
.
0
2
5
0
.
0
3
1
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
0
%
5
0
%
1
0
0
%
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
L
T
C
M
f
i
r
s
t
p
e
r
i
o
d
f a c t o r
0
%
5
0
%
1
0
0
%
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
L
T
C
M
s
e
c
o
n
d
p
e
r
i
o
d
f a c t o r
0
%
5
0
%
1
0
0
%
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M
E
q
u
i
t
y
B
u
b
b
l
e
f
i
r
s
t
p
e
r
i
o
d
f a c t o r
0
%
5
0
%
1
0
0
%
T
F
B
T
F
C
u
T
F
C
o
S
P
5
0
0
S
S
d
B
d
M
t
C
S
E
M E
q
u
i
t
y
B
u
b
b
l
e
s
e
c
o
n
d
p
e
r
i
o
d
f a c t o r
π
A −
π
0
π
A +
C
T
A
s
M
e
r
g
e
L
e
g
e
n
d
:
1
)
T
F
B
:
B
o
n
d
T
r
e
n
d
−
F
o
l
l
o
w
i
n
g
f
a
c
t
o
r
2
)
B
T
C
o
:
C
o
m
m
o
d
i
t
y
T
r
e
n
d
−
F
o
l
l
o
w
i
n
g
f
a
c
t
o
r
3
)
T
F
C
u
:
C
u
r
r
e
n
c
y
T
r
e
n
d
−
F
o
l
l
o
w
i
n
g
f
a
c
t
o
r
4
)
S
P
5
0
0
:
S
&
P
5
0
0
5
)
S
S
d
:
S
i
z
e
S
p
r
e
a
d
f
a
c
t
o
r
6
)
B
M
:
B
o
n
d
m
a
r
k
e
t
f
a
c
t
o
r
7
)
C
S
:
C
r
e
d
i
t
S
p
r
e
a
d
f
a
c
t
o
r
8
)
E
M
:
E
m
e
r
g
i
n
g
M
a
r
k
e
t
f
a
c
t
o
r
S
e
e
A
p
p
e
n
d
i
x
I
f
o
r
d
e
ﬁ
n
i
t
i
o
n
s
o
f
f
u
n
d
t
y
p
e
s
.
T
h
e
n
u
m
b
e
r
o
f
f
u
n
d
s
c
o
v
e
r
i
n
g
t
h
e
p
e
r
i
o
d
i
s
e
q
u
a
l
t
o
1
3
9
4
.
T
h
e
b
a
r
ﬁ
g
u
r
e
s
i
l
l
u
s
t
r
a
t
e
t
h
e
p
r
o
p
o
r
t
i
o
n
o
f
f
u
n
d
s
h
a
v
i
n
g
m
o
r
e
t
h
a
n
1
0
p
e
r
c
e
n
t
d
e
c
r
e
a
s
e
(
b
l
u
e
o
r
π
− A
)
,
c
o
n
s
t
a
n
t
(
w
h
i
t
e
o
r
π
0
)
,
a
n
d
i
n
c
r
e
a
s
e
(
r
e
d
o
r
π
+ A
)
m
a
r
k
e
t
e
x
p
o
s
u
r
e
d
u
r
i
n
g
t
h
e
2
c
r
i
s
i
s
.
E
a
c
h
c
r
i
s
i
s
i
s
d
i
v
i
d
e
d
i
n
2
p
e
r
i
o
d
(
s
e
e
s
e
c
t
i
o
n
7
:
m
e
t
h
o
d
o
l
o
g
y
)
.
.
T
h
e
s
e
c
o
n
d
ﬁ
g
u
r
e
s
(
F
a
c
t
o
r
R
a
d
a
r
C
h
a
r
t
)
i
n
d
i
c
a
t
e
s
t
h
e
s
t
r
a
t
e
g
y
’
s
s
e
n
s
i
t
i
v
i
t
i
e
s
(
p
e
r
c
e
n
t
a
g
e
c
h
a
n
g
e
)
t
o
v
a
r
i
o
u
s
f
a
c
t
o
r
s
i
n
r
e
g
a
r
d
s
t
o
t
h
e
2
c
r
i
s
i
s
.
42
Appendix I
Deﬁnition of Strategies
• The emerging markets strategy attempts to capture gains from ineﬃciencies in emerging
markets.
• The Equity Long/Short strategy refers to taking both long and short positions in equities.
• The Market Timer focus on securities associated with companies that will soon experience
a signiﬁcant event.
The Distressed Securities Strategy focuses on asset of distressed companies. Buys equity,
debt, or trade claims at deep discounts of companies in or facing bankruptcy or reorganization.
• The Merger Arbitrage Strategy also called risk arbitrage strategy exploit pricing inef
ﬁciencies associated with a merger or acquisition.
• The event driven multistrategy can use both the distressed securities style and/or the
merger arbitrage style.
• The Relative Value arbitrage style take positions in 2 securities that are mispriced relative
to each other, with the expectation that their prices will converge to appropriate values in the
future(Arbitrage, Market neutral.
• The arbitrage involves simultaneously purchasing and selling related securities that are mis
priced relative to each other.
• Convertible Arbitrage Strategy can be described by taking a long position in a convertible
bond and sells short the associated stock. Convertible arbitrage  exploit pricing ineﬃciencies
between convertible securities and the corresponding stocks.
• Fixed Income Arbitrage Strategies encompass a wide range of strategies in both domestic
and global ﬁxedincome markets. Fixed income arbitrage  exploit pricing ineﬃciencies between
related ﬁxed income securities.
• Equity Market Neutral style creates a position that attempts to hedge out most market
risk by taking oﬀsetting positions. This strategy exploits the mispricing between a stock
which is overvalued and one that is undervalued such that beta of the combined position is
zero. Statistical arbitrage  equity market neutral strategy using statistical models.
• Index arbitrage style generally attempts to exploit mispricing between an index and deriva
tives on that index.
• Mortgagebacked securities arbitrage style exploit the mispricing of mortgagebacked
assets relative to Treasury securities.
• multistrategy style uses diﬀerent styles and may change exposures to diﬀerent styles based
upon changing market conditions.Multi strategy in Macro strategy  combination of discre
tionary and systematic macro. Multi strategy in FoHF  a hedge fund exploiting a combination
of diﬀerent hedge fund strategies to reduce market risk.
• dedicated short selling style only takes short equity positions.
• Global Macro Discretionary macro  trading is done by investment managers instead of
generated by software.
Systematic macro (Systematic diversiﬁed)  trading is done mathematically, generated by
software without human intervention.
43
• Sector funds  expertise in niche areas such as technology, health care, biotechnology, phar
maceuticals, energy, basic materials.
• Fundamental value  invest in undervalued companies.
• Fundamental growth  invest in companies with more earnings growth than the broad equity
market.
• Quantitative Directional, statistical arbitrage  equity trading using quantitative tech
niques.
• Multi manager  a hedge fund where the investment is spread along separate sub managers
investing in their own strategy.
• Trend following  longterm or shortterm. Nontrend following (Counter trend)  proﬁt
from trend reversals.
• Regulation D  specialized in private equities.
• Credit arbitrage or ﬁxed income arbitrage strategy  specialized in corporate ﬁxed
income securities.
• Fixed Income asset backed  ﬁxed income arbitrage strategy using assetbacked securities.
• Volatility arbitrage  exploit the change in implied volatility instead of the change in price.
• Yield alternatives  non ﬁxed income arbitrage strategies based on the yield instead of the
price.
• Capital Structure Arbitrage  involves taking long and short positions in diﬀerent ﬁnancial
instruments of a company’s capital structure, particularly between a company’s debt and
equity product.
44
Twostep TimeVarying Coeﬃcient Model
The varying coeﬃcient model assumes the following conditional linear structure:
Y
t
=
p
j=1
β
j
(t)X
jt
+ε
t
= α(t) +Xβ(t) +ε
t
for a given covariates (t, X
1
, ..., X
p
)
and variable Y with
E[εt, X
1
, ..., X
p
] = 0,
V ar[εt, X
1
, ..., X
p
] = σ
2
(t),
In this study, we took X
1
= 1 as the intercept term and t = time.
if we consider that β
i
depends on t: (β
i
(t)), we can approximate the function locally as
β
i
(t) ≈ a
i
+b
i
(t −t
0
). This leads to the following local leastsquares problem:
minimize
n
i=1
_
_
Y
i
−
p
j=1
{a
j
+b
j
(T
i
−t
0
)}X
ij
_
_
2
K
h
(T
i
−t
0
)
for a given kernel function K and bandwidth h, where K
h
(.) = K(./h)/h.
In matrix notation:
Let
X
0
=
_
_
_
X
11
X
11
(T
1
−t
0
) . . . X
1p
X
1p
(T
1
−t
0
)
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
X
n1
X
n1
(T
n
−t
0
) . . . X
np
X
np
(T
p
−t
0
)
_
_
_
Y = (Y
1
, ..., Y
n
)
and W
0
= diag(K
h
0
(T
1
−t
0
), ..., K
h
0
(T
n
−t
0
))
Then the solution to the leastsquares problem can be expressed as:
´a
j,0
= e
2j−1,2p
(X
0
W
0
X
0
)
−1
X
0
W
0
Y
With these estimates, ´a
j,0
, a local leastsquare regression is ﬁtted again via substituting the
initial estimate into the local leastsquares problem:
n
i=1
_
_
Y
i
−
p−1
j=1
ˆ a
j,0
(T
i
)X
ij
−
_
a
p
+b
p
(T
i
−t
0
) +c
p
(T
i
−t
0
)
2
+d
p
(T
i
−t
0
)
3
_
X
ip
_
_
2
×K
h
2
(T
i
−t
0
)
where h
2
is the bandwidth in the second step. In this way, a twostep estimator is obtained.
Fan and Zhang showed that the bias of the twostep estimator is of O(h
4
2
) and the variance
O
_
(nh
2
)
−1
_
45
Twostep TimeVarying Coeﬃcient Model using Bsplines
As mentioned by Fan and Zhang (1999), other techniques such as smoothing splines can also
be used in the second stage of ﬁtting. Therefore we built the second twostep estimator based
on the same article but we used during the second step a smoothing splines instead of local
regression.
From the ﬁrst step, we obtained the estimates:
´a
j,0
= e
2j−1,2p
(X
0
W
0
X
0
)
−1
X
0
W
0
Y
In a second step, knowing that we can approximate each β
p
(t) by a basis function expansion
β
p
(t)
K
k=0
γ
∗
pk
B
pk
(t),
We can now minimized in order to estimate γ
∗
kp
:
n
i=1
w
i
_
_
Y
i
−
p−1
j=1
ˆ a
j,0
(T
i
)X
ij
−
_
K
k=0
γ
kp
B
kp
_
X
ip
_
_
2
then it is natural to estimate β
p
(t) by
ˆ
β
p
(t) =
K
k=1
ˆ γ
kp
B
kp
(t).
46
Robustness: local and Bspline TimeVarying Coeﬃcient Model
The following example will be used to illustrate the performance of our estimator. We created two “betas
”(called β
i
created
) which represent a possible structural change for a Hedge Fund. We put up a simulated Hedge
Fund track (R
HF
) in this manner:
R
HF
= β
1
created
X
1
+ β
2
created
X2 + ε
where X
1
, X
2
are the S&P500 and the monthly change in the 10year treasury constant maturity yield respec
tively. The random variable ε follows a normal distribution with mean zero and variance 1.
We called the local timevarying coeﬃcient model: LTVCM and the Bspline time varying coeﬃcient model:
BTVCM.
Short sample: 50 months
0 20 40 60
−2
−1
0
1
2
ˆ
β
1
and true β
1
using LTVCM
months
0 20 40 60
−5
0
5
10
ˆ
β
2
and true β
2
using LTVCM
months
0 20 40 60
−2
0
2
4
ˆ
β
2
and true β
2
using BTVCM
months
0 20 40 60
−2
0
2
4
6
ˆ
β
2
and true β
2
using BTVCM
months
FIG.: Comparison of the performance between the onestep estimator (longdashed curve) and the true coeﬃ
cient function (the solid curve).
47
Medium sample: 100 months
0 50 100
−2
−1
0
1
ˆ
β
1
and true β
1
using LTVCM
months
0 50 100
−2
0
2
4
ˆ
β
2
and true β
2
using LTVCM
months
0 50 100
−1
−0.5
0
0.5
1
ˆ
β
2
and true β
2
using BTVCM
months
0 50 100
−2
0
2
4
6
ˆ
β
2
and true β
2
using BTVCM
months
FIG.: Comparison of the performance between the onestep estimator (longdashed curve) and the true coeﬃ
cient function (the solid curve).
Long sample: 150 months
0 50 100 150
−2
−1
0
1
ˆ
β
1
and true β
1
using LTVCM
months
0 50 100 150
−2
0
2
4
ˆ
β
2
and true β
2
using LTVCM
months
0 50 100 150
−1
−0.5
0
0.5
1
ˆ
β
2
and true β
2
using BTVCM
months
0 50 100 150
−2
−1
0
1
2
ˆ
β
2
and true β
2
using BTVCM
months
true β
ˆ
β
FIG.: Comparison of the performance between the onestep estimator (longdashed curve) and the true coeﬃ
cient function (the solid curve).
48
Bootstrap Conﬁdence intervals
We summarize the methodology from Colin and Chiang (2000) in order to create conﬁdence
regions.
