“Kat’s Meow” – An occasional guest column for AllAboutAlpha
By: Professor Harry Kat, Cass Business School, City University (London)
Judging by the number of investment banks and asset managers jumping on the bandwagon, the media attention and the number of forthcoming conferences on the subject, hedge fund replication has definitely arrived in 2007. However, since it is all a bit new and all a bit much so suddenly, many investors seem to be somewhat confused and unable to ask the right questions. This brief note is meant to provide some guidance in this area.
A little bit of history first. Initially, hedge funds were sold on the promise of superior performance – the story being that hedge fund managers’ long experience and proven investment skills were a virtual guarantee for superior returns. Especially high net worth investors were sensitive to these arguments and fuelled much of the early growth of the industry. Starting in the late 1990s, hedge fund performance took a turn for the worst, however, with every next year being worse than the year before. The story changed accordingly. No longer were hedge funds sold purely on the promise of superior performance, but more and more on the basis of a diversification argument, pointing at hedge funds’ relatively low correlation with stocks and bonds and the beneficial effects on risk and return from including hedge funds in the traditional investment portfolio. Especially institutional investors proved sensitive to these arguments and started to pour large amounts of money into the sector, in the process making up for the outflow of some of the early private investors. Nowadays, there aren’t many serious investors left who still expect hedge funds to provide them with superior returns. The hedge fund game these days is all about diversification.
Now let’s have a closer look at hedge fund replication. What are the attractions of replicated hedge funds? The main attraction is that there are no expensive managers to pay. Since the bulk of the hedge fund and fund of funds managers (we estimate 80%) are unable to make up for the fees that they charge, that is a very good thing. In addition, replicas solve a range of other problems as well, including liquidity, transparency and capacity problems, just to name a few. Altogether, this makes for a very attractive package: hedge fund returns at lower cost and without the usual hassle. The above almost sounds too good to be true, and unfortunately in many cases it is. Research has shown that the models used by the big name hedge fund replicators (see later) are unable to accurately replicate individual hedge funds as well as most hedge fund indices.
The only indices that can be replicated with reasonably accuracy are the indices that contain so many different funds that there is nothing hedge fund-like or â€˜alternative’ about them. Since replication accuracy is on top of their wish list, these are the indices that the main hedge fund replicators aim to replicate.
At the time of writing only one replication product is really live: the Merrill Lynch Factor Index. Goldman Sachs announced a very similar product, to be called the Goldman Sachs Absolute Return Tracker Index, last year, but hasn’t actually launched it yet. Recently, JP Morgan announced that in the second quarter of this year it intends to launch what will be called the JP Morgan Alternative Beta Index. State Street Global Advisors as well as some others also look set to launch hedge fund replication products.
In what follows I will concentrate on the Merrill Lynch Factor Index. Not because I have an axe to grind with Merrill, but simply because there is not yet enough information about the forthcoming Goldman, JP Morgan and State Street products. It is, however, quite likely that these products will be very similar to the Merrill product (although the designers and marketers of these products will of course claim otherwise).
The ML Factor Index aims to replicate the HFRI Composite index, a basket of 1800 hedge funds following all kinds of strategies. Since it is extremely diversified, the peculiarities of the various strategies simply diversify away. The result is an index with mainly traditional risk exposures and very little is â€˜alternative’ about it. This is exactly why this index can be accurately replicated. It is, however, also exactly why investors should not be interested in this index. They are already exposed to all these risks and adding more of the same to a portfolio doesn’t provide any diversification benefits.
To illustrate the above point, Figure 1 shows the evolution of the S&P 500, the HFRI Composite index, the ML Factor Index and the HFRI Fund of Funds index since January 2003. Comparing the plots for the three hedge fund indices with that for the S&P 500 we see that the differences are only small. None of the three hedge fund indices seems to contain many â€˜alternative’ ingredients. To a large extent, they simply mimic the S&P 500. This is also evident from the correlation coefficients. Over the period studied, both the HFRI Composite and the ML Factor Index had a correlation of 0.76 with the S&P 500. Instead of buying into the ML Factor Index for 100bps you might as well buy S&P 500 futures for 1bp and add the 99bps difference to your expected return.
As Figure 2 shows, the forthcoming JP Morgan product suffers from the same deficiency. Similar to the ML Factor Index, the JP Morgan Alternative Beta Index tends to move in virtual lock step with the HFRI Composite and Fund of Funds indices.
A second indication of the lack of â€˜alternative’ elements in the current replication products can be obtained by looking at the composition of the ML Factor Index portfolio, which is shown in Table 1. From Table 1 we see how traditional the ML Factor Index portfolio really is. It contains the S&P 500, USD, MSCI EAFE, MSCI Emerging Markets, Russell 2000 and 1-month Libor. Nothing alternative about it. The result is evident: a high correlation with traditional asset classes and therefore little or no diversification benefits. This clearly shows what focussing on replication accuracy does: it leads one to replicate those indices that are most traditional in nature. Unfortunately, these indices are also the least interesting from a diversification perspective.
So maybe it is better to forget about replication accuracy altogether and just invest in a replication strategy of which we know upfront that it doesn’t work? Doing so, we at least have a chance of ending up with something that adds value in a portfolio context. Although this is true, the problem with such a fatalistic approach is that it leaves an awful lot to chance. With an index that can be replicated we can study the track record of the index to get an idea of the kind of returns we can expect. With an index that can’t be replicated, we don’t have that luxury. In that case we can only look at the backtested strategy results, which will necessarily cover a significantly shorter period of time than the track record of the index that is being replicated.
Another area where hedge fund replication products fail is the stability of the risk profile generated. Since the prime focus is on replication accuracy instead of a stable bottom-line risk profile, the risk profiles of the resulting portfolios can be quite unstable. Figure 3 for example, shows the weights of the various assets making up the ML Factor Index portfolio. From this graph it is obvious that the composition, and thereby the risk profile, of the replicating portfolio is capable of changing very substantially over time. For example, in February 2002 the portfolio was 20% long S&P 500, but in July 2005 it was 20% short S&P 500.
What does the future hold for hedge fund replication? Given the global distribution and marketing power of Merrill, Goldman, JP Morgan and State Street, there is no doubt that they will manage to raise serious money with these hedge fund replication products. Investors buying into these products with the aim to diversify their traditional portfolios, however, are likely to be disappointed as there is very little â€˜alternative’ about these products, resulting in very high correlation with traditional asset classes and correspondingly low diversification benefits. Caveat Emptor!
– Harry M. Kat, March 3, 2007
Professor Harry Kat is an occasional contributor to AllAboutAlpha.com. The opinions expressed herein are those of Professor Kat and not necessarily those of Alpha Male or AllAboutAlpha.com.