Alternative Beta: Old Wine in New Bottles

By: Guest Contributor to AllAboutAlpha, Professor Harry Kat, Cass Business School, City University (London)
February 4, 2007 – Currently, there is a lot of talk among investment practitioners about “Alternative Beta”, with the latter roughly defined as the return derived from exposure to “alternative” risk sources. The way in which Alternative Beta is being presented in the media and at conferences suggests that it is the next big thing since electricity and sliced bread. The truth, however, is that Alternative Beta has always been around, although under a different name. Below I will explain this comment further.

“Alpha-beta thinking” is based on factor models. A factor model aims to link the return on a fund to a number of risk factors. Given the relevant risk factors and the fund’s track record, the fund’s sensitivity to the risk factors is estimated by some form of regression analysis. The resulting sensitivity estimates are subsequently referred to as the fund’s “betas” and the part of the average return that cannot be explained by these betas as the fund’s “alpha”.

When it comes to risk factors it is now customary to distinguish between “traditional” and “alternative” risk factors, the sensitivities to which are referred to as “traditional betas” and “alternative betas”. Although it all seems to make perfect sense, it is at this point that the confusion starts, as it is not at all clear what is what. Stock market and bond market risk clearly are traditional risks. But so are FX risk, credit risk, commodity risk, liquidity risk, etc. one might argue. Investors have been taking these risks for centuries, so why would they suddenly be “alternative” instead of “traditional”. This strongly suggests that the distinction between “traditional” and “alternative” is nothing more than artificial; just another one of those pseudoscientific marketing gimmicks underwritten by an industry that is desperate to hide the fact that most players are offering more or less the same product.

To put the above hypothesis to the test, let’s have a look at what risk factors are popular with the Alternative Beta crowd these days and see if there is anything in there that is truly alternative in the sense that traditional investors haven’t seen it before very often. In a recent study of hedge fund risk factor exposure, Lo and Hasanhodzic (2006) used the following factors: (1) the US Dollar Index, (2) the Lehman Corporate AA Intermediate Bond Index, (3) the spread between the Lehman Corporate AA Intermediate Bond Index and the Lehman Treasury Index, (4) the S&P 500, and (5) the GSCI.  [Initially Lo and Hasanhodzic (2006) also include the CBOE Volatility Index as a risk factor. Going forward, however, they drop this index “as it is not easily realized with liquid instruments” (p. 20)]. In another recent study, Jaeger and Wagner (2005) use different factors for different hedge fund strategies. Among the factors used, however, we again find a large number of very traditional indices, such as the S&P 500, the CSFB High Yield Bond Index, the S&P 600 Small Cap Index, the GSCI, the Lehman World Government Bond Index, and the JP Morgan Emerging Market Global Bond Index.

Don’t these risk factors sound but all too familiar? Aren’t they exactly the indicators that many traditional investors tend to look at every day? This confirms the above hypothesis. The bulk of the returns in so-called “alternative” factor models is derived from old-fashioned, traditional risk factors, with truly alternative risk factors only playing a supporting role at best.

What does this imply for factor model based hedge fund replication? It means that the resulting replicating portfolios will primarily be made up of traditional assets, managed in a manner not too different from what traditional managers do. One might argue that this is exactly what replication is about: to generate hedge fund returns by trading traditional asset classes only. However, that begs the question why it is necessary to talk about alternative beta all the time if the aim is to produce a mainly traditional portfolio? Could it have something to do with marketing?

The above observation is quite useful in explaining why factor models are only able to replicate the most diversified indices, like the HFRI Composite Index for example. It is only in these indices that most of the truly alternative risk is diversified away. In other words, it is only in this kind of indices that returns are driven largely by systematic, traditional factors. It also explains product providers’ eagerness to offer factor model based replication products: it allows them to charge 100bps for what is very much an old-fashioned traditional portfolio!

So is it really necessary to think about investments in terms of alphas, betas and risk factors? Of course not. In the old days one would simply look at the securities making up a portfolio, instead of the risk factors thought to be driving the portfolio’s returns. This form of “asset class thinking” may not have the pseudo-scientific intellectual appeal that alpha-beta thinking has, but that doesn’t necessarily make it less effective or essentially different. In the old days somebody that expected credit spreads to narrow would say: “Let’s buy some more corporate bonds”. Nowadays that same person is likely to say: “Let’s increase our credit risk beta”, and then buy some more corporate bonds. Now how different is that?

So what do we conclude from this? Basically, that it is a word game. Alternative betas are largely traditional betas called differently. Alpha-beta thinking is a fancy form of asset class thinking and factor model based hedge fund replicating portfolios are a lot more traditional than many suspect. Alternative Beta? I wouldn’t lose sleep over it. You know most of it all already.

– Harry M. Kat, February 4, 2007

References

Hasanhodzic, J. and A. Lo, Can Hedge Fund Returns Be Replicated? The Linear Case, Working Paper MIT Laboratory for Financial Engineering, 2006. (ed: link to AllAboutAlpha posting on this paper)

Jaeger, L. and C. Wagner, Factor Modelling and Benchmarking of Hedge Funds: Can Passive Investments in Hedge Fund Strategies Deliver?, Journal of Alternative Investments, Winter 2005, pp. 9-36. (ed: link to AllAboutAlpha posting on this paper)

Be Sociable, Share!

2 Comments

Leave A Reply

← AlphaSwiss Group to Harry Kat: Not So Fast... Bill Fung's Alpha/Beta Slides from Last Year's GAIM Conference →