According to the author this following naive bootstrap procedure can be used to construct
approximate pointwise percentile conﬁdence intervals for β
i
(t):
• 1) Randomly sample n subjects with replacement from the original data set, and let
{(t
∗
ij
, X
∗
i
), Y
∗
ij
; 1 = i = n, 1 = j = n
i
} be the longitudinal bootstrap sample.
• 2) Compute the kernel estimator
¯
β
boot
i
(t)
• 3) Repeat the above 2 steps B times, so that B bootstrap estimators
¯
β
boot
i
(t) of β
i
(t) are
obtained.
• 4) Let L
α/2
(t) and U
(α/2)
(t) be the (α/2)
th
and (1 − α)
th
i.e. lower and upper (α/2
th
) per
centiles, respectively, calculated on the B bootstrap estimators. An approximate (1 − α)
bootstrap conﬁdence interval for β
i
(t) is given by (L
(α/2)
(t), U
(α/2)
(t)).
bandwidth ”leaveonesubjectout” crossvalidation methodology
The leaveout method is based on regression smoothers in which one, say the jth, observation
is left out. So, for N values,
• 1) Compute the leaveout estimate ˆ m
h,j
(X
j
) = n
−1
i=j
W
hi
(X
j
)Y
i
.
• 2) Construct the cross validation function CV (h) = n
−1
n
j=1
(Y
j
− ˆ m
h,j
(X
j
))
2
w(X
j
). where w
denotes a weight function.
• 3) With this N CV(h), we can, now, deﬁne the automatic bandwidth as
ˆ
h = arg min
h∈H
n
[CV (h)]
49
1
Introduction
Nowadays, it is well known that investing in Mutual Funds, on average, underperform passive investment strategies. One class commonly called Hedge Funds are deﬁned as pooled investment vehicles that are privately organized, and not widely available to the general investing public. Hedge Funds are private partnerships that use advanced investment strategies and can use derivatives, leverage and short selling. Over the last few years, Hedge Funds have become the favorites of many private as well as institutional investors. Hedge Funds follow investment strategies that are substantially diﬀerent from the nonleveraged, longonly strategies conventionally followed by investors. During alternative investment seminars and conferences, Hedge Fund managers boast about their ability to produce something they refer to as “alpha”or “absolute return”in the sense that performance is not due to primary asset classes performance. They do not aim to track and try to beat a certain benchmark, but instead are focused on pure return generation. Hence Hedge Funds generate alpha, as opposed to depending on beta and performance should normally result from active management decisions combined with the skills of their advisors. However statistical analysis shows that many of them retain signiﬁcant exposure to diﬀerent types of market risk factors. Consequently, it is essential for qualiﬁed investors to determine whether or not these strategies are sensitive to the market and whether they can ﬁnd alpha through the manager’s skill. For these reasons, an increasing focus has been directed to the performance of Hedge Funds and their factor exposures. Mostly, owing to the theory of CAPM or APT, fund performances are assessed by a parametric model with the hypothesis of normality and linearity of coeﬃcients, as well as, the nondynamic behavior of beta. Some researchers expanded the parametric model to the world of Hedge Funds. Fung and Hsieh (2001, 2004a) developed factors used to replicate trendfollowing strategies. Agarwal and Naik (2000) suggested an approach using optionbased returns in order to capture the nonlinearity in beta. Recently, Bollen and Whaley (2008) tested for structural change and used two econometric models in order to capture the dynamic in beta. In this paper, we attempt to go a step further than the previous research. First, by developing the result from Bollen and Whaley (2008) by testing up to ﬁve multiple structural changes. We showed that the relative number of breaks by fund has increased over the last few years, resulting in the increase of the use of leverage. 2
Secondly, by developing a new model1 for Hedge Funds which can take into consideration their distinctly nonnormal characteristics, the dynamic in manager skill, as well as, the dynamic trading represented respectively by alpha and beta. This model allows us to relax traditional parametric models and to exploit possible hidden structure. For example, what manager skill is the most important? Is it the skill to pick and choose the right stocks, bonds, any other ﬁnancial products, or is it the skill to anticipate market events? If we consider that alpha estimated from a linear model contains the two skills, how can we separate and study them? Can we ﬁnd a diﬀerent proportion of positivealpha funds? This model speciﬁcally allows us to look into how managers behave in relation to strong events and whether or not they succeed to generate alpha. Moreover, Hedge Fund managers are likely to change trading strategies in order to obtain “absolute return”2 . Therefore, the exposure to a risk factor can change during special events and can have some deep implications for a Fund of Funds or a portfolio. Often it has been merely stated that during an event, risk to a speciﬁc factor can increase or decrease. Our study also portrays how the beta exposure or risk behaves in relation to market reaction, as well as, alpha during two agitated periods; the equity bubble crisis (the Equity Bubble hereafter)3 and LTCM crisis (LTCM hereafter) 4 . Thirdly, with the methodology used for mutual Funds by Barras, Scaillet and Wermers (2008) called False Discovery Rate, we look into the proportion of the fund’s population which has an increase in diﬀerent market exposures, as well as, the proportion of skilled, unskilled and zeroalpha funds. We showed that the proportion of “skilled”funds are higher with our model than with a static linear factor model. We explained such a diﬀerence by our model’s ability to capture the dynamic part of alpha that reﬂects Hedge Fund managers’ market timing ability. We found that the quantity of new alpha substantially increased the proportion of positive and negative alpha funds. Nevertheless, we showed that the majority of Hedge Funds are zeroalpha funds as Barras, Scaillet and Wermers (2008) portrayed for Mutual Funds. Furthermore, we look into the possibility for a strategy to have a common increasing trend to a market exposure during a crisis. For all strategies, the main result was a common trend to an increase in the credit spread risk factor.
This timevarying coeﬃcient model has been fully explored by the seminal work of Cleveland, Grosse and Shyu (1991). 2 positive returns in all market conditions, up or down. 3 February, March 2000 4 July, Auguste, and September 1998
1
3
To the best of our knowledge, this is the ﬁrst paper which tries to look into how the risk of Hedge Funds and CTAs exposure changes. This paper also looks into the skill of manager anticipation, part of alpha. The rest of our paper is organized as follows; Section 2 reviews related literature about Hedge Funds modeling, and the dynamic in beta5 . The data is described in section 3. Section 4 summarizes the risk factors deﬁned in Fung and Hsieh’s paper (2001, 2004). Section 5 tests for structural change by applying the method of Bai and Perron (1998). Section 6 deﬁnes our timevarying coeﬃcient model. Our methodology in which we apply our model is deﬁned in section 7. Section 8 provides results of our timevarying coeﬃcient model, as well as, the application of the False discovery Rate (FDR hereafter) to alpha and beta bringing, in this way, a new tool for Hedge Fund analysis. Section 9 presents our conclusion.
2
Literature review
It is wellknown that there are some drawbacks to model Hedge Funds. Various fundamental and statistical multifactor models for Hedge Funds have been analyzed by Agarwal and Naik (2000), Edwards and Caglayan (2001), Fung and Hsieh (1997, 2001, 2004a), Lhabitant (2001), Liang (2001), Schneeweis and Spurgin (1998), among others. Research has speciﬁcally been focused on identifying dynamic trading strategies in order to mimic the actual trades of Hedge Funds (for example: Fung and Hsieh (1997), Agarwal and Naik (2000)). Contrary to mutual funds, performance analyzes of Hedge Funds are completely diﬀerent. One reason for this, concerns the returns of Hedge Funds which usually have low correlations with market indices, and therefore a traditional CAPM analysis using the Jensen measure is typically nonappropriate. Another reason concerns the linear regression which works well for mutual funds because its strategies tend to follow a buyandhold pattern. Brown, Goetzmann and Ibbotson (1999) have shown that Hedge Funds pursue many diﬀerent styles which are completely diﬀerent to the buy and hold strategy followed by Mutual Funds and therefore can put the hypothesis of linearity to question. Liang (2001) also documents that Hedge Fund investment strategies are dramatically diﬀerent from those of Mutual Funds. To the best of our knowledge, the ﬁrst paper which suggested another approach in order to better analyze the returns of Hedge Funds was written by Fung and Hsieh (1997). They attempted
5
the readers can ﬁnd a detailed literature review into the book from Agarwal and Naik (2005).
4
Brown and al. and found that a put or a call option on an underlying variable is the most signiﬁcant factor in the case of 54 percent of their funds. commodities. Agarwal and Naik (2000) extended this analysis of Hedge Fund performance by recognizing that funds may follow dynamic nonlinear trading strategies. Fung and Hsieh found that in most cases their option strategies only played a marginal role. systems/trend following and distressed) explain about 45 percent of the crosssectional variation in Hedge Fund returns.67 for Hedge Funds that pursue primarily a long equity investment strategy. Also. value. In contrary to Agarwal and Naik’s (2000) discoverers. (2001). bonds. but they focused on exploring the risk of ﬁxed income Hedge Fund styles and did not consider performance explicitly. The paper written by Mitchell and Pulvino (2001) on mergerarbitrage strategies are also begun to produce useful explicit links between HedgeFund strategies and observable asset returns. and currency returns. Mitchell and Pulvino (2001)). Following Fung and Hsieh (1997). In order to ameliorate the explanation of Hedge Fund returns by asset class. They used stepwise regression to identify the independent variables. As we will be using their methodology in this paper we will introduce and develop it further later on in our work. which concludes that Hedge Fund return patterns do not map as well to the ﬁnancial asset as do mutual fund returns. One reason given by the authors for the varied results mentioned above is due to the fact that Fung and Hsieh (2002) used active and advanced straddle strategies.2000. They found that ﬁve different trading style (system/opportunistic. They found that Rsquare measures that range from near zero for relative value Hedge Funds to 0. they found that the amount of variation in Hedge Fund returns that is explained by asset class return is low 6 . Fung and Hsieh (1997) applied a factor analysis based on a Hedge Fund’s trading style. Mitchell and Pulvino (2001) and Agarwal and Naik (2004) show that Hedge Fund returns relate to conventional 6 RSquare measures were less than twenty ﬁve percent for almost half of the Hedge Fund studied 5 . Fung and Hsieh (2002) incorporated option strategies into a Sharpe style model. In summary.2001). global/macro. In their paper.to analyze the returns to Hedge Funds by applying the factor or style analysis conducted by William Sharpe with respect to Mutual funds. Schneeweis and Spurgin (1998) conﬁrm this result by conducting a regression analysis of the returns of stocks. a lot of articles have been developed which seek to understand the trading strategies and characteristics of Hedge Funds by regressing their returns on explanatory factors (Agarwal and Naik (2000). Mitchell and Pulvino (2001) simulated the returns of a merger arbitrage strategy applied to announced takeover transactions from 1968 until 1998. Fung and Hsieh (1999.
Fung and Hsieh. The latter.asset class returns and optionbased strategy returns. Indeed they found that their CUSUMs’ tests of the stability of factor loadings rejected the null hypothesis of stable coeﬃcients at the 5 percent level in the case of three quarters of the funds. Lee. Lo. They study vendorprovided fundoffund indices. The primer methodology was developed by Andrews. They found that the break points coincide with extreme market events7 . 7 6 . 2001). With strong evidence of non normality (Agarwal and Naik. Their interest is in the broader issue of how well data on Fund returns can be used to measure “changes”in factor exposure. and performed a modiﬁedCUSUM test to ﬁnd structural break points in fund factor loadings. the paper from Bollen and Whaley8 studied 2 econometric techniques that accommodate changes in risk exposure. the collapse of LongTerm Capital Management in September 1998. Fung et al. and Ploberger (1996). 2001. the mean and standard deviations are not suﬃcient to describe the return distribution.(2006) paper was concerned with estimating factor exposures at the time of particular crises. the returns also exhibit third and fourth moment attributes10 as well as positive ﬁrstorder serial correlation11 . Amin and Kat. 2006). Brooks and Kat (2002) show that although Hedge Funds oﬀer high mean returns and low standard deviations. Recently. Nevertheless the 2 precedent methodologies use a hypothesis of normality which is not adapted to the world of Hedge Funds. 10 Skewness and kurtosis. This problem is less acute for Mutual Funds than for HedgeFunds. 1999. and higher moments need to be considered. uses maximum likelihood and a Kalman ﬁlter under an AR(1) model. (2004)) that monthly Hedge Fund returns may exhibit high levels of autocorrelation. They studied a class of optimal tests9 for multiple changes. and the peak of the technology bubble in March 2000 8 At the moment where we write this article. we do not know when this paper will be publish in Journal of Finance. 11 They showed (as Lo & al. They relate another strong diﬀerence between Mutual Funds and Hedge Funds which concerns the problem of managers changing their investment style over time. Kat and Lu (2002). They found signiﬁcant changes in the risk factor parameters in about 40 percent of their sample of Hedge Funds. Brealey and Kaplanis (2001) presented evidence that within each category funds tend to make similar changes to their factor exposures. Furthermore these results about the distribution of Hedge Funds could be diﬀerent depending on their strategies (Anson. namely the AvgW and ExpW. 9 In the sense that they maximize a weighted average power. 2003. In a similar way.
