Siegel: Carefully selecting “exotic betas” a worthwhile pursuit

We are pleased to bring you a special guest posting today by one of the big names in institutional money management.  But unlike most big names, this one comes from the ranks of investors, not asset managers.  Laurence Siegel (see related posting) is the Director of Research at the Ford Foundation.  He was co-author of a hugely popular article in the spring 2006 edition of the Financial Analysts Journal called “The Myth of The Absolute Return Investor” that took issue with many of the popular assumptions about absolute return investing.  (Siegel’s co-author, BGI’s Barton Waring is featured in a summer 2006 webcast on the topic).  In addition, Siegel was also co-author of a great piece called “Five Myths About Fees“, which originally appeared in the Journal of Portfolio Management.

Siegel is speaking at a conference next week in New York on the topic of alternative beta and hedge fund replication.  In this exclusive, he gives us an overview of his thoughts on “exotic beta”. 

Are Exotic Betas Worth Investing In? A Brief Note.

By Laurence Siegel, Special to

What are exotic betas?  What are clone funds?  Has the sober science of money management been taken over by aliens from outer space?

No.  Let’s start with a brief (but not brief enough) history lesson.  Hedge funds evolved out of active management, the attempt to beat a benchmark using security holdings weights that differ from those in the benchmark.  Hedge funds differed from traditional actively managed funds in being able to sell short and use leverage; and, typically but not always, in having no fixed mandate (thus no fixed benchmark).  Nevertheless active management, including hedge fund management, is always about beating some sort of neutral or normal portfolio that you would hold in the absence of any views.  That neutral portfolio is the benchmark; if the fund is perfectly hedged to all systematic market factors, then there is still a benchmark, but it’s cash.

To achieve this result, many hedge funds adopted approaches to active management that involved making factor bets rather than individual security bets.  This behavior is a natural outcome of the attempt to simplify what would otherwise be an overly complex process of analyzing every security in one’s opportunity set.  Almost all managers do this, not just hedge fund managers. 

In U.S. equities, the common factors that seemed to work over long periods of time included: value minus growth, small cap minus large cap, momentum, selling volatility (writing portfolio insurance), and various arbitrage strategies (for example, buying takeover targets and shorting the acquirers, or buying convertible bonds and shorting the stock of the issuer).  Other asset classes have other factor-based strategies that work on average.  Note that any systematic strategy works only part of the time; for example, the value/growth and small/large cap effects are strongly cyclical.  Writing portfolio insurance works only when the market does not crash.

This emphasis on factor bets gives the appearance of the hedge fund community practicing what amounts to passive management – investing in widely recognized factors – but charging active (very active) fees for the payoff from the factor bets.  But if this is what hedge funds are really doing, then one shouldn’t invest in them.  Instead, one should invest directly in the factors, through portfolios of index funds, index derivatives, or funds specifically designed to generate the factor returns.  This last type of fund, which engages in hedge fund [return] replication, is sometimes called a clone fund, because it attempts to mimic an important part of the return of actively managed hedge funds (the beta part) without charging full hedge fund fees and without engaging in active security-level management. 

Should investors take seriously the opportunity presented by clone funds?  The answer depends on whether the exotic betas captured by clone funds have good risk-return characteristics, competitive with those offered by ordinary beta exposures.  If, and only if, the risk-return characteristics of the exotic betas are attractive, then buying clone funds is superior to buying, say, a hedge-fund index fund because you get the beta part of the return to the strategy without paying high hedge fund fees. 

However, if you have skill at selecting individual hedge funds, you would do even better, since you would also get the alpha part of the return.  (Having skill means the realized alpha, after fees, is a positive number.)

How can we determine whether the exotic betas offered by clone funds have attractive risk-return characteristics?  We can start by estimating their Sharpe ratios and comparing them to that of, say, the Sharpe ratio from investing in equities.  Let’s start with the Sharpe ratio of the stock market, proxied by the total return on the S&P 500.  In 2002, Richard Grinold and Ken Kroner of Barclays Global Investors, in a broadly collaborative effort that included my participation, estimated the equity risk premium (over intermediate-term Treasury bonds) at 2.5%.  The market has risen greatly since then, but so have earnings, so not much has changed in terms of overall market valuation levels – so let’s leave this estimate of the risk premium alone.  Meanwhile, the standard deviation of the S&P 500 since 1980, which is a good proxy for the expected standard deviation, has been 14.9%.  The expected Sharpe ratio of the S&P 500 is the ratio of these two numbers or 0.17.

The expected Sharpe ratios of the various exotic betas can then be compared to this number.  To give an example, let’s first calculate the historical Sharpe ratio of the value-minus-growth factor.  I am sitting on an airplane and do not have either current data or U.S. data handy, but I do have the EAFE value and growth index values from their inception (January 1, 1975) until March 31, 2000, just before the growth boom turned to bust and the subsequent value boom began.  Over this period, value outperformed growth by 3.1% per year, with the difference between value and growth returns having a standard deviation of 6.9% per year.  The Sharpe ratio of the EAFE value-minus-growth factor was thus 0.45, far higher than the expected Sharpe ratio of the S&P 500.

But is the historical Sharpe ratio a good forecast, a good expected value?  Note that the period studied excludes the huge value boom of 2000-2007, which would have further boosted the Sharpe ratio of value-minus-growth.  It also covers quite a long historical time span.  So a Bayesian adjustment of cutting the expected Sharpe ratio in half – from 0.45 to 0.22 – seems not only reasonable but conservative.  The adjusted Sharpe ratio of value-minus-growth is still very competitive with that of the equity risk premium.  Although value stocks do not seem like a  timely investment at the moment, this analysis shows that the value-minus-growth factor is worth investing in, at least as a very long-run, equilibrium proposition.

The Sharpe ratio, of course, does not tell the full story.  Some exotic betas have distributions for which the standard deviation does not capture the true risk.  A strategy that resembles picking up pennies in front of a steamroller – where small gains accumulate except in very rare periods when large losses occur – needs different analytical tools.  But at least we have gotten off to a start.  It is possible to evaluate the exotic beta opportunities presented by clone funds, using the tools we would use to analyze any investment.  The exotic beta we have analyzed is worthwhile, if not very exotic, and it is likely that other, carefully selected exotic beta opportunities are too. 

– Laurence B. Siegel

Be Sociable, Share!


  1. Walt French
    October 24, 2007 at 11:09 pm

    Of course, the real beauty of the long-short strategies such as Larry mentions is that their Sharpe Ratios can be had without giving up a dime of your equity premium. Unless their pattern of variabilities are identical (and they’re closer to orthogonal, at least compared to the broad equity market), a combination will have a higher Sharpe Ratio than either component — possibly, quite a bit higher.

Leave A Reply

← Alpha, once beatified, now "beta-fied" Paying Tribute to the Clones →