Replicating Hedge Funds: Traditional beta or alternative beta?

We are pleased to bring you a guest posting today from Dr. Lars Jaeger of Partners Group.  Regular readers and students of hedge fund replication will recall that Dr. Jaeger edited one of the first books covering the topic back in 2003.   Since then, he has written a number of papers on hedge fund replication (see related posting) and has spoken at many conferences.  He’s also one of the speakers on the programme for Terrapinn’s second Alternative Beta & Hedge Fund Replication conference in London (March 10-12).

Hedge Fund Replication and Alternative Beta: Two different ways of looking at replicating hedge funds

Special to AllAboutAlpha.com by: Lars Jaeger, Ph.D., CFA, FRM, Partner, Partners Group

The discussion on alpha versus beta in the breakdown of investment returns is as old as the capital asset pricing model which defined these terms some 40 years ago.  Today, it has finally reached the hedge fund industry – an area of investing traditionally associated with pure alpha or absolute return.

Investors and researchers alike realize that hedge fund returns are largely composed of betas. However, hedge fund beta is different from traditional beta.  While both are the result of exposures to systematic risks in the global capital markets, beta in hedge fund investing can be significantly more complex than traditional beta (and can be often be “hidden in the alpha smokescreen”).

We therefore refer to this beta as “alternative beta”, a term that now part of the hedge fund lexicon.  In fact, the increased academic and non-academic effort to model and understand hedge fund return sources has also reached Wall Street.  The new buzzword: “hedge fund replication”.

Most providers of “hedge fund replication” use a trading model application of a linear factor decomposition of hedge fund returns (as represented by particular return time series).  However, these models yield satisfactory results for only a few hedge fund strategies, namely directional ones such as Long/Short Equity or Event Driven (see Table 1 for the R-squared values for various strategy sectors).

When linear factor models work, traditional beta (i.e., equity market factors) are all that is required.  In other words, hedge fund replication of this kind works when investors need it least to work – in the directional equity space.

On the other hand, linear factor models are not satisfactorily successful when applied to most other hedge fund strategy sectors, specifically non-directional and Relative Value hedge funds or indices.  In order to cope with these strategies we have to employ more complex market factors and/or dynamically adjust factors to include non-linear profiles.  These factors are likely to represent dynamic trading strategies that capture alternative risk premia with their contingent and non-linear payout profiles.

In addition, linear models are backward looking only.  In contrast to hedge fund managers who base their decisions on current data, these models adjust exposures with a significant time lag.  This can be problematic in a fast changing environment!

Many hedge fund “replicators” use these backward looking regression models as the basis to estimate current and future hedge fund exposures.  They essentially ask “Why not give up on alternative beta and model hedge funds with traditional beta only?”

Actually, this approach would have proven reasonably successful in the last four years (which is exactly the period most providers choose to display when they show their back-tested performance to attract investors).   However, even in this benign environment for equity directional strategies, alternative betas can add value to investors’ portfolios.  Figure 1 shows a simple risk weighted average of the following generic alternative beta factors:

  • Deutsche Bank Carry Index,
  • CDX High Yield Index,
  • CRB Commodity Index,
  • value versus growth spread,
  • small cap versus large cap spread,
  • BXM covered call writing (BXM Index – 0.5*SPX Index),
  • the Merger Arbitrage Fund (MERFX  Index), and,
  • the spread between emerging market equity returns and developed equity markets (MSCI Emerging Markets – MSCI World).

The second chart represents the Merrill Lynch factor model which we believe is a very good proxy for the equity component in hedge fund returns.

Source: Bloomberg, PG calculation*

We see that alternative beta can yield similarly attractive results as equity beta in hedge funds during periods of equity bull markets. The result looks very different, however, when we include the bear market from March 2000 to March 2003, a period when hedge funds, in aggregate, made money despite heavy losses in the equity markets.  This is shown in the figure below, which is based on the same data – just extended further into the past.  Here we see the real attractiveness of alternative betas.

Source: Bloomberg, PG calculation

However, the large majority of alternative beta can only be extracted by conditional trading rules directly aimed at benefiting from particular risk premia in the global capital markets.  Instead of naively replicating past properties of doubtful (i.e., biased) time series with inappropriate (i.e., linear) models, it seems more appropriate to tackle the hedge fund risk premia/alternative betas one by one.  We believe that the hedge fund investor is well advised to stay with the alternative beta returns rather than chase the pseudo hedge fund returns of traditional beta.  Hedge funds may be simpler than what many investors have believed so far, but they are not that simple.

How recent market developments have impacted the field of “hedge fund replication”

In Roman mythology, Janus is the god of gates and doors as well as of beginnings and endings.  He is usually depicted with two faces looking in opposite directions.  The year 2007 will be remembered by investors as a Janus year.  The first half was sparked with euphoria and buoyant equity markets, only shortly interrupted by the turbulences in late February/early March.

Retrospectively, many consider this period now as the end of the four year period of low volatility and low investor risk aversion that began in 2003.  In contrast, the second half of 2007 proved one of the most difficult periods for investors in the last decade – comparable perhaps only to the summer of 1998.  The second half of 2007 have represented the beginning of a new phase in the global markets characterized by higher volatility, more uncertainty and hence higher risk premiums across asset classes.

A closer look reveals that the crisis triggered by the US sub-prime market, which sent shock waves around the world is mostly a non-equity crisis.  While the credit and the inter-bank lending market was sent into paralysis, the FX markets turned the carry trade sour, small cap and value underperformed large cap and growth, volatility soared, and merger deal spreads widened dramatically.  But the global equity markets themselves experienced limited losses (albeit showing much higher volatility).  The MSCI World index closed 2007 only about 5% below its all time high.

This caused hedge fund replication programs that focus on alternative (hedge fund) beta to underperform those based on traditional equity beta.  The first 20 days of trading in 2008 shows that this is about to change.

Note: The opinions expressed in this guest posting are those of the author and not necessarily those of AllAboutAlpha.com.

Be Sociable, Share!

3 Comments

Leave A Reply

← AAA Exclusive: Survey contains some surprises about how hedge fund managers now view hedge fund replication AAA Exclusive: An interview with Prof. Harry Kat about his newest project, "super-diversification" →