Autoregressive multifactor APT model for U.S. Equity Markets

Arbitrage Pricing Theory is a one period asset pricing model used to predict equity returns based on a multivariate linear regression. We choose three sets of factors – Market specific, firm specific, and an autoregressive return term to explain returns on twenty U.S. stocks, using monthly data over the period 2000-2005. Our findings indicate that, apart from the CAPM beta factor, at least five other factors are significant in determining time series and cross sectional variations in returns. The times series regression establishes factor loadings and the cross sectional regression gives the risk premiums associated with these factors.