Presentation is loading. Please wait.

Presentation is loading. Please wait.

B A R C L A Y S G L O B A L I N V E S T O R S 0 Stan Beckers Simon Weinberger Barclays Global Investors Fundamental Factors in Hedge Fund Returns Spitalfields.

Similar presentations


Presentation on theme: "B A R C L A Y S G L O B A L I N V E S T O R S 0 Stan Beckers Simon Weinberger Barclays Global Investors Fundamental Factors in Hedge Fund Returns Spitalfields."— Presentation transcript:

1 B A R C L A Y S G L O B A L I N V E S T O R S 0 Stan Beckers Simon Weinberger Barclays Global Investors Fundamental Factors in Hedge Fund Returns Spitalfields Day Cambridge 10 March 2005

2 B A R C L A Y S G L O B A L I N V E S T O R S 1 Overview  The Raw Data: Issues  Skewness, Kurtosis and Autocorrelation  Communality in Hedge Fund Returns  Systematic and Residual Factors

3 B A R C L A Y S G L O B A L I N V E S T O R S 2 1. Hedge Fund Returns : Data Issues  Return data only (no transparency)  Bias in Pricing/ Returns Survivorship Bias Instant History Bias Self-Reporting Bias  Short histories, low frequency data  Fund size ignored

4 B A R C L A Y S G L O B A L I N V E S T O R S 3 HFR Database : Fund inception date, Reporting start and end date

5 B A R C L A Y S G L O B A L I N V E S T O R S 4 Short histories

6 B A R C L A Y S G L O B A L I N V E S T O R S 5 Histogram of Hedge Fund AUM (Sept 2004): Not all funds are equally important

7 B A R C L A Y S G L O B A L I N V E S T O R S 6 2. More data issues : these things aren’t normal !?  Skewness and Kurtosis Downside Protection Use of derivatives Non-Linear Factors  Autocorrelation in return series Data Smoothing

8 B A R C L A Y S G L O B A L I N V E S T O R S 7 Skewness and Kurtosis -1.7-1.4-1.1 -0.8 -0.5-0.2 0.10.4 0.7 1 1.31.61.92.2 2.52.83.1 3.4 Standard Deviations

9 B A R C L A Y S G L O B A L I N V E S T O R S 8 Non-Linearity in Factors Period: January 1997- May 2004

10 B A R C L A Y S G L O B A L I N V E S T O R S 9 A Broad Cross Section of Funds Daily data

11 B A R C L A Y S G L O B A L I N V E S T O R S 10 A Broad Cross Section of Funds Monthly data

12 B A R C L A Y S G L O B A L I N V E S T O R S 11 Characteristics Return Distribution Fund X June 2002 – December 2004

13 B A R C L A Y S G L O B A L I N V E S T O R S 12 Characteristics Return Distribution Fund X January 1994 – June 2004

14 B A R C L A Y S G L O B A L I N V E S T O R S 13 3. Looking for Communality in Hedge Fund Returns  Hedge Fund Styles as defined by the Index Providers Self-Declared Opportunistic  Statistical Approaches Cluster Analysis Principal Component Analysis

15 B A R C L A Y S G L O B A L I N V E S T O R S 14 The HFR Hedge Fund Style Classification HFRI Convertible Arbitrage Index HFRI Distressed Securities Index HFRI Emerging Markets (Total) HFRI Equity Hedge Index HFRI Equity Market Neutral Index HFRI Equity Non-Hedge Index HFRI Event-Driven Index HFRI Fixed Income (Total) HFRI Fixed Income: Arbitrage Index HFRI Fixed Income: Convertible Bonds Index HFRI Fixed Income: Diversified Index HFRI Fixed Income: High Yield Index HFRI Fixed Income: Mortgage-Backed Index HFRI Macro Index HFRI Market Timing Index HFRI Merger Arbitrage Index HFRI Relative Value Arbitrage Index HFRI Short Selling Index

16 B A R C L A Y S G L O B A L I N V E S T O R S 15 Defining the Number of Hedge Fund Styles: Cluster Analysis[1][1] [1] We require 60 months of data (199907-200406), which leaves us with 676 funds for this analysis.

