Estimating hedge fund risk factor exposures

In this paper, we propose a novel approach for decomposing financial market returns into observable risk factors and idiosyncratic risk. We utilize a vector stochastic-volatility model to extract the potentially time-varying exposure of low frequency hedge fund performance on high frequency data. By making use of a particle filter with Rao-Blackwellization, we reduce the dimension of the space where we generate particles, which results in more accurate estimates of the posterior and predictive distributions of the unknowns. This approach can be used for analyzing hedge fund performance and their advertised strategies as well as in forensic risk-management. The latter is a critical need given the generally low transparency of the hedge fund industry. We illustrate our results using simulations and real hedge fund and S&P 500 index data from 1994-2011.