Hedge fund replication

Many asset managers have recently produced products that are described as hedge fund replication product. Now there are a few methodologies that have been used in replication. Factor model replication, return distribution replication and position tracking. The most successful or at least the most widely utilized are factor models. Factor models are attractive as they are relatively easy to utilize and understand, and also unlike return distribution replication (Harry Kat) the successfulness of the product can be evaluated monthly.

As a very quick primer on factor model replication:-
These products work by evaluating the beta exposures to a number of factors. Running, normally an ordinary least squares regression between the index that is to be replicated and an array of alternative betas leads to the weightings of each market factor.
Those factor weighting are usually reevaluated each month and new hedge fund data becomes available. That’s basically it, however most companies have some slight variation.

Index selection / creation
One of the questions to be asked is “what should be replicated”? A standard hedge fund index, a fund of fund index or a customized index. Some companies create their own index, the argument being that this allows them to remove funds that they do not want exposure to, or to overweight exposure to fund that are “easier” to replicate.

Potential factor exposures
Generally a basket of factors is monitored. Now, one of the advantages of replication is the transparency and the increased liquidity, (as well as the lower fees). This means that even if it is clear that Ugandan equity is a significant factor, it is not going to be possible to provide a product with such an exposure. This is important, as many hedge fund replication products do not include exposure to factors such as credit, volatility, convertibles or some emerging market.

Weightings
The classic methodology is an ordinary least squares regression on rolling window. Normally something between 24 and 36 months. Some replication products utilize techniques such a kalman filter to be able to adapt quickly to any structural change in market structure. The argument being that a rolling window technique will take time to fully adjust to any significant change in weights.

These products have had a degree of success. However, there are a number of problems. Firstly, what are the advantages of these products? Well the two main headline attributes are liquidity and transparency. Both of these are attractive attributs to some degree, however one has to ask what level of liquidity and transparency investors really require. Investing in a hedge fund replication product has generally been considered in a core satellite approach. The replication product providing the strategic core whilst the investor can allocate tactically to specific hedge funds that they believe will outperform or provide alpha. Now, if an investor has chosen to invest like this, the additional liquidity is of little advantage (unless of cause the investor changes strategic allocations intra-month). The liquidity offered by most fund of hedge funds would be adequate to provide investor with the liquidity needed in a core satellite approach. Transparency is indeed important; however there are some important questions.
1) What do you do with this transparency? Full transparency of a Fund of hedge fund portfolio down to hedge funds underlying security positions is quite daunting. Most investors would use a factor model to break the returns down in order to find out where the risk lies. This can be done without knowledge of the underlying positions.
2) What makes an investor think that they are in any way better informed at evaluating the risks in the underlying positions? This is done by the hedge fund who will (such have) intimate knowledge of the position and the fund of hedge fund.

The disadvantage of these product revolve around the fact that because of there aims of providing liquidity, they are unable to provide exposure to (or replicate) all hedge fund strategies. None of the academic papers have been able to show a significant out performance over the index that they are trying to replicate. Most of these products are trying to be correlated to the index after fees, so by the time the replication products fees are included; the returns are not particularly attractive.

One of the way that I want to look utilizing these know market risk premiums is not to replicate hedge funds, uses these risk premiums to create an efficient portfolio.