17 B A R C L A Y S G L O B A L I N V E S T O R S 16 Defining the Number of Hedge Fund Styles: Average Return Correlation with Peers

18 B A R C L A Y S G L O B A L I N V E S T O R S 17 Defining the Number of Hedge Fund Styles: Explanatory Power of first Principal Component

19 B A R C L A Y S G L O B A L I N V E S T O R S 18 HFR Hedge Fund Style Classification

20 B A R C L A Y S G L O B A L I N V E S T O R S 19 Histograms of average correlation with peers (same hedge fund style) versus non-peers Correlation Structure with Peers and Non-Peers

21 B A R C L A Y S G L O B A L I N V E S T O R S 20 Cluster Analysis within Broadly Defined Styles

22 B A R C L A Y S G L O B A L I N V E S T O R S 21 4. Dissecting Within-Style Hedge Fund Returns: Systematic and Residual factors  A Primer on Multiple Factor Models Differentiating between Alpha and Beta  Identifying systematic factors Principal Component analysis Fundamental Factors  Where is the hedge?

23 B A R C L A Y S G L O B A L I N V E S T O R S 22 The Academic Background on Multiple Factor Models  The CAPM Single Factor : The Market Portfolio  APT : Multiple Factors Factors Undefined but the academic world would probably agree that – for equities - they include —Small versus Large —Value Versus Growth —Momentum Broad Approaches for Factor Indentification —Macro-Economic Models —Fundamental Models —Statistical Models  Virtually all models are Linear

24 B A R C L A Y S G L O B A L I N V E S T O R S 23 Factor Model Selection Criteria  Academically Sound?  Best Fit?  Economic Interpretation?  Out of sample explanatory power?  Tradeable?

25 B A R C L A Y S G L O B A L I N V E S T O R S 24 A First Cut at Identifying Factors within each Style: Principal Component Analysis

26 B A R C L A Y S G L O B A L I N V E S T O R S 25 Selected Findings from the Hedge Fund Literature  Equity Strategies tend to have market exposure and exposure to Fama/French Factors SMB and HML (among others Fung and Hsieh, 2003)  30% of market-neutral funds have market risk exposure (Patton, 2004)  Option Strategies have explanatory power for non-directional strategies (Agarwal and Naik, 2000)  Trend Following Strategy exhibits payoff similar to by lookback straddle (Fung and Hsieh, 2001)  Merger Arbitrage exhibits payoff like uncovered put on Equity Index (Mitchell and Pulvino, 2000)

27 B A R C L A Y S G L O B A L I N V E S T O R S 26 Mean/Median Forecast Error at successive steps Equity Hedge Funds

28 B A R C L A Y S G L O B A L I N V E S T O R S 27 Explanatory power of a factor risk model (Equity Funds) Median : 35% in sample, 23% out of sample Truncated Median: 41% in sample, 34% out of sample

29 B A R C L A Y S G L O B A L I N V E S T O R S 28 Equity Hedge Fund In-Sample Alpha

30 B A R C L A Y S G L O B A L I N V E S T O R S 29 Mean/Median Forecast Error at successive steps Fixed Income Hedge Funds

31 B A R C L A Y S G L O B A L I N V E S T O R S 30 Explanatory power of a factor risk model (Fixed Income Funds) Histogram of in-sample (199401-200406) and out-of-sample fit (regression of fund returns on return explained by risk model (product sum of prevailing exposure estimate and realised factor return) Median * disregards observations with extreme fit ( 70%).

32 B A R C L A Y S G L O B A L I N V E S T O R S 31 Fixed Income Fund In Sample Alpha Fund-level rolling 36m estimation

33 B A R C L A Y S G L O B A L I N V E S T O R S 32 HFR Hedge Fund Index Returns: Where is the Hedge? T-Stat of Systematic Factors

34 B A R C L A Y S G L O B A L I N V E S T O R S 33 Summary  Given the Quality of the Data, all Hedge Fund Empirical Research has to be taken with a pinch of salt  Skewness, Kurtosis and Autocorrelation are less of an issue than some people would lead you to believe  Hedge Fund Styles are not clearly delineated and somewhat arbitrary  Even so, Common factors can be identified within broad hedge fund style classifications  Significant systematic factors are present in most hedge fund returns  Alpha does remain after taking systematic factors into account  Hedge Fund is somewhat of a misnomer


Download ppt "B A R C L A Y S G L O B A L I N V E S T O R S 0 Stan Beckers Simon Weinberger Barclays Global Investors Fundamental Factors in Hedge Fund Returns Spitalfields."

Similar presentations


Ads by Google