Research and publications

An Analysis of Hedge Fund Performance Using Loess Fit Regression

In this article, we analyze the returns distribution of Hedge Funds strategies, the average returns obtained over the past ten years and their correlation with a traditional portfolio. The aim is to identify the characteristics of each Hedge Fund investment strategy in order to be able to construct an optimal Hedge Fund portfolio for a Swiss pension fund. We will show that the classical linear correlation and the classical linear regression cannot be applied for Hedge Funds. Moreover, we will show that only three strategies, Convertible Arbitrage, Market Neutral and CTA, give diversification during market downturns. The techniques used are non-linear regressions and local correlations. A revisited version of this paper was published in the Spring 2002 issue of the Journal of Alternative Investments.

Author(s):

Laurent Favre, José-Antonio Galeano

Summary:

In this article, we analyze the returns distribution of Hedge Funds strategies, the average returns obtained over the past ten years and their correlation with a traditional portfolio. The aim is to identify the characteristics of each Hedge Fund investment strategy in order to be able to construct an optimal Hedge Fund portfolio for a Swiss pension fund. We will show that the classical linear correlation and the classical linear regression cannot be applied for Hedge Funds. Moreover, we will show that only three strategies, Convertible Arbitrage, Market Neutral and CTA, give diversification during market downturns. The techniques used are non-linear regressions and local correlations. A revisited version of this paper was published in the Spring 2002 issue of the Journal of Alternative Investments.

Register to download PDF

Register/Log in
Type : Working paper
Date : 17/03/2002
Keywords :

Asset Allocation