Jonathon Reeves, UNSW

Abstract

Ever since the inception of betas as a measure of systematic risk, forecast error in relation to this parameter has been of major concern to both academics and practitioners in finance. In order to alleviate forecast error, this paper compares a series of competing models to forecast beta. Realized measures of asset return covariance and variance are computed and applied to forecast beta following the advancements in methodology by Andersen, Bollerslev, Diebold and Wu (2005a and 2005b). This approach is compared with the constant beta model (the industry standard) and a variant, the random walk model. It is shown that an autoregressive model with two or three lags produces the lowest or close to the lowest error for quarterly stock beta forecasts. In general, the AR(3) model has a mean-squared-forecast-error half that of the constant beta model. This reduction in the forecast error is a dramatic improvement over the benchmark constant model.