It covers the period from May 1975 through to October 2008 but we are only concerned with the period from January 1991 until October 2008. and 600 CTAs. we use the Center for International Securities and Derivatives Markets (CISDM) and HedgeFund. The sample with all funds contains approximately 9800 funds (Hedge Funds. 7 . Fung and Hsieh (1997) and Brown and Goetzmann (2003) have identiﬁed between ﬁve and eight investment styles. CTA and Fund of Funds). Veeraraghavan. {please insert Table I here} HedgeFund. 12 13 This database combines four main group. whereas. Funds of Funds. one for Hedge Funds and the other for CTAs. we have collected the returns. It has approximately 3000 Funds of Funds. We select funds for a minimum of 30 months covering the period May 1998 until June 2000.Net database. Returns are net of management and performance based fees. especially because we are not persuaded by the precedent results.13 . Nevertheless. The primer covers the period from January 1994 through to July 2007 with the advantage of the inclusion of dead funds. The latter is the largest commercial database of active Hedge Fund. and Whelan (2005) have found the presence of only three diﬀerent Hedge Fund styles. Bianchi Drew. For every fund. Fund of Fund and CTA products with over 8500. CTA. Hedge Funds. and CPO. it will be the topic we are going to turn to next. it seems better for this study to follow the 23 strategies deﬁned by the provider plus the CTA as another strategy as well as Fund of Funds. One of the main studies in this article is to analyze whether or not Hedge Funds and CTA were able to generate alpha during market events or on the contrary. Hedge Fund database providers classify Hedge Funds into between 11 and 31 investment styles. and approximately 3000 live funds. Therefore choosing a number of strategies is not an easy task.3 Database For this study. generate negative alpha. whether they generated negative alpha. There is a vast selection of academic literature on how to classify Hedge Funds. We have created 2 groups in order to analyze them separately. 4900 Hedge Funds. In order not to create a bias we have included all funds in our analysis even though dead funds mean that they did not necessarily generate alpha or worst. Thus. it is more pertinent to study Hedge Funds and CTAs depending on their strategies. the strategy and fund type12 . In contrast.net uses 31 strategies that we have grouped in order to obtain the same 23 strategies from the CISDM.
{please insert Table II here} The main analyzes consider the merge database14 due to the speciﬁcity of these products that they tend to avoid direct regulation (by the SEC or any regulatory authorities). 2001. each strategy can also have a speciﬁc exposure to another risk factors and therefore risk exposure can change drastically in regards to the strategy. and the diagnostic identiﬁes which variable are involved in each linear dependency. knowing the right number of factors is major in order to capture risk correctly. 4 Factors Hedge Funds can be divided into four main groups. it is not the only problem that we have with factors. market directional. we can have a problem of multicollinearity. Short selling and equity market timing. If there are too many factor. the most accomplished article about Hedge Fund factors was written by Fung and Hsieh (2004). The exposure to the stock market is the major risk aﬀecting market directional funds 15 . Consequently. We can add to each of them a major risk factor. 17 failure of the proposed transaction. the model is misspeciﬁed and we can question whether alpha corresponds to the manager skill. Indeed. convergence trading fund and opportunistic funds. they managed to obtain similar results using the Agarwal and Naik (2004) optionbased factor We merged the HedgeFund. and relative value arbitrage. statistical arbitrage. The major risk aﬀecting corporate restructuring funds 16 is exposure to the event risk17 which is the same for the convergence trading fund 18 . equity market neutral. Nevertheless. equity Long/Short. deﬁning factors is not an easy task. 16 distressed securities. Into every group. 15 14 8 . Precedent articles have shown that the relation between Hedge Fund returns and market returns are nonlinear. 2004). and Welsh (1980). To the best of our knowledge. convertible bond arbitrage. Mitchell and Pulvino (2001) and Agarwal and Naik (2004)). Kuh. Moreover. merger arbitrage and event driven.Net database to the CISDM and found that they are 925 funds in common. We do not detect any multicollinearity with these eight factors. 19 We use the diagnostic technique presented in chap 3 of Regression Diagnostic by Belsley. The diagnostic is capable of determining the number of near linear dependencies in a given data matrix X. 2000. Moreover. If some risk factors are missing. corporate restructuring fund. 18 ﬁxed income arbitrage. Therefore many researchers have suggested factors in order to explain Hedge Fund returns (Fung and Hsieh (1999. Hedge Funds have a systematic risk but it is not possible to capture this risk with standard asset benchmarks. They showed that their seven factor model strongly explains variation in Hedge Fund returns and at the same time avoid multicollinearity19 .
The problem of critical values was solved simultaneously in the early 1990s by several sets of authors.edu/ dah7/. the best known solution was provided by Andrews (1993). • 2 Equityoriented Risk Factors: S&P500 minus risk free rate23 and Size Spread Factors deﬁned by the Russell 2000 index monthly total return less S&P500 monthly total return. the MSCI Emerging market minus the risk free rate. Their paper included seven factors and they added an eighth factor on their website20 that we have also added. and a Credit Spread Factor formed by the monthly change in the Moody’s Baa yield less 10year treasury constant maturity yield. The problem with these tests in the case of multiple structural change is practical implementation. Currency and Commodity22 which capture a non linear exposure. 22 These factors are downloadable on http://faculty. For more details about the construction of these factors see Fung and Hsieh. in this case. Therefore we will follow the eight Hedge Fund risk factors deﬁned in Fung and Hsieh’s paper (2004)21 . According to Perron. the classical test for structural change is typically attributed to Chow (1960) with the weakness that the breakdate must be known ` priori.duke. 2004a. and Andrews and Plobergers (1994). Bolley and Whaley implemented this test using just one changepoint. Nevertheless. However.edu/ dah7/DataLibrary/TFFAC. 5 Structural Change Few researchers have tested if there has been a structural change.fuqua. 2001. the AvgW and ExpW 20 21 http://faculty. The tests used considered only one structural change and most of them needed to know when the break point should happened. Lee and Ploberger (1996) developed the precedent results into a class of optimal tests for linear models with known variance.model. 23 3month USD LIBOR 9 . Andrews. Quandt (1960) treated a the breakdate as unknown but.xls . This class of optimal tests allows an arbitrary number of changepoints.duke.fuqua. These factors are : Three TrendFollowing Factors: Bond. 1997. the chisquare critical values are inappropriate. • One Emerging Market Risk Factor. • 2 Bondoriented Risk Factors: a Bond Market Factor represented by the monthly change in the 10year treasury constant maturity yield. Indeed.
This code is a companion to the paper: Estimating and testing structural changes in multivariate regressions (Econometrica. The latter is a test of the null hypothesis of l breaks against the alternative that an additional break exists. Therefore another interesting view is to look at the frequency of break’s dates in the Hedge Fund’s universe by considering the number of the breaks at time t.edu/perron/. 62 percent present some structural changes. relative to one fund. . the Russian government defaulted on the payment of its outstanding bonds. Whatever the strategy. Although Bai and Perron (1998. It is a generalization of the sup F test considered by Andrews (1993). in our opinion. we have applied just one of them which is.tests require the computation of the Wtest over all permissible partitions of the sample. the most signiﬁcant for this analysis. the UDmax and the WDmax which diﬀer by their weight methodology26 . Bai and Perron’s tests also give the break’s date. 2142 Hedge Funds). we have noted an increase in the ratio over the period 20022007 as well as an increase in the frequency of breaks. The results are conclusive. In this category. It stops when the sup F (l + 1l) test is not signiﬁcant.. This is the second test provided by Bai and Perron. All these precedent results took into consideration track records with more than 30 months.bu. 2004) solved this problem with a very eﬃcient algorithm which is available on their website24 . The Russian debt http://people. 24 10 . we have suggested to create a ratio called Rbreaks . This default caused a worldwide liquidity crisis with credit spreads expanding rapidly all around the globe. which is deﬁned as the number of breaks at time t divided by the number of funds at time t. By strategy. This sequential procedure estimates each break one at a time. 5 breaks. It is a test of no structural breaks against an unknown number of breaks given some upper bound (M = 5 in our application).III. There are two interesting things to say about the two crisis. the increase in the amount of Hedge Funds is considerable and we must take this huge growth into consideration within our analysis. called Double Maximum Test. there are two tests.II. {please insert Table III here} The majority of Hedge Funds present structural breaks... the results are conclusive. the minimum percentage of Hedge Funds with breaks is 31 percent (Other Relative Value) and the maximum is 70 percent for Event Driven Multi Strategy. {please insert Graph I. In August 1998. 2004) proposed three diﬀerent tests25 .IV here} Once again. 25 The primer test considers no break versus ﬁxed number of breaks (up to 5 breaks)This test considers the sup F type test of no structural break (m = 0) versus the alternative hypothesis that there are m = 1. 2007) (developed by Zhongjun Qu). For this reason. If we consider the most representative strategy (Equity Long/Short. Bai and Perron (1998. 26 See Bai and Perron (1998) for a full explanation on the diﬀerent weights. Indeed.
6 Our Model Stone (1977) introduced local linear least squares kernel estimators as a regression estimator which has been generalized by Cleveland (1979)27 . showing fewer structural changes. During this period a bubble developed. 11 . This section has shown that it is not enough to purely take into consideration diﬀerent distributions but also that alpha and beta could be dynamic and consequently depend on time. Kernel estimators u 27 The robust local regression estimators. The conditions for ﬁnancial markets in 1999 were very good. better suited for Hedge Funds. They concluded that Hedge Funds deliberately held technology stocks and were able to exploit this opportunity. We suggest a timevarying coeﬃcient model with the Fung and Hsieh factors which are. Stone (1980. we noticed the opposite for the equity bubble crisis which corroborated the conclusions of Brunnermeier and Nagel (2002). Furthermore. 1993) showed in the univariate case that another important advantage of local linear least squares kernel estimators is that the asymptotic bias and variance expressions are particularly interesting and appear to be superior to those of the NadarayaWatson or GasserM¨ller kernel estimators. The next section presents an answer to the problem of dynamic for alpha and beta. in this way.crisis (LTCM) was a crisis that materially aﬀected Hedge Fund returns that was conﬁrmed by the results from Rbreaks . 1982) used local linear least squares kernel and its generalization to higherorder polynomials to show the achievement of his bounds on rates of convergence of estimators of a function m and its derivatives. the precedent results for a majority of strategies were conﬁrmed. According to Brunnermeier and Nagel (2002) most Hedge Funds. decided to ride the bubble rather than clear their positions. These arguments are conﬁrmed by the fact that Rbreaks is very low. They explained that Hedge Funds heavily tilted their portfolios towards technology stocks without oﬀsetting this long exposure by short or derivatives. we showed that the risk in Hedge Funds increased over the last few years essentially due to the use of leverage in order to reach the investors desired return target. despite irrational levels of valuation. Surprisingly. Furthermore. especially for riskier assets. Fan (1992. Furthermore. for the majority of the strategies.
Thus βi (t) depend on t. for this study we need an automatic bandwidth selection. Coling and Chiang (2000) recommend another alternative bootstrap procedure suggested by Hoover et al. provided the density f of Xi s satisﬁes certain assumptions. Kauermann. See appendix for more details on the TimeVarying Coeﬃcient Model. both bootstrap procedures may lead to good approximations of the actual (1 − α) conﬁdence intervals when the biases of βi (t) are negligible31 . The varying coeﬃcient model assumes the following conditional linear structure: p Yt = j=1 βj (t)Xjt + εt = α(t) + Xβ(t) + εt for a given covariates (t. To practice statistical inferences such as the construction of conﬁdence interval for βi (t) different methods have been suggested. (1998). and they are consistent for any smooth m. this hypothesis can signiﬁcantly reduce the modeling bias and especially avoid the “curse”of dimensionality28 . In practice. Another paper of Galindo. We illustrated the performance of our estimator by a simulation study. According to the authors. . It is wellknown in kernel regression that the selection of bandwidths is more important than the selection of kernel function. X1 . Xp ) and variable Y . 32 Appendix I gives a summary to the algorithm. and Carroll (2000) suggest another bootstrap method based on the wildbootstrap of H¨rdle and Marron (1991) a 30 Construct pointwise intervals of the form 29 28 βi (t) ± z(1−α/2) se∗ (t). 31 They point out that theoretical properties of these bootstrap procedures have not yet been developed. We used the bandwidth deﬁned above and a number of data equal to 50 in order to respect Kernel depends. Colin... 12 . only on t. in this context.have the advantage of being simple to understand and globally used by researchers. Coling and Chiang (2000)29 suggest a naive bootstrap procedure that we have applied in this article. Wu and Chiang (2000) suggest that we apply the choice of bandwidth “leaveonesubjectout”crossvalidation32 . The main advantage of this naive bootstrap procedure is that it does not rely on the asymptotic distributions of βi (t). The mathematical analyze and the implement into a computer is easy. Moreover.. B where se∗ (t) is the estimated standard error of βi (t) from the B bootstrap estimators and z(1−α/2) is the (1 − α/2th ) B percentile of the standard Gaussian distribution. bandwidth may be selected by examining the plots of the ﬁtted curves. which relies on normal approximations of the critical values30 .
