Stability of Risk Premia in the Italian Stock Market

In this work, we estimate the arbitrage pricing theory (APT) on the Italian Stock Market using the reduced‐rank regression technique recently propossed by Bekker et al. (1996). Due to its computational simplicity, this technique allows extensive empirical analysis of the properties of the estimator employed. In this work, we carry out an initial exploration of the cross‐sectional stability of the risk premia estimates in relation to the stocks' sample composition. We show that, by choosing an appropriately diversified sample, some acceptable degree of stability may be obtained. We also investigate, using the bootstrap method, the small sample properties of the estimator.