Hedge fund returns - Make them yourself

Discussion in 'Educational Resources' started by Helder, Sep 29, 2005.

  1. Helder


    For those interested in academic papers....

    We know that hedge fund returns usually are non-normally distributed and non-linearly related with market returns. These characteristics of hedge fund returns can affect traditional measures of performance, like the Sharpe Ratio.
    One alternative approach would be to evaluate a Hedge Fund by the cost of a replicating procedure, which would produce the same distribution of returns in the long-run trading liquid future contracts.

    There is a paper about this study in:

  2. nitro



    Thanks for the interesting article. When I downloaded it, there is a section after Vega Fund returns that says

    << Insert Figure 3 and 4 here >>

    Can you fix that?

  3. Nitro, the figures are at the end.
  4. Helder, thanks for sharing.

    A few thoughts...

    a) I'd be interested to see how you came up with the trading strategy to replicate the returns.

    b) I assume you tried to match net returns? I don't believe that is explicited stated.

    c) I'd be curious to see the costs to implement the replication trading strategy. ie. hardware, transaction, and development costs? [Edit: These costs are not accounted for in the article.]

    d) as you stated, the more HF money that flows in, the more the machine grinds to a halt. So by that extension, as each individual successfully implements the replication strategy, each succesive year will produce less profit potential because the individual will be reinvesting the previous years gains. So the machine -slowly- breaks?
  5. good luck helder ... gosh ... if this is some sort of breakthrough ... then I hope you and the prof.
    get some well deserved recognition and consulting work

    I assume this cannot work in some styles of HF
    that cannot be replicated or would not want to be

    ( i.e. LTCM )

  6. Helder


    Dear Waterloo and SethArb,
    Thanks for your comments.

    The idea is to create a kind of "exotic option". If you want to replicate the hedge fund distribution and you have some benchmark index (for example S&P500), you can create a payoff function f(S&P500) which produces the desired distribution. If you want to replicate not only the distribution of the hedge fund returns but also its dependence with the S&P500, then you need to trade another asset (for example future of Tbonds). Then you would have a bivariate exotic option f(S&P500, Tbonds). Each day you can adjust your position in future contracts. The trading costs are there, but are not so large in the case of futures.

    The basic idea is not to replicate month-to-month returns , but the same distribution of returns in the long run. So we don't need to know what is the strategy of the hedge fund, just its distribution of returns.

    After constructing the payoff function, you can price it using standard option pricing technology (for example Monte Carlo). Suppose that you invest 100 in the beginning of each month in the hedge fund. If the cost of the payoff function is smaller then 100, then the replication strategy is cheaper than to invest in the fund, and this fund is deemed as "inefficient"). If the cost is greater than 100, it is the other way around.

    If we try to replicate an efficient fund, we will get the same characteristics (standard deviation, skewness, kurtosis, correlation with index), but a lower expected return. So we are not claiming to replicate all hedge funds, but the inefficient ones. But there are many funds which appear to be efficient considering only the Sharpe Ratio or Alpha, for example, but which have low skewness and high kurtosis (so a high probability of some extreme losses). The cost of our payoff function will capture this "inefficiency", and if the fund is not efficiency due to this we can replicate it with the same or a higher expected return.
  7. This is very similar to something Fung and Hsieh did many years ago. They recreated CTA hedge fund returns by using lookback straddles.
  8. squeeze


    I wasn't convinced by the paper which was very short on real detail and seemed to involve over-fitting.
    There were parts that I thought showed the naivity of the academics who wrote this.

    For example:

    1. Since we trade three of the most liquid futures markets in the world, we are
    not confronted with liquidity or capacity problems. Want to add or take out a
    couple of hundred million? It won&#146;t be a problem.
    2. We don&#146;t have any transparency problems either. All we do is trade futures
    and anyone who wants to see what the portfolio looks like can do so at any
    point in time.
    3. Investors do not have to be afraid of style drift. Real-life managers can and do
    change their strategy over time, which changes the statistical properties of the
    fund return. Taking a fund&#146;s historical track record as our point of departure,
    we aim to extrapolate the past into the future, i.e. we aim to generate returns
    with exactly those properties that attracted investors to the fund in question in
    the first place.

    Even if it did work on a broader base of HF returns, it is still a technique for reproducing averageness and I don't see much value in this.
  9. nitro


    It would have great value to a fund of funds allocator.

  10. squeeze


    Yes, if they are setting out to achieve averageness then they will find it helpful.
    But a fund of funds allocator that is average does not have much value.
    #10     Dec 4, 2006