1−α End of month of July. In each group obtained. March and April 200035 . we were able to determine whether or not managers were able to react quickly and also if we can capture change exposure superior to 10%. we have built 2 indicators for α36 and 2 indicators for βi .. Brooks and Kat (2002) show that the netoffees monthly returns of the average individual Hedge Funds exhibit positive ﬁrstorder serial correlation which is due.β = βi /σβi for every αj (t) and βij (t). robust ability). some others indicators could be more appropriate for a speciﬁc analysis of these market events. assure that the linear relation follows also a normal distribution. August. Using these 3 months’ data. The second aspect treats the ability to anticipate market events or to cope with them (i. i = 1.37 The 2 indicators for beta38 have the same diﬀerences but are augmented by 10 percent39 .Nj . The second has the ﬁrst requirement (more than 30 months) but only has funds which were available during the 2 crisis. 8). 8. By doing this. The ﬁrst is composed by all tracks record which is superior to 30 months. February.. 13 . The observed (or smoothed) value Vt∗ of a Hedge Fund at time t could be expressed as a weighted average of the true ∗ value at time t..1) × βt−1 39 This methodology is simply a linear relation between 2 independent variables which. Vt−1 : ∗ Vt∗ = αVt + (1 − α)Vt−1 . we have used the tstatistic. rt = 35 36 ∗ ∗ rt − αrt−1 . . For every indicator 33 Kat and Lu (2002). March. . Auguste. .. For the LTCM period we will focus on July. In each case. (i = 1.the average size of Hedge Fund tracks (see appendix). we have selected 3 consecutive months. we have formed 2 groups. 38 ˆ ˆ βt − (1. and September and end of month of February. Vt and the smoothed value at time t − 1. and September 1998. ti. To properly cover these periods. {please insert Table I here} Indeed.e. we are concerned with the 2 diﬀerent aspects. In order to capture anticipation in alpha. according to the authors. and April. We have removed serial correlation by applying the same methodology as used in Brooks and Kat’s paper (2001). Nevertheless.. We regress the netoffee monthly excess return (in excess of the riskfree rate) of a Hedge Fund on the excess returns earned by traditional buy and hold and primitive trend following strategies deﬁned above33 . called the simple Blundellward ﬁlter34 . The ﬁrst focusses on the security selection ability (α(t)).. We have grouped the estimates together into the 23 strategies deﬁned by the CISDM. we are going to look at 2 strong market events: the LTCM crisis and the Equity Bubble crisis. and for the Equity Bubble period. αt − αt−1 ˆ ˆ 37 These two indicators were created in order to show a new methodology that this model bring to the analysis of Hedge Funds.. Nj being the number of funds.α = αi /σαi and ti. under the condition of ˆ normality for βjt j = 1. to markingtomarket problems.. The 2 primers are the diﬀerence between the second month and the ﬁrst month and the diﬀerence between the last month and the second month.
Now. They analyzed whether or not Hedge Fund performance can be explained by luck and if Hedge Fund performance persists at annual horizons. Therefore we have used their methodology called False Discovery Rate (FDR hereafter) in our multiplehypothesis test. On the one hand. and Wermers (2008) suggested another approach which is notably informative with regards to the prevalence of outstanding managers in the whole fund population. In order to overcome the nonnormality. Scaillet. the test can accept H0 whereas it is H1 which is true. a multiplehypothesis testing.e. we were able ˆ ˆ to determine.D. Nevertheless. Once it was done.e. Either we reject H0 when X ∈ Γ or we accept H0 when X ∈ Γ. using a bootstrap procedure. In this way. Their approach simultaneously estimated the prevalence and location of multiple outperforming fund within a group. from a more general perspective. Barras. for each strategy. Their methodology tested the skills of a single fund that was chosen from the universe of alpharanked funds. there is the possibility that there is an error in the test. the test can reject H0 whereas H0 is true. we have applied the FDR on α (security selection ability) in order to determine the proportion of zeroalpha funds and skilled and unskilled funds.40 If we examine just 1 fund. examining fund performance. we only have to test a null hypothesis H0 versus an alternative H1 based on a statistic (let’s say X). A ﬁrst approach was suggested by Kosowski. and Wermers (2008). in fact.” of the possibility change in market exposure. we applied the FDR on the “beta” indicators deﬁned above during the 2 crisis in order to determine if we could observe a common increasing trend for change in market exposure i. On the other hand. The tstatistic distributions for individual Hedge Funds are generally nonnormal. a common increase in beta value of more than 10%. we use the same approach as Barras. This error is called a type II error. consisting of the use of a bootstrap to more accurately estimate the distribution of tstatistics for each Hedge Funds (and their associated pvalues). Naik and Teo (2005). in this manner. Scaillet. we calculated the pvalue for every indicator and the precedent means. the proportion of the security selection ability as well as the proportion of the timing ability.covering the 2 periods i. but also for the indicator deﬁned above αt − αt−1 (timing ability) for the 2 crisis. for each strategy. if we want to test for numerous Hedge Funds this problem becomes much more complicated. called type I error. Moreover. Step 2. We have 2 possibilities for a given rejection region Γ. Step 1. 4 indicators by fund for αj (t) and 32 indicators by fund for βij (t). 40 14 . we calculated the median of the percentage change for each beta of each strategy in order to give an “I. Testing numerous Hedge Funds is to do.
On the other hand. {please insert Table V here} Concerning the risk behavior. a view about common trend change and secondly about what are the highest impacted beta.6 percent of Mutual Funds generate positive alpha and a majority of them can also be considered as zeroalpha funds. Although The CTA coped 41 42 refer to the time between making an investment and needing the funds. Therefore Mutual Funds are often associated with Funds that have a relative performance. Barras. 15 . Whereas the TVCM roughly gave nineteen and twenty eight percent. So the legitimate question is: Are the results relatively the same for Hedge Funds or are they completely diﬀerent? On the one hand. we brought 2 diﬀerent views about exposure change. Mutual Fund managers use a buy and hold strategy consisting of buying a range of ﬁnancial products according to their investment strategy and then they hold them according to the time horizon (or investment horizon)41 . if we apply a static linear factor model such as the NeweyWest (1987) heteroscedasticity and autocorrelation consistent estimator. we were able to capture other skills from Hedge Fund managers. Firstly. we showed that the majority of Hedge Funds have a tendency to get an increase in credit spread as well as bond market risk factors. resulting in obtaining us a non negligible increase of positive alpha funds. whatever the strategies. Results for the CTA We observed a strong diﬀerence for estimated alpha between the static factors model (SFM hereafter) and our timevarying coeﬃcient model (TVCM hereafter). The results and the graphs for the seventeen remaining strategies are available upon request. The estimated alpha from the two diﬀerent crisis are interesting to point out. we determine a “static”alpha that does not capture the particularities of Hedge Fund Strategies. Scaillet and Wermers (2008) have shown that only 0. Unsurprisingly. the majority of Hedge Funds are zeroalpha funds. We are going to look into the results for seven out of the twenty four strategies in the following part of this paper42 . And therefore we have showed that. 7 Results Generally speaking. when we applied our timevarying coeﬃcient model.With these results. The SFM gave a very small percentage of positive and negative alpha funds with two percent and one percent respectively.
The majority of CTAs had a relatively stable exposure. The credit spread risk factor was the most sensitive factor during LTCM. We obtained the same results using the SFM. During the Summer of 1998. of positive alpha funds. The proportion of dynamic skilled funds was very good during the two crisis with a huge number of roughly forty percent. during the two events. whereas. {please insert Graph XI here} Emerging Markets The emerging market strategy showed a good proportion of stockpicker skilled funds where approximately twelve percent generated positivealpha which pointed out that the majority of managers were fundamental bottomup stockpickers. four out of the eight factors showed a sensitivity during the Equity Bubble crisis. Nevertheless this sensitivity concerned only a small percentage of our population. The SFM gave a small four percent and three percent of negative alpha funds. and. It obtained a better percentage with twenty four percent using the TVCM. the CTA had a tendency to show a slight increase in the credit spread risk factor and the emerging market risk factor.5 percent in September according to CSFB/Tremont Managed Futures 16 . Unsurprisingly. Generally speaking.during the two crisis it obtained a strong percentage of positive alpha funds during the Equity Bubble Crisis where approximately twenty seven percent were positive alpha funds. the 43 Approximately 10 percent in August and 7. unsurprisingly. Certain Equity Long/Short specialize in a speciﬁc sector like technology. During the Equity Bubble and LTCM we noticed a greater dynamic strategy than previously seen where approximately twentyﬁve and ﬁfty percent of our population showed an increase in two risk factors: the credit spread risk factor and the bond market risk factor. It indicated that the estimated alpha was more “static”than with other strategies. {please insert Graph V here} Equity Long/Short Equity Long/Short obtained approximately the same results as the CTA apart from the Equity Bubble crisis. the timer ability had been more impact during the Equity Bubble than during LTCM. CTAs had one of its best performances43 while all other Hedge Fund strategies were struggling. These results conﬁrmed that emerging market equity hedge fund managers perceived the high volatility of emerging markets as an asset.
A small percentage of the population showed an increase or a decrease in exposure during the two crisis.proportion stayed relatively consistent pointingout their ability to switch from the short to the long position and visa versa. Therefore the two results were relatively closer than with the other strategies. and only the second crisis was more proﬁtable. This was surprising because the two crisis created several opportunities44 . spinoﬀs. 17 . for a minority of the population we noticed that the credit spread and the commodity showed the biggest sensitivity during LTCM and the emerging market factor during the Equity Bubble. On the other hand it was the strategy which obtained the smallest proportion of unskilled funds. We noticed a tiny four percent during the ﬁrst period which reached a strong sixteen percent in the second period. concerned the percentage of the population to show a variation in factor exposure. {please insert Graph VIII here} Another strength. We still noticed a sensitivity to the credit spread risk factor and to the commodities factors during LTCM. for this strategy. and other announced events. {please insert Graph VI here} Equity Market Neutral This strategy produced a very interesting result where the proportion of stockpicker skilled funds was three percent higher than the estimated proportion using the SFM. The result was conﬁrmed during the two crisis with zero percent during LTCM. {please insert Graph VII here} It was the most robust strategy in terms of change in exposure. Furthermore this result was conﬁrmed by the obtained percentage of estimated alpha during LTCM with zero percent. The 44 Invests in mergers. Whereas emerging risk factor was the most sensitive during the Equity Bubble crisis. Therefore we can deﬁne it as a zeroalpha fund strategy. Nevertheless. A negligible part showed an instability during the crisis. reorganizations. It didn’t cope as well during the equity bubble. Event Driven Multi Strategy Event Driven Multi Strategy obtained the worst percentage of skilledfunds with zero percent for the SFM and a small two point ﬁve percent for the TVCM.
These results conﬁrmed that global macro managers have the most extensive investment universe and that they were able to ﬁnd some arbitrages. The Equity Bubble crisis also gave a good percentage of positive alpha funds with eighteen percent. It obtained one of the bigger groups of unskilled funds with roughly twenty eight percent. this strong percentage was less than the seventy one percent of positive alpha funds found during LTCM. Credit spread was still present during LTCM while Sizespread risk factor had an non negligible sensitivity to bond. Furthermore. Whereas. The SFM gave ﬁve percent of skilled funds whereas the TVCM gave thirty ﬁve percent. the emerging market risk factor and the credit spread risk factor (slightly) were concerned about the change in exposure. so the result. Moreover. The Equity Bubble obtained a tiny twenty seven percent: This impressive result conﬁrmed the strong dynamic within the strategy.factors concerned were the credit spread the emerging market factor and the commodity factor for LTCM and the emerging risk factor for the equity bubble. commodity and emerging market risk factors during the Equity Bubble. the percentage of positive alpha during LTCM increased up to twenty six percent. More than sixty percent of our population showed an increase in diﬀerent market risk factors. in our opinion. it is important to say that the number of Hedge Funds following this strategy was relatively small. Unsurprisingly. could be debatable. {please insert Graph X here} 18 . {please insert Graph IX here} Short Bias Step one. the short bias produced the best percentage of variation exposure in. Credit Spread risk factor stayed the most sensitive during LTCM. Less than ten percent of our population showed an increase or decrease in our general exposure. the size spread factor. Global Macro Global Macro showed a proportion of stockpicker skilled funds equal to ﬁve percent using TVCM and one percent using the SFM.
