Harry Kat: The FundCreator view on hedge fund replication and synthetic funds

Date: Monday, January 22, 2007
Author: Opalesque.com

 Harry Kat calls his model `hedge fund inspired`, `hedge fund resembling funds` rather than replication

Since the launch of FundCreator we have received various emails from people who, put simply, stated that “You guys are mad, there is no way you can replicate the best hedge funds”. Comments like these betray a serious misconception of what synthetic funds are really about. In this brief note I will attempt to correct this.

First of all, it should be noted that, despite their popularity with the media and conference organizers, the use of terms like “replication” and “clones” is somewhat unfortunate as such terminology creates expectations, which neither FundCreator nor factor model based funds live up to. FundCreator aims to provide returns with predefined statistical properties. Nothing more, nothing less. When the desired properties are derived from an existing fund.s or index.s track record we refer to this as “replication”. We do not, however, attempt to generate the same month-to-month returns, just returns with the same statistical properties as a given fund or index.

Factor models do try to generate identical month-to-month returns and as such the goal is indeed true replication or cloning. However, except for some extremely well diversified indices, the accuracy of factor model based strategies tends to be low, which means that in practice these strategies are not true replication strategies either. In sum, FundCreator and factor model based synthetic funds should preferably be referred to as “hedge fund inspired” or “hedge fund resembling” funds, as that is exactly what they are: funds with risk characteristics that resemble those of hedge funds.

What is driving the development of hedge fund inspired funds? We do not claim that there is no skill among hedge fund managers. What we do claim, however, is that many hedge fund (of funds) managers do not have enough skill to make up for the massive fees that they charge. It is investors in these managers that synthetic funds are aimed at. If you knew a manager that over the past 20 years had generated 20% per annum with 3% volatility, our advice would be to try to invest more with him, not to replace him by a synthetic fund.

Based on our own research, we estimate that in the current environment in 70-80% of the cases a hedge fund (of funds) manager.s contribution to the after-fee bottom line is zero or negative. Whenever this is the case, investors basically allow the manager to make a (very good) living using their money, without getting anything in return. Unless one is genuinely attracted to this form of charity, in these cases it is worth replacing the manager in question by a synthetic fund. A synthetic fund produces no pre-fee alpha, but it doesn’t cost a fortune to run either. In addition to generating hedge fund resembling returns, synthetic funds also come with a number of other benefits, such as great improvements in liquidity, transparency, capacity, etc., which makes the decision to go for the synthetic fund even easier.

So who should invest in synthetic funds? In the end, it all depends on how confident you are of your ability to find those truly skilled hedge fund (of funds) managers (and talk them into letting you invest with them). Investors who are confident they have enough skill to identify those managers that will more than make up for the fees they charge, should do so. Synthetic funds are not for them. If you take a more realistic view, however, and realise how good your manager selection skills will need to be, to be successful at this game, then synthetic funds are a very worthwhile alternative.

A final point concerns the difference between FundCreator and the more traditional factor model approach to synthetic fund creation, such as used by Merrill Lynch and Goldman Sachs for example. Both approaches aim to create synthetic funds with risk profiles similar to hedge funds. Both do so, however, in a completely different way. Factor model based synthetic funds aim to duplicate the various risk exposures of a fund or index. If successful, this will generate returns with properties similar to the target fund or index. If not, however, it is unclear what the properties of the fund return will be like. Unlike factor models, FundCreator skips over the actual return generating process and aims directly for the bottom line. A FundCreator based synthetic fund is explicitly designed to generate returns with particular statistical properties. With FundCreator therefore, you get the properties that you asked for, while with factor models it is very much a matter of wait and see. I know what I would prefer.