Modeling Hedge Fund Returns Using the Kalman Filter: An Errors-in-Variables Perspective

Many studies on Hedge fund returns are static, resorting frequently to a multifactor model composed of mimicking portfolios or benchmarks that account for the style of the different strategies followed by hedge funds. Yet hedge fund strategies are essentially dynamic. As conditional models appear that permit to explain the dynamic strategies of portfolio managers, we study in this article how the portfolio manager alphas and betas react to financial market variables like market risk premium, interest rate, and market volatility. In this paper, we thus propose an innovative application of the Kalman Filter to dynamize these important risk factors and to account for measurement errors which create bias in the estimations process of these important parameters.