Our result are less accurate than Zhang. We retested the ability to capture a structural break (see appendix I). We allowed for three diﬀerent sizes: ﬁfty. For these reasons the diﬀerent graphics truly represent the obtained result for each fund. one hundred and one hundred and ﬁfty months which akin the size of Hedge Funds that we can ﬁnd in diﬀerent databases. Because of the size of our database. the use of a second bootstrap algorithm for each fund (the ﬁrst algorithm calculates the pvalue) would have made the estimation process too costly. And therefore these diﬀerences slightly alter the performance of our estimator. even in the short sample. we applied the false discovery rate to the estimated alpha and betas obtained by Bspline. we managed to obtain some very good results. Thirdly. In a second step. we did not apply a monte carlo study in order to ﬁnd the optimal bandwidth (or knot for the second model). we used the S&P500 and the monthly change in the 10year treasury constant maturity yield as factors which are not a perfect factors set. Secondly. 19 . we created a track of return where the two betas had a structural change. Lee and Song (2002). Firstly. The results gave approximately the same percentage of unskilled. The main problem with structural change concerns some technical conditions. Lee and Song due to three reasons. For this.Robustness The goal of this section is to demonstrate that our model is robust to structural change and that we succeeded to capture a jump in estimated alpha or beta. our sample contains between 50 and 150 values in opposite to the range in 2501000. zero and skilled funds for our population. Furthermore. We used one of the same example as in Zhang. Nevertheless in second step to our estimator we used a Bspline smoothing instead of a local regression. Mainly that the second derivative is required to be continuous in a neighborhood of t. The results were also very good. As showed in appendix I. in order to show that our results were independent to our model we created another timevarying coeﬃcient model which was also based on the work of Fan and Zhang (1999).
a proportion of negative alpha funds. the use of leverage and their structural fees. The use of the factors from Fung and Hsieh allowed us to capture the nonlinearity in beta and therefore gave the best overall risk factors. We found a diﬀerent result for Hedge Funds. Another advantage of this model was its capacity to analyze change in factorial exposure. If we want to model Hedge Funds. we applied the FDR to determine the 20 . we examined the percentage change between our eight factors and saw if there was a persistent for beta parameters. Scaillet and Wermers (2008) showed us that only 0. We showed that this dynamic was not wellcaptured by a static factor model where the results considered Hedge Funds as zeroalpha funds. 2004). the dynamic in beta as well as non linearity exposure to the market. Hedge Fund managers seek absolute returns and try to outperform the market whatever the market conditions are. No assumption about statistical distribution. We saw three main results. Whereas our model found a better proportion of positive alpha funds but also. Whereas Barras. that a majority of Hedge Funds with a track record superior to 30 months had a minimum of one structural break. In order to overcome these drawbacks. The model allowed us to postulate that alpha and beta are a function which depended on time and avoided parametric distribution. Last but not least. Firstly. Secondly. This model allowed us to separate manager skill into two components pointedout by the (stock or bond or funds)picking and the ability to anticipate market events. Hedge Fund managers try to focus on positive return (whatever the market conditions). deﬁned in Fung and Hsieh’s paper (1997. Furthermore we noticed that the frequency of breaks has increased over the last few years. It also allowed us to see what were the changes in beta exposure or the managers’ reactions according to the market’s conditions.8 Conclusion Hedge Funds cover many diﬀerent strategies which radically vary in terms of market exposure and risk. However there are some common characteristics. unfortunately. by strategy. Our results suggest that the common increasing percentage change of diﬀerent Hedge Funds in the downstate of the market is strongly represented by one exposure. these characteristics imply a model which takes some new assumptions into consideration.6 percent of Mutual Funds generated positive alpha and therefore a majority of them can be considered as zeroalpha funds. 2001. we used a timevarying coeﬃcient model as well as the whole factors. it was the credit spread risk factor.
proportion of the fund’s population which had an increase. a decrease. The changes in factorial exposure and the proportion of funds give us a strong tool for risk managers and speciﬁcally for stresstesting. 21 . or stayed constant in the market’s exposure during the two crisis (LTCM and the Equity Bubble). We show that within a strategy. Hedge funds have a tendency to have an increase in market exposure.
. and Naik N. Daniel N. 2000. [6] Agarwal V. Do Hedge Funds Manage their Reported Returns. [19] Bailey W. [7] Agarwal V. Portfolio of Hedge Funds. [4] Agarwal V. [3] Agarwal V.... 2002. Hedge Fund Performance Evaluation: A Stochastic Discount Approach.. and Huij J. P. Working Paper.. Working paper. Y.. and Naik N. 54:833–874.. 2000. The journal of ﬁnance. and Kat H. [14] Amin G. Hedgequest.F. 2005. November 2008. Leelahanon S. Working Paper. January 4. Working Paper. [2] Agarwal V... Working paper. 2007. Performance Evaluation of Hedge Funds with Optionbased and BuyandHold Strategies. [5] Agarwal V. M. 2004. [10] Agarwal V. [8] Agarwal V. 22 . Poor Man’s Hedge Funds?.References [1] Ackermann C.. [9] Agarwal V. Performance. S. 2002. and Incentives. and Teo M. D. and Bollen N. 2003. and Fung W. Working Paper. Y.and Loon Y. [12] Amenc N.. Flows. Hedge Fund Performance 19902000. Naik N. Li H.. Working Paper. Risks and Portfolio Decisions Involving Hedge Funds. May 2008. and li Q. and Martellini L. and Zhang X. H. [13] Amenc N. Working Paper. The Journal of Financial and Quantitative Analysis. and Martellini L. March 2002. December 2004. Evidence of Predictability in Hedge Fund Returns and MultiStyle MultiClass Tactical Style Allocation Decisions. and Naik N. M. S. Working paper. Naik N. Portfolio Optimization and Hedge Fund Style Allocation Decisions. Dynamic Investment Opportunities and the CrossSection of HedgeFund Returns: Implications of HigherMoment Risks for Performance. Y. and Daniel N. El Bied S. 1999. The Performance of Hedge Funds: Risk. [15] Amin G. MultiPeriod Performance Persistence Analysis of Hedge Funds. 2005. Working Paper. Risk and Return in Convertible Arbitrage: Evidence from the Convertible Bond Market.. and Naik N. [17] Avramov D. Kosowski R. and Kat H. 1(1):22–26. Role of Managerial Incentives and Discretion in Hedge Fund Performance. The annals of Statistics.and Naik N. Return.. and Ravenscraft D. [18] Bacmann J. Working Parer. 2006. [16] Ang A. Working paper. McEnally R. 2006. Working paper. and Daniel N.. Working Paper. Working Paper. Performance and Risktaking of Hedged Mutual Funds. D. Y. 2002. C. 2002.. Boyson N. [11] Ahmad I.. Locked Up by a Lockup: Valuing Liquidity as a Real Option. M. 33(1):258–283. Y. Y.B. Bakshi G. Y. Eﬃcient Estimation of a Semiparametric Partially Linear Varying Coeﬃcient Model. 2008. Funds of Funds Are Still Providing Alpha. Y. 35:327–342.. investing in Hedge Funds when Returns are Predictable. and Naik N. and Managerial Incentives in Hedge Funds. and Jeanneret A.
NonParametric Analysis of Hedge Fund Returns: New Insights from High Frequency Data. A Quantitative Analysis of Hedge Fund Style and Performance. 4(2):161–187. N.. Careers and Survival: Competition and Risk in the Hedge Fund and CTA Industry. Working paper. 2001. [36] Chan N. J.and Liang B. Dec. 198995. [37] Christiansen C. Working Paper. and Hastie T. 2002. 20(4):1970–1984. estimating Operational Risk for Hedge Funds the ωScore. and ChinTsang C. [33] Carol A. [38] Colin O. [26] Brooks C. Consistency and Limiting Distribution of the Least Squares Estimator of a Treshold Autoregressive Model. 1993. Fees on Fees in Funds of Funds. and Peare P. The annals of Statistics. Getmansky M. and Ibbotson R. 2004. N. Systemic Risk and Hedge Funds. 72:91–117. K. E. 1992. 2008.. Getmansky M. [39] Darolles S. Journal of Portfolio Management. Statistica Sinica.. P.. Kernel Smoothing on Varying Coeﬃcient Models with Longitudinal Dependent Variable. and Schwarz C. [25] Brealey R. estimating Coeﬃcient Distributions in Random Coeﬃcient Regressions. S. G. and Dimitriu A. International Finance. 2002. 23 . Arbitrage at its Limits: Hedge Funds and the Technology Bubble. Journal of Alternative investments. [28] Brown S.. [24] Bollen N. Hedge Funds and Financial Stability: an Analysis of Factor Exposures. 29(2):101–112. [27] brown S. Working Paper.. [34] Chambers J.. 1992. 1999. and Goetzmann W. Working paper. No arbitrage and arbitrage pricing: A new approach. [31] Brown S. J. May 2008. The annals of Statistics. 2008. Hedge Fund Risk Dynamics: Implications for Performance Appraisal. 2004.. B.. [23] Billio M. [32] Brunnermeier M. Working paper. and Kat H... 21(1):520–533. Oﬀshore Hedge Funds: Survival and Performance. Statistical Models in S. The Journal of Finance. Haas S. 48:1231–1262.A. [21] Bartholdy J. [29] Brown S. Paciﬁc Grove. W.and Simon G. 2009. The Journal of Business. 1993. J.and Liang B. and Christensen M. Working paper. and Kaplanis E. Working Paper.and Goetzmann W. J. N. Hegde Funds and Style. Working Paper. B. The Statistical Properties of Hedge Fund Index Returns and their Implications for Investors. and Park J. M. P. Estimation of Expected Return: Capm vs Fama and French.. Goetzmann W. Madsen P. Wadsworth.. [35] Chan K. 2001. 2000.and Nagel S. Working Paper. 2003. and Whaley R. and Pelizzon L. and Florens J. [22] Beran R. M. The journal of Finance. 2005. 2004. and Hall P. J. and Lo A. 56:1869–1886. A Comparison of Cointegration and Tracking Error Models for Mutual Funds and Hedge Funds. Hedge Fund Duration: Endogeneity of Performance and Assets Under Management. 2004. Goetzmann W. Goetzmann W. [30] Brown S.. N. 10:433–456.[20] Bansal R. Working Paper.B. and Viswanathan S. M.
and Hsieh D. [53] Fung W. A. Statist.. A Primer of Hedge Funds. 58(5):16–27. 2003. 10:275–302. A. 21:196–216. and Madan D. [51] Fung W.[40] Darolles S. and Hsieh D. 20:2008–2036. and Galeano J. [47] Fan J. [52] Fung W. A. [41] Davies R. 1999. and Hsieh D. [59] Fung W. and Signer A. A. Proﬁle Likelihood Inferences on Semiparametric VaryingCoeﬃcient Partially Linear Models. J. 24 . and Huang T. Journal of Fixed Income.. A Monte Carlo Method for Optimal Portfolios. Ann. [43] Eberlein E.. Analyst Journal. and Rindisbacher M. Kat H. MeanModiﬁed ValueatRisk Optimization with Hedge Funds. 2002. Working Paper. and Zhang W. 2002. 2005. The Risk in Hedge Fund Strategies: Theory and Evidence from Trend Followers. and Hsieh D. and Gijbels I. Assoc. The Review of Financial Studies.. and Hsieh D. 11(6):1031–1057. Empirical Characteristics of Dynamic Trading Strategies: The case of Hedge Funds. Variable Bandwidth and Local Linear Regression Smoothers. Journal of Investment Management. 2004. A. LPerformance with an Application to Hedge Funds. [55] Fung W. The Risk in Fixed Income Hedge Fund Styles. 2002. 2000. July 2008. and Hsieh D. Hedge Fund Performance: Sources and Measures. [48] Fan J. Spurious Biases. 1997. and Hsieh D. Working paper. and Lu S. Performance Characteristics of Hedge Funds and Commodity Funds: Natural vs. Hedge Fund Benchmarks: A Risk Approach. 62:53–58. [56] Fung W. Extracting Portable Alphas from Equity Long/Short Hedge Funds. The Diﬃculties of Measuring the beneﬁts of Hedge Funds. [50] Favre L. A. Financial Analyst Journal. B. Amer. [54] Fung W. The Journal of Financial and Quantitative Analysis. The journal of Finance. [57] Fung W. 35:291–307.. Local Linear Regression Smoothers and their Minimax Eﬃciencies. 2004. Oct. Fund of Hedge Funds Portfolio Selection: A MultipleObjective Approach. Working Paper.. Economics Letters.A. 14:313–341. pages 65–80.. A. and Jasiak J. Garcia R. The annals of Statistics. The Review of Financial Studies. Journal of alternative investments. and Hsieh D. [44] Fan J. AssetBased Style Factors for Hedge Funds. 1992. A. and Hsieh D. J. [45] Fan J. 27(5):1491. Is MeanVariance Analysis Applicable to Hedge Funds.2005. 1999. Ann.B. [46] Fan J. Bernoulli. 1999. August. Statist. The Journal of Alternative Investments. 87:998–1004. DesignAdaptive Nonparametric Regression. 5:21–25. A. 2008.. 2001. M. 2:1–19. Statistical Estimation in Varying Coeﬃcient Models. [58] Fung W. Gourieroux C. 1992. [42] Detemple J. Statist. Working Paper. 1993. 58:401–446. [49] Favre L. 2002.
2005. andKauermann G.. 2000. Managed Futures and Hedge Funds: A Match Made in Heaven.. Hsieh D. 4:327–345. 55(4):757–796. A Contingent Claim Approach to Performance Evaluation. and Lieberman O. and Jackwerth J. Bootstrap Conﬁdence Intervals for Local Likelihood. 63(322):584–595. 1991. Smoothing in AppraisalBased Returns. 1994. 25 . Jun. Working Paper. Working Paper. [68] Goodworth T.2003. 1968. and Hsieh D. ¨ [69] Hardle W. 2006. [63] Gayer G. [77] Kat H. 1:133–160. M. J.. The European Journal of Finance. [72] Hodder J. 8:325–345. January 2006... [64] Geltner D. A. RuleBased and CaseBased Reasoning in Housing Prices.. C. 1991. working paper. Mar. The annals of Statistics. Integrating Hedge Funds into the Traditional Portfolio. A. Nonparametric Risk Measurement Framework for Hedge Funds and FundsofFunds. Local Estimating Equations and Varying Coeﬃcient Models. Economic Review. Series B(methodological). M. M. sept. 2003. 1993. [74] Jaeger L. Naik N.R. Hedge Funds: Performance. Hedge Funds: An Industry in its Adolescence. Journal of Real Estate Research. [71] Hildreth C. Bootstrap Simultaneous Error Bars for Nonparametric Regression. Factor Modelling and Benchmarking of Hedge Funds: Can Passive Investments in Hedge Fund Strategies Deliver. VaryingCoeﬃcient Models. [61] Fung W. 2005 january. FactorBased. R. P. 13:7:645– 655. Working paper.. [76] Kat H. Lo A. Journal of the american Statistical Association. and Marron J. 1975. Gilboa I. November. [62] Galindo C. and houck J. 1993. 43(2):305–325. [73] Hsiao C. Y.. Replication Evaluation of Fund of Hedge Funds Returns. [78] Kat H. 19:778–796. [65] Geltner D. Working Paper n28. S. [67] Glosten L. and Ramadorai T. and Jones C. Working Paper. journal of Real Estate Finance and Economics. 2005.2002. Some Estimation Methods for a Random Coeﬃcient Model. Journal of the Royal Statistical Society. Fourth Quarter 2006. 2004. Working Paper. An Econometric Model of Serial Correlation and Illiquidity in Hedge Fund Returns. Risk and Capital Formation. Working paper. [75] Kat H. and Carroll R.and Jagannathan R. October 2007. November. [66] Getmansky M..[60] Fung W.. Estimating Market Values from Appraisal Values without Assuming an Eﬃcient Market. and Makarov I. Working paper 0022. Incentive Contracts and Hedge Fund Management. liang H. Working paper. D. Some Estimators for a Linear Model with Random Coeﬃcients. The Dangers of Mechanical Investment DecisionMaking: The Case of Hedge Funds. W. Econometrica. E. [70] Hastie T. and Wagner C.. Journal of Empirical Finance. and Palaro H. and Tibshirani R.
