The association between arbitrage pricing theory risk measures and traditional accounting variables

According to the Arbitrage Pricing Theory (APT), actual security returns depend on a variety of pervasive economic and financial risk factors; as well as firm or industry specific influences. The sensitivity of an asset\u27s returns to unanticipated changes in the pervasive risk factors reflects the security\u27s measure of systematic risk. In equilibrium, the expected security return is a linear function of the sensitivities of actual security returns to unanticipated changes in the pervasive risk factors. The APT does not specify the number or the nature of the pervasive risk factors. Factor analysis of stock returns can be used to determine sensitivities of individual securities to pervasive risk factors without having to identify these risk factors. In this dissertation we empirically tested the following question: \u27Can we use traditional accounting risk measures from the current period to explain cross-sectional variations of the APT risk measures (sensitivities) in the next period?\u27. The empirical analysis was carried out using a sample of manufacturing companies from the Compustat data tapes. The study covered two time periods: 1983-1986 and 1988-1991. The dependent variables were the APT risk measures, derived from a principal factor analysis of daily stock returns. The set of independent variables was an extensive list of traditional accounting risk measures associated with a firm\u27s operating and financial activities. The accounting risk measures used in this study represented the firm\u27s liquidity, debt management, profitability and efficiency, business risk, and market value ratios, as well as the size of the company. Relying on predictive correlation analysis and multiple regression analysis, an association was established between a firm\u27s basic operating and financial characteristics, and the variations of the APT risk measures across a sample of individual companies as well as portfolios of manufacturing companies. Several variables measuring various characteristics of firms were found to be significant in explaining each one of the APT risk measures. Traditional accounting variables, can be used to explain the APT risk measures across different sample specifications, as well as across different time periods. In this respect variables showing the greatest promise are measures of size, business risk, financial risk, and market power