Αναπαραγωγή των αποδόσεων των hedge funds

Abstract

In this thesis, they use a broad set of 21 traditional and option-like risk factors, significantly larger than the sets used in previous papers, in an attempt to capture as much systematic risk as possible. Our full sample analysis of HFR indices over the period from January 1995 to September 2007 yields similar systematic exposure and alphas as those reported in previous papers. In a further step, they use a rolling window version of the linear model in order to account for the flexible trading environment of hedge funds by allowing the beta coefficients to change. Motivated by the higher R2 values and the Sharpe ratios they get, they compare they cumulative performance of the replicating portfolios to that of the corresponding hedge funds indices. The comparison extends to a ten year long period, which is much longer than that used in previous works. This allows to examine whether replication is persistently successful. With the exception of the Short Bias strategy, they find that the replicating portfolios considerably under perform the original hedge funds indices. This is in contrast to Jaegner and Wagner (2005), who use a much shorter (two and a half year long) period and report encouraging results as regards to the ability of the replicating portfolios to provide similar cumulative returns to those of the corresponding indices.