. [89] Linton O. The journal of Finance. and Li S. Proﬁle Likelihood and Conditionally Parametric Models. and Patton A. Statist. M. 1977. and Pulvino T. P. Multivariate Locally Weighted Least Squares Regression. and Spurgin R. Working paper. May 2006. Taking the Sting Out of Hedge Funds. [95] Ruppert D. 22(3):1346–1370. Multifactor Analysis of Hedge Fund and Managed Futures.. M. M. Consistent Nonparametric Regression. W. An Excursion into the Statistical Properties of Hedge Funds. 2002. October. Popova E. Working Paper. and Palaro H.. [98] Stone C. 26 . Working paper. 2005. [83] Kat H. and Miffre J. [85] Kosowski R. 2003. [88] Liang B. [96] Schneeweis T. and Patton A. and Miffre J. [84] Kat H.. [91] McFall Lamm R. Naik N. May 2007.2002. M. [93] Pezier J. [92] Mitchell M. and Teo M. 10:1040–1053. The annals of Statistics. Dec. Y. Who Needs Hedge Funds ? Working Paper 0027. Hedge Funds: Risk and Return. Ann. Working Paper. 20(4):1768–1802. J. Performance Evaluation and Conditioning Information: The Case of Hedge Funds. Morton D. 2004. and H. Working paper. 2001. Working Paper. Do Hedge Fund Deliver Alpha ? A Bayesian and Bootstrap Analysis. Working Paper. Journal of Financial and Quantitative Analysis. Journal of Alternative Investments. Ann. 2002. [81] Kat H.[79] Kat H. [90] Malkiel B. 1992. 5:595–620. 8:1348–1360. Journal of Financial Economics. 1:1–24. Working Paper.. and Yau J. 2006. Wong.. J. Statist. Evaluating Hedge Fund Performance: a Stochastic Dominance Approach. Statist. J. Ann. London school of Economics. J. J. and Saha A. and Lu S. TimeVarying Liquidity in Hedge Fund Returns. [94] Popova I. 2006. Optimal Rates of Convergence for Nonparametric Estimator. 2005. optimal Global Rates of Convergence for Nonparametric Regression. In Search of the Optimal Fund of Hedge Funds. and Wand M. The journal of Alternative Investments. 56:2135–2175. TimeVarying liquidity in Hedge Fund Returns. and White A. Optimal Hedge Fund Allocation with Asymmetric Preferences and Distributions. The Impact of NonNormality Risks and Tactical Trading on Hedge Fund Alphas. [80] Kat H. M. 1980. M. 2000. 1998. Asymmetric Returns and Optimal Hedge Fund Portfolios. P. 35:309–336. Hedge Funds: The Living and the Dead. [99] Stone C. characteristics of Risk and Return in Risk Arbitrage. [100] Stone C. G. [82] Kat H. [86] Li S. [97] Severini Thomas A. The Annals of Statistics. The Relative Merits of Investable Hedge Fund Indices and of Funds of Hedge Funds in Optimal Passsive Portfolios. [87] Li S. October. 1994. Working paper. Return and Risk Characteristics. Working Paper. Working paper. 2007.2002.
Journal of the American Statistical Association. jan. 82:166–188. Lee S. and Mehta J. Estimation of Linear Models with Time and CrossSectionally Varying Coeﬃcients. 2002. 72(360):890– 898. 27 . [102] zhang W.[101] Swamy P. journal of Multivariate Analysis..Y. and Song X. Dec. A.. V. S.. Local Polynomial Fitting in Semivarying Coeﬃcent Model. 1977.
Auguste.Table I See Appendix I for deﬁnitions of fund types. March. Certain of the funds used are considered as dead funds i. and September 1998) and the Equity Bubble years (February.e. and April 2000) Strategies Equity Long/Short Multi Strategy Emerging Markets Sector Equity Market Neutral Event Driven Multi Strategy Global Macro Equity Long Only Single Strategy Fixed Income Distressed Securities Fixed Income Arbitrage Convertible Arbitrage Relative Value Multi Strategy Fixed Income .MBS Option Arbitrage Merger Arbitrage Other relative Value Short bias Regulation D Capital Structure Arbitrage Market Timing Unclassiﬁed CTAs (systematic) CTAs (discretionary) FoHFs (multi strategy) Number of Hedge Funds CISDM 2141 251 445 387 322 224 303 148 114 153 162 190 207 89 74 29 126 16 51 16 21 2 58 1003 283 1837 1915 merge 3519 734 635 630 698 384 534 276 112 326 268 314 271 179 98 126 139 32 74 56 30 3 521 CISDM* 1561 142 331 296 227 168 205 99 50 109 131 143 181 74 59 17 111 7 32 13 13 1 37 759 202 1390 1394 merge* 2525 462 461 468 519 293 377 174 50 260 232 237 227 137 69 76 122 17 56 44 21 1 369 28 .e. The funds used at minimum cover the LTCM and bubble period which represent a track record of a minimum of 30 months. they stopped their activities. The database marked by a asterisk give the number of funds covering the 2 speciﬁc periods i. LTCM period (July.
Table II See Appendix I for deﬁnitions of fund types. CISDM database 23 Strategies Multi Strategy Equity Long/Short Short bias Event Driven Multi Strategy Emerging Markets Merger Arbitrage Fixed Income Equity Market Neutral Global Macro Relative Value Multi Strategy HedgeFund. This table shows how we have grouped the 31 strategies from HedgeFund.MBS Convertible Arbitrage Fixed Income Arbitrage Other relative Value Market Timing Option Arbitrage Regulation D Capital Structure Arbitrage Mortgage Convertible Arbitrage Fixed Income Arbitrage Other Arbitrage Market Timer Option Strategies Regulation D Capital Structure Arbitrage 29 .Net and the 23 strategies from CISDM .Net database 31 Strategies Multi Strategy Statistical Arbitrage Equity Long/Short Short bias Event Driven Emerging Markets Merger (risk) Arbitrage Fixed Income (non arbitrage) Market Neutral Equity Macro Value Small/Micro Cap Finance Sector Technology Sector Energy Sector Healthare Sector Sector Equity Long Only Distressed Securities Single Strategy Unclassiﬁed Long Only Distressed FoF Market Neutral Asset Based Lending country speciﬁc Special situations shortterm trading Fixed Income .
Listed is a test provided by Bai and Perron (1998) to analyze whether Hedge Funds have some structural.Table III: Test of Multiple Structural Changes See Appendix I for deﬁnitions of fund types. Strategy Equity Long/Short Multi Strategy Emerging Markets Sector Equity Market Neutral Event Driven Multi Strategy Global Macro Equity Long Only Single Strategy Fixed Income Distressed Securities Fixed Income Arbitrage Convertible Arbitrage Relative Value Multi Strategy Fixed Income .MBS Option Arbitrage Merger Arbitrage Other relative Value Short bias Regulation D Capital Structure Arbitrage Market Timing Unclassiﬁed Fund of Hedge Funds CTA Systematic CTA Discretionary Double Maximum test UDmax 38% 46% 34% 41% 49% 35% 42% 53% 71% 40% 20% 41% 41% 30% 58% 57% 27% 75% 51% 29% 53% 100% 38% 65% 70% 66% WDmax 38% 46% 34% 41% 49% 35% 42% 53% 71% 40% 20% 41% 41% 30% 58% 57% 27% 75% 51% 29% 53% 100% 38% 66% 70% 66% 30 . We note the percentage of Hedge Funds signiﬁcant to the test after applying the False Discovery Rate methodology . The funds used at minimum cover the LTCM and bubble period which represent a track record a minimum of 30 months. The test considers tests of no structural break against an unknown number of breaks given some upper bound (m = 5).
6% Event Driven M. Long/Short TVCM 18.2% 90.9% 17.0% 31.9% 14.0% 3% 28.2% 17.2% Nwest 1% Short bias TVCM 35.7% 59.9% 22.8% 17.Table V: estimated proportions of zeroalpha.6% 13.5% 17.6% 31 Nwest 12.0% 60.net databases.6% 16.5% Nwest 0% Global Macro TVCM 5.2% 0. Scaillet and Wermers (2008). This table displays the estimated proportions of zeroalpha.5% 100% 67.2% 90.6% 16.0% αL2 αB1 π0 58.8% 39.4% 9.0% 27.5% 8.1% 69.8% 93.8% 39.2% 60.2% 0.5% 87. We also gave the result using a linear factor model: the NeweyWest (1987) heteroscedasticity and autocorrelation consistent estimator.5% 49.5% 8.7%% 17.5% 63.1% 17.2% 16. TVCM 2.6% 31.1% 22. S.2% 66.8% 7. we create the mean (α) which represents the stockpicker ability and the indicators deﬁned in section 7 during the 2 crisis:αL1 and αL2 for LTCM and αB1 and αB2 for the Equity Bubble which represents the diﬀerent market timer abilities.7% 59.2% Nwest 5. We estimate alpha with the timevarying coeﬃcient model deﬁned in section 6.0% 55.4% 0. Furthermore the results about the other strategies are available upon request.6% 71.5% 0.9% 19.3% CTA TVCM 18.4% 21.6% 31.5% 74.2% 83. Market Neutral TVCM 11.2% 58.2% 0. We do not give some results for Single Strategy because of too few data.3% Eq.1% 89% 97.9% 18.2% 71. and skilled funds from merge database See Appendix I for deﬁnitions of fund types.6% 53.8%% Emerging Markets TVCM 12.3% − πA + πA αB2 π0 24.4% 0.4% 22.8% 2.9% 17.2% 7.8% Nwest 2% . With the estimates. Strategy π0 58.3% 0.5% 41.0% 61.0% 0. The funds used at minimum cover the LTCM and bubble period which represent a track record a minimum of 30 months.0% 92. unskilled.8% 4.3% − πA used + πA Eq.9% 43.2% 55.4% 0.2% 22.0% 28.2% 83.2% 22.4% 12.0% 43.5% Nwest 3.0% 90.3% 45.1% 4.8% 26.8% 62.8% 27.0% 2.9% 19.6% Nwest 8.6% 24.0% − πA + πA Model π0 − πA + πA α π0 − πA + πA αL1 17.2% 38.6% 49.2% 15.0% 97. unskilled.1% 97% 1% 28.0% 27.0% 96% 9.5% 0.0% 55.4% 21.6% 0.9% 44.3% 44. and skilled fund for each strategy after applying the false Discovery rate methodology developed by Barras.2% 42.8% 0.0% 18. These results come from the merge between the CISDM and the HegdeFund.5% 2.8% 0% 26.4% 99.
The bar ﬁgures illustrate the number of breakdates relative to one fund.05 0 31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007 Distressed Securities 0. In this way.1 0.15 0. The number of hedge Funds has increased drastically since 2000 therefore the number of breaks in 1995.05 0 31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007 Emerging Markets 0.8 0. 32 .15 LTCM Equity Bubble Rbreaks 0. In order to overcome this problem and to study correctly the dynamic in beta we deﬁne a ratio: Rbreaks which is equal to the number of funds at time t divided by the number of break at time t.15 LTCM Equity Bubble Rbreaks 0.2 0.2 0.4 0. we are able to study how the dynamic in hedge Funds behaves during a long period.05 0 31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007 CTA 0.2 0.1 0.6 0.Figure I: Capital Structure Arbitrage 1 0.2 0 31−Mar−1997 30−Nov−1998 31−Jul−2000 31−Mar−2002 30−Nov−2003 31−Jul−2005 31−Mar−2007 LTCM Equity Bubble Rbreaks Convertible Arbitrage 0.05 0 31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007 LTCM Equity Bubble Rbreaks Equity Long Only 0.3 LTCM Equity Bubble Rbreaks 0.1 0.25 0.2 0.4 0.2 0.1 0 31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007 See Appendix I for deﬁnitions of fund types. has not the same signiﬁcance as the number of breaks in 2002.15 LTCM Equity Bubble Rbreaks 0. for instance.1 0.
