A Markov Regime-Switching ARMA Approach for Hedging Stock Indices.

Abstract

This study considers the hedging effectiveness of applying the N-state Markov regime-switching autoregressive moving-average (MRS-ARMA) model to the S&P-500 and FTSE-100 markets. The distinguishing feature of this study is to incorporate the observations of serially correlated stock returns into the hedging analysis. To resolve the problem of NT possible routes induced by the presence of MA parameters associated with the algorithm of Hamilton JD (1989) and a sample of size T, we propose an algorithm by combining the ideas of Hamilton JD (1989) and Gray SF (1996). We nd that the hedging performances of the three proposed MRS-MA(1) strategies herein are superior to their corresponding MRS counterparts considered in Alizadeh A and Nomikos N (2004) over the out-of-sample periods, even when we realistically track the transaction costs generated from rebalancing the hedged portfolios.