05 0 31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007 Event Driven Multi Strategy 0.2 0.3 LTCM Equity Bubble Rbreaks 0.05 0 31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007 LTCM Equity Bubble Rbreaks See Appendix I for deﬁnitions of fund types.2 0.15 LTCM Equity Bubble Rbreaks 0.Figure II: Equity Long/Short 0.2 0. for instance. In this way.25 0. 33 . The number of hedge Funds has increased drastically since 2000 therefore the number of breaks in 1995.5 0.4 0. In order to overcome this problem and to study correctly the dynamic in beta we deﬁne a ratio: Rbreaks which is equal to the number of funds at time t divided by the number of break at time t.15 0.1 0 31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007 LTCM Equity Bubble Rbreaks Fixed Income Arbitrage 0.05 0 31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007 Equity Market Neutral 0.1 0 31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007 Fixed Income 0.15 0.4 0. has not the same signiﬁcance as the number of breaks in 2002.2 0.2 0.1 0.25 0.05 0 31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007 LTCM Equity Bubble Rbreaks Fixed Income − MBS 0.1 0.15 LTCM Equity Bubble Rbreaks 0. The bar ﬁgures illustrate the number of breakdates relative to one fund.1 0. we are able to study how the dynamic in hedge Funds behaves during a long period.3 0.1 0.2 0.
Figure III: Global Macro 0.05 0 31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007 Merger Arbitrage 0. The bar ﬁgures illustrate the number of breakdates relative to one fund. The number of hedge Funds has increased drastically since 2000 therefore the number of breaks in 1995. for instance.2 0.1 0.1 0 31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007 Multi Strategy 0. has not the same signiﬁcance as the number of breaks in 2002.4 0.3 LTCM Equity Bubble Rbreaks 0. In this way.3 0. In order to overcome this problem and to study correctly the dynamic in beta we deﬁne a ratio: Rbreaks which is equal to the number of funds at time t divided by the number of break at time t. 34 .5 0.2 0.4 0.2 0.1 0 31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007 LTCM Equity Bubble Rbreaks Option Arbitrage LTCM Equity Bubble 31−Dec−1993 30−Nov−1995 30−Sep−1997 31−Aug−1999 30−Jun−2001 31−May−2003 31−Mar−2005 28−Feb−2007 31−Dec−1993 Other Relative Value LTCM Equity Bubble 31−Dec−1993 30−Nov−1995 30−Sep−1997 31−Aug−1999 30−Jun−2001 31−May−2003 31−Mar−2005 28−Feb−2007 31−Dec−1993 Regulation D LTCM Equity Bubble 31−Dec−1993 30−Nov−1995 30−Sep−1997 31−Aug−1999 30−Jun−2001 31−May−2003 31−Mar−2005 28−Feb−2007 31−Dec−1993 See Appendix I for deﬁnitions of fund types.15 LTCM Equity Bubble Rbreaks 0. we are able to study how the dynamic in hedge Funds behaves during a long period.
2 0.1 0 31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007 LTCM Equity Bubble Rbreaks Single Strategy 1 0. In order to overcome this problem and to study correctly the dynamic in beta we deﬁne a ratio: Rbreaks which is equal to the number of funds at time t divided by the number of break at time t. The number of hedge Funds has increased drastically since 2000 therefore the number of breaks in 1995. In this way.2 0 28−Feb−1994 31−Jan−1996 31−Dec−1997 30−Nov−1999 30−Sep−2001 31−Aug−2003 31−Jul−2005 30−Jun−2007 28−Feb−1994 LTCM Equity Bubble Rbreaks unclassified 0.25 0. we are able to study how the dynamic in hedge Funds behaves during a long period.2 0.05 0 31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007 LTCM Equity Bubble Rbreaks Short Bias 0.2 0.5 0.1 0.25 0.05 0 31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007 See Appendix I for deﬁnitions of fund types. has not the same signiﬁcance as the number of breaks in 2002.15 0.15 0.6 0.4 0.4 0.8 0. The bar ﬁgures illustrate the number of breakdates relative to one fund.2 0.15 LTCM Equity Bubble Rbreaks 0. 35 .05 0 31−Dec−1993 31−Aug−1995 30−Apr−1997 31−Dec−1998 31−Aug−2000 30−Apr−2002 31−Dec−2003 31−Aug−2005 30−Apr−2007 LTCM Equity Bubble Rbreaks Sector 0.3 0.1 0. for instance.Figure IV: Relative Value Multi Strategy 0.1 0.
02 0.02 TFCo TFCu 0.019 0. .Figure V: Emerging Markets π− A 5) SSd 6) BM 7) CS 8) EM : Size Spread factor : Bond market factor : Credit Spread factor : Emerging Market factor π0 π+ A LTCM second period Equity Bubble first period Equity Bubble second period Legend: 1) TFB : Bond Trend−Following factor 2) BTCo : Commodity Trend−Following factor 3) TFCu : Currency Trend−Following factor 4) SP500 : S&P 500 LTCM first period factor factor factor EM CS BdMt SSd SP500 TFCo TFCu TFB factor 50% 100% 100% 0% 50% 100% 0% 0% 50% 100% EM CS BdMt SSd SP500 TFCo TFCu TFB EM CS BdMt SSd SP500 TFCo TFCu TFB EM CS BdMt SSd SP500 TFCo TFCu TFB 0% 50% 36 LTCM Second Period TFCo 0.02 Equity Bubble Second Period TFCo 0.049 0.01 0. Each crisis is divided in 2 period (see section 7: methodology).0041 TFCu 0.016 0.03 0. The number of funds covering the period is equal to 461. constant (white or π0 ).0082 0.038 SP500 0. . The bar ﬁgures illustrate the proportion of funds having more than 10 − + percent decrease (blue or πA ).012 0.025 LTCM first Period TFCo TFCu 0.029 0.0099 0.005 Equity Bubble First Period SP500 0. and increase (red or πA ) market exposure during the 2 crisis.048 SP500 0.0096 SSd TFB SSd TFB SSd TFB SSd TFB BdMt CS EM BdMt EM BdMt CS EM BdMt CS EM CS See Appendix I for deﬁnitions of fund types.04 TFCu SP500 0. The second ﬁgures (Factor Radar Chart) indicates the strategy’s sensitivities (percentage change) to various factors in regards to the 2 crisis.015 0.
0043 SSd TFB SSd TFB SSd TFB SSd TFB BdMt CS EM BdMt EM BdMt CS EM BdMt CS EM CS See Appendix I for deﬁnitions of fund types.023 Equity Bubble Second Period TFCo 0.013 0.029 TFCo TFCu 0.023 0. .022 SP500 0. The second ﬁgures (Factor Radar Chart) indicates the strategy’s sensitivities (percentage change) to various factors in regards to the 2 crisis.021 0.0057 Equity Bubble First Period SP500 0.011 0.Figure VI: Equity Long/Short π− A 5) SSd 6) BM 7) CS 8) EM : Size Spread factor : Bond market factor : Credit Spread factor : Emerging Market factor π0 π+ A LTCM second period Equity Bubble first period Equity Bubble second period Legend: 1) TFB : Bond Trend−Following factor 2) BTCo : Commodity Trend−Following factor 3) TFCu : Currency Trend−Following factor 4) SP500 : S&P 500 LTCM first period factor factor factor EM CS BdMt SSd SP500 TFCo TFCu TFB factor 50% 100% 100% 0% 50% 100% 0% 0% 50% 100% EM CS BdMt SSd SP500 TFCo TFCu TFB EM CS BdMt SSd SP500 TFCo TFCu TFB EM CS BdMt SSd SP500 TFCo TFCu TFB 0% 50% 37 LTCM Second Period TFCo 0.0041 0. Each crisis is divided in 2 period (see section 7: methodology).017 0. .012 0.0087 0. and increase (red or πA ) market exposure during the 2 crisis.012 0.017 SP500 0.028 LTCM first Period TFCo TFCu 0. constant (white or π0 ). The number of funds covering the period is equal to 2525.017 0.0058 TFCu 0. The bar ﬁgures illustrate the proportion of funds having more than − + 10 percent decrease (blue or πA ).0082 0.016 TFCu SP500 0.
The bar ﬁgures illustrate the proportion of funds having more than 10 − + percent decrease (blue or πA ).02 LTCM first Period TFCo TFCu 0.0043 TFCu 0. and increase (red or πA ) market exposure during the 2 crisis.017 0.032 TFCu SP500 0.045 SP500 0.013 0.024 0.Figure VII: Equity Market Neutral π− A 5) SSd 6) BM 7) CS 8) EM : Size Spread factor : Bond market factor : Credit Spread factor : Emerging Market factor π0 π+ A LTCM second period Equity Bubble first period Equity Bubble second period Legend: 1) TFB : Bond Trend−Following factor 2) BTCo : Commodity Trend−Following factor 3) TFCu : Currency Trend−Following factor 4) SP500 : S&P 500 LTCM first period factor factor factor EM CS BdMt SSd SP500 TFCo TFCu TFB factor 50% 100% 100% 0% 50% 100% 0% 0% 50% 100% EM CS BdMt SSd SP500 TFCo TFCu TFB EM CS BdMt SSd SP500 TFCo TFCu TFB EM CS BdMt SSd SP500 TFCo TFCu TFB 0% 50% 38 LTCM Second Period TFCo 0.0041 Equity Bubble First Period SP500 0. constant (white or π0 ).012 0. The second ﬁgures (Factor Radar Chart) indicates the strategy’s sensitivities (percentage change) to various factors in regards to the 2 crisis.0085 0. . . The number of funds covering the period is equal to 519.027 0. Each crisis is divided in 2 period (see section 7: methodology).016 0.036 SP500 0.021 TFCo TFCu 0.018 0.016 Equity Bubble Second Period TFCo 0.039 0.0081 0.0079 0.0091 SSd TFB SSd TFB SSd TFB SSd TFB BdMt CS EM BdMt EM BdMt CS EM BdMt CS EM CS See Appendix I for deﬁnitions of fund types.
.021 0. and increase (red or πA ) market exposure during the 2 crisis.034 0.0085 0.042 TFCo TFCu 0.0057 SSd TFB SSd TFB SSd TFB SSd TFB BdMt CS EM BdMt EM BdMt CS EM BdMt CS EM CS See Appendix I for deﬁnitions of fund types.025 0. The number of funds covering the period is equal to 293.025 0. The bar ﬁgures illustrate the proportion of funds having more than 10 − + percent decrease (blue or πA ).011 0. .0084 Equity Bubble First Period SP500 0.034 Equity Bubble Second Period TFCo 0. Each crisis is divided in 2 period (see section 7: methodology). constant (white or π0 ).Figure VIII: Event Driven Multi Strategy π− A 5) SSd 6) BM 7) CS 8) EM : Size Spread factor : Bond market factor : Credit Spread factor : Emerging Market factor π0 π+ A LTCM second period Equity Bubble first period Equity Bubble second period Legend: 1) TFB : Bond Trend−Following factor 2) BTCo : Commodity Trend−Following factor 3) TFCu : Currency Trend−Following factor 4) SP500 : S&P 500 LTCM first period factor factor factor EM CS BdMt SSd SP500 TFCo TFCu TFB factor 50% 100% 100% 0% 50% 100% 0% 0% 50% 100% EM CS BdMt SSd SP500 TFCo TFCu TFB EM CS BdMt SSd SP500 TFCo TFCu TFB EM CS BdMt SSd SP500 TFCo TFCu TFB 0% 50% 39 LTCM Second Period TFCo 0.0084 TFCu 0.029 SP500 0.023 SP500 0.017 TFCu SP500 0.017 0. The second ﬁgures (Factor Radar Chart) indicates the strategy’s sensitivities (percentage change) to various factors in regards to the 2 crisis.017 0.042 LTCM first Period TFCo TFCu 0.013 0.017 0.0042 0.
constant (white or π0 ). . The second ﬁgures (Factor Radar Chart) indicates the strategy’s sensitivities (percentage change) to various factors in regards to the 2 crisis. The number of funds covering the period is equal to 377. .006 0.013 0.0085 0. The bar ﬁgures illustrate the proportion of funds having more than 10 − + percent decrease (blue or πA ).018 0.Figure IX: Global Macro π− A 5) SSd 6) BM 7) CS 8) EM : Size Spread factor : Bond market factor : Credit Spread factor : Emerging Market factor π0 π+ A LTCM second period Equity Bubble first period Equity Bubble second period Legend: 1) TFB : Bond Trend−Following factor 2) BTCo : Commodity Trend−Following factor 3) TFCu : Currency Trend−Following factor 4) SP500 : S&P 500 LTCM first period factor factor factor EM CS BdMt SSd SP500 TFCo TFCu TFB factor 50% 100% 100% 0% 50% 100% 0% 0% 50% 100% EM CS BdMt SSd SP500 TFCo TFCu TFB EM CS BdMt SSd SP500 TFCo TFCu TFB EM CS BdMt SSd SP500 TFCo TFCu TFB 0% 50% 40 LTCM Second Period TFCo 0.024 TFCu SP500 0.017 TFCo TFCu 0.03 0.0034 TFCu 0.01 0.02 0.017 Equity Bubble Second Period TFCo 0.0067 SSd TFB SSd TFB SSd TFB SSd TFB BdMt CS EM BdMt EM BdMt CS EM BdMt CS EM CS See Appendix I for deﬁnitions of fund types.0069 0.012 0.013 0.014 0. Each crisis is divided in 2 period (see section 7: methodology). and increase (red or πA ) market exposure during the 2 crisis.034 SP500 0.021 LTCM first Period TFCo TFCu 0.027 SP500 0.0043 Equity Bubble First Period SP500 0.
constant (white or π0 ).015 0. The bar ﬁgures illustrate the proportion of funds having more than 10 − + percent decrease (blue or πA ).02 0.0067 0.0051 Equity Bubble First Period SP500 0.021 Equity Bubble Second Period TFCo 0.03 0.011 0.022 0.01 0.Figure X: Short Bias π− A 5) SSd 6) BM 7) CS 8) EM : Size Spread factor : Bond market factor : Credit Spread factor : Emerging Market factor π0 π+ A LTCM second period Equity Bubble first period Equity Bubble second period Legend: 1) TFB : Bond Trend−Following factor 2) BTCo : Commodity Trend−Following factor 3) TFCu : Currency Trend−Following factor 4) SP500 : S&P 500 LTCM first period factor factor factor EM CS BdMt SSd SP500 TFCo TFCu TFB factor 50% 100% 100% 0% 50% 100% 0% 0% 50% 100% EM CS BdMt SSd SP500 TFCo TFCu TFB EM CS BdMt SSd SP500 TFCo TFCu TFB EM CS BdMt SSd SP500 TFCo TFCu TFB 0% 50% 41 LTCM Second Period TFCo 0. The second ﬁgures (Factor Radar Chart) indicates the strategy’s sensitivities (percentage change) to various factors in regards to the 2 crisis.027 TFCo TFCu 0.01 SSd TFB SSd TFB SSd TFB SSd TFB BdMt CS EM BdMt EM BdMt CS EM BdMt CS EM CS See Appendix I for deﬁnitions of fund types. and increase (red or πA ) market exposure during the 2 crisis.0054 TFCu 0.013 0. The number of funds covering the period is equal to 56.027 TFCu SP500 0. Each crisis is divided in 2 period (see section 7: methodology). .05 SP500 0.016 0.02 0. .04 SP500 0.033 0.026 LTCM first Period TFCo TFCu 0.
0062 TFCu 0. constant (white or π0 ).019 SP500 0.0076 0. Each crisis is divided in 2 period (see section 7: methodology).0039 0.019 0.015 TFCu SP500 0.019 0. . and increase (red or πA ) market exposure during the 2 crisis.0064 Equity Bubble First Period SP500 0.015 SP500 0.Figure XI: CTAsMerge π− A 5) SSd 6) BM 7) CS 8) EM : Size Spread factor : Bond market factor : Credit Spread factor : Emerging Market factor π0 π+ A LTCM second period Equity Bubble first period Equity Bubble second period Legend: 1) TFB : Bond Trend−Following factor 2) BTCo : Commodity Trend−Following factor 3) TFCu : Currency Trend−Following factor 4) SP500 : S&P 500 LTCM first period factor factor factor EM CS BdMt SSd SP500 TFCo TFCu TFB factor 50% 100% 100% 0% 50% 100% 0% 0% 50% 100% EM CS BdMt SSd SP500 TFCo TFCu TFB EM CS BdMt SSd SP500 TFCo TFCu TFB EM CS BdMt SSd SP500 TFCo TFCu TFB 0% 50% 42 LTCM Second Period TFCo 0.0077 0.012 0.032 LTCM first Period TFCo TFCu 0.018 0. The second ﬁgures (Factor Radar Chart) indicates the strategy’s sensitivities (percentage change) to various factors in regards to the 2 crisis. .013 0. The number of funds covering the period is equal to 1394. The bar ﬁgures illustrate the proportion of funds having more than − + 10 percent decrease (blue or πA ).011 0.031 TFCo TFCu 0.0038 SSd TFB SSd TFB SSd TFB SSd TFB BdMt CS EM BdMt EM BdMt CS EM BdMt CS EM CS See Appendix I for deﬁnitions of fund types.026 Equity Bubble Second Period TFCo 0.012 0.025 0.
• The Equity Long/Short strategy refers to taking both long and short positions in equities.combination of discretionary and systematic macro. Systematic macro (Systematic diversiﬁed) . or trade claims at deep discounts of companies in or facing bankruptcy or reorganization. • Mortgagebacked securities arbitrage style exploit the mispricing of mortgagebacked assets relative to Treasury securities.exploit pricing ineﬃciencies between convertible securities and the corresponding stocks. • Fixed Income Arbitrage Strategies encompass a wide range of strategies in both domestic and global ﬁxedincome markets.Multi strategy in Macro strategy . with the expectation that their prices will converge to appropriate values in the future(Arbitrage.Appendix I Deﬁnition of Strategies • The emerging markets strategy attempts to capture gains from ineﬃciencies in emerging markets. • The event driven multistrategy can use both the distressed securities style and/or the merger arbitrage style. • The Relative Value arbitrage style take positions in 2 securities that are mispriced relative to each other.a hedge fund exploiting a combination of diﬀerent hedge fund strategies to reduce market risk. This strategy exploits the mispricing between a stock which is overvalued and one that is undervalued such that beta of the combined position is zero. generated by software without human intervention. Convertible arbitrage . The Distressed Securities Strategy focuses on asset of distressed companies. • The Merger Arbitrage Strategy also called risk arbitrage strategy exploit pricing inefﬁciencies associated with a merger or acquisition. • Equity Market Neutral style creates a position that attempts to hedge out most market risk by taking oﬀsetting positions. • Index arbitrage style generally attempts to exploit mispricing between an index and derivatives on that index.trading is done mathematically.exploit pricing ineﬃciencies between related ﬁxed income securities. • Global Macro Discretionary macro . Fixed income arbitrage . • multistrategy style uses diﬀerent styles and may change exposures to diﬀerent styles based upon changing market conditions. • The arbitrage involves simultaneously purchasing and selling related securities that are mispriced relative to each other. • The Market Timer focus on securities associated with companies that will soon experience a signiﬁcant event. Multi strategy in FoHF . • Convertible Arbitrage Strategy can be described by taking a long position in a convertible bond and sells short the associated stock. • dedicated short selling style only takes short equity positions. Statistical arbitrage . 43 .trading is done by investment managers instead of generated by software. Market neutral. debt. Buys equity.equity market neutral strategy using statistical models.
longterm or shortterm.involves taking long and short positions in diﬀerent ﬁnancial instruments of a company’s capital structure.expertise in niche areas such as technology.specialized in corporate ﬁxed income securities. basic materials. • Regulation D .proﬁt from trend reversals. 44 . • Credit arbitrage or ﬁxed income arbitrage strategy . • Multi manager . • Fundamental growth . Nontrend following (Counter trend) .non ﬁxed income arbitrage strategies based on the yield instead of the price.ﬁxed income arbitrage strategy using assetbacked securities. • Fundamental value .• Sector funds . • Volatility arbitrage . • Quantitative Directional. • Yield alternatives .invest in companies with more earnings growth than the broad equity market. statistical arbitrage . particularly between a company’s debt and equity product. energy.exploit the change in implied volatility instead of the change in price. • Capital Structure Arbitrage . • Fixed Income asset backed .a hedge fund where the investment is spread along separate sub managers investing in their own strategy. • Trend following . health care.equity trading using quantitative techniques.invest in undervalued companies. biotechnology. pharmaceuticals.specialized in private equities.
X1 . .2p (X0 W0 X0 )−1 X0 W0 Y With these estimates./h)/h. . Xp ] = 0. .. Xn1 Xn1 (Tn − t0 ) . Yn ) and W0 = diag(Kh0 (T1 − t0 ). . . Xp ] = σ 2 (t). ... a twostep estimator is obtained. This leads to the following local leastsquares problem: n Yi − p j=1 2 {aj + bj (Ti − t0 )}Xij Kh (Ti − t0 ) minimize i=1 for a given kernel function K and bandwidth h. where Kh (. Fan and Zhang showed that the bias of the twostep estimator is of O(h4 ) and the variance 2 O (nh2 )−1 45 .. . X1 . In matrix notation: Let X11 X11 (T1 − t0 ) . Kh0 (Tn − t0 )) Then the solution to the leastsquares problem can be expressed as: aj.. X1 .) = K(. V ar[εt. Xnp X1p (T1 − t0 ) .. In this way. we can approximate the function locally as βi (t) ≈ ai + bi (t − t0 ).0 . . . . X0 = . ... In this study. Xnp (Tp − t0 ) Y = (Y1 ... . ..Twostep TimeVarying Coeﬃcient Model The varying coeﬃcient model assumes the following conditional linear structure: p Yt = j=1 βj (t)Xjt + εt = α(t) + Xβ(t) + εt for a given covariates (t. .. X1p . if we consider that βi depends on t: (βi (t)). . .. . we took X1 = 1 as the intercept term and t = time. aj..0 (Ti )Xij − ap + bp (Ti − t0 ) + cp (Ti − t0 )2 + dp (Ti − t0 )3 Xip ×Kh2 (Ti −t0 ) ˆ where h2 is the bandwidth in the second step. . Xp ) and variable Y with E[εt.0 = e2j−1. .. a local leastsquare regression is ﬁtted again via substituting the initial estimate into the local leastsquares problem: Yi − i=1 j=1 n p−1 2 aj.. .
knowing that we can approximate each βp (t) by a basis function expansion K βp (t) k=0 ∗ γpk Bpk (t). From the ﬁrst step.0 = e2j−1. ˆ k=1 46 .2p (X0 W0 X0 )−1 X0 W0 Y In a second step.0 (Ti )Xij − ˆ j=1 k=0 then it is natural to estimate βp (t) by ˆ βp (t) = K γkp Bkp (t). ∗ We can now minimized in order to estimate γkp : n i=1 wi Yi − p−1 K 2 γkp Bkp Xip aj.Twostep TimeVarying Coeﬃcient Model using Bsplines As mentioned by Fan and Zhang (1999). Therefore we built the second twostep estimator based on the same article but we used during the second step a smoothing splines instead of local regression. other techniques such as smoothing splines can also be used in the second stage of ﬁtting. we obtained the estimates: aj.
We created two “betas i ”(called βcreated ) which represent a possible structural change for a Hedge Fund. The random variable ε follows a normal distribution with mean zero and variance 1. Short sample: 50 months ˆ β1 and true β1 using LTVCM 2 1 0 −1 −2 0 10 ˆ β2 and true β2 using LTVCM 5 0 20 40 60 −5 0 months ˆ β2 and true β2 using BTVCM 4 6 4 2 0 0 −2 40 60 months ˆ β2 and true β2 using BTVCM 20 2 −2 0 20 months 40 60 0 20 months 40 60 FIG. We called the local timevarying coeﬃcient model: LTVCM and the Bspline time varying coeﬃcient model: BTVCM.: Comparison of the performance between the onestep estimator (longdashed curve) and the true coeﬃcient function (the solid curve). X2 are the S&P500 and the monthly change in the 10year treasury constant maturity yield respectively. We put up a simulated Hedge Fund track (RHF ) in this manner: 1 2 RHF = βcreated X1 + βcreated X2 + ε where X1 .Robustness: local and Bspline TimeVarying Coeﬃcient Model The following example will be used to illustrate the performance of our estimator. 47 .
5 0 −0.: Comparison of the performance between the onestep estimator (longdashed curve) and the true coeﬃcient function (the solid curve).5 −1 2 1 0 −1 −2 0 50 100 months 150 0 50 100 months 150 FIG. 48 . Long sample: 150 months ˆ β1 and true β1 using LTVCM 1 4 ˆ β2 and true β2 using LTVCM 0 2 −1 0 −2 0 100 150 months ˆ β2 and true β2 using BTVCM 50 −2 0 100 150 months ˆ β2 and true β2 using BTVCM true β ˆ β 50 1 0.5 0 −0.: Comparison of the performance between the onestep estimator (longdashed curve) and the true coeﬃcient function (the solid curve).Medium sample: 100 months ˆ β1 and true β1 using LTVCM 1 4 ˆ β2 and true β2 using LTVCM 0 2 −1 0 −2 0 50 100 months ˆ β2 and true β2 using BTVCM −2 0 50 100 months ˆ β2 and true β2 using BTVCM 1 0.5 −1 6 4 2 0 −2 0 50 months 100 0 50 months 100 FIG.
now. U(α/2) (t)). so that B bootstrap estimators βi (t) of βi (t) are obtained. 1 = j = ni } be the longitudinal bootstrap sample. • 4) Let Lα/2 (t) and U(α/2) (t) be the (α/2)th and (1 − α)th i.j (Xj ) = n−1 ˆ i=j Whi (Xj )Yi .Bootstrap Conﬁdence intervals We summarize the methodology from Colin and Chiang (2000) in order to create conﬁdence regions. According to the author this following naive bootstrap procedure can be used to construct approximate pointwise percentile conﬁdence intervals for βi (t): • 1) Randomly sample n subjects with replacement from the original data set. j=1 (Yj − mh. n • 2) Construct the cross validation function CV (h) = n−1 denotes a weight function. say the jth. bandwidth ”leaveonesubjectout” crossvalidation methodology The leaveout method is based on regression smoothers in which one. lower and upper (α/2th ) percentiles. So. deﬁne the automatic bandwidth as h = arg min [CV (h)] h∈Hn 49 . An approximate (1 − α) bootstrap conﬁdence interval for βi (t) is given by (L(α/2) (t). calculated on the B bootstrap estimators. ij i boot • 2) Compute the kernel estimator βi (t) boot • 3) Repeat the above 2 steps B times. Yij . where w ˆ ˆ • 3) With this N CV(h). X∗ ).e. observation is left out.j (Xj ))2 w(Xj ). • 1) Compute the leaveout estimate mh. we can. 1 = i = n. for N values. respectively. and let ∗ {(t∗ .
This action might not be possible to undo. Are you sure you want to continue?