CONSIDERABLE EMPIRICAL RESEARCH HAS been directed to the relationship between financial and accounting variables and market based measures of risk.1 The results of this research indicate that some financial (accounting) variables are highly correlated with a market based measure of risk (beta) and are useful in the prediction of future risk. However, there has been relatively little research into the theoretical relationship between financial variables and market determined risk.Hamada [13,14] has researched the relationship between portfolio analysis and corporate finance. More specifically, he has shown that the systematic risk of a firm's common stock should be positively correlated with the firm's leverage. The analytical approach used by Hamada will be discussed more thoroughly below. The approach used in this paper to develop a relationship between systematic risk and leverage draws on the earlier work of Hamada. In the later paper, he adopted a different approach to arrive at a similar conclusion. Lev [19] has shown, using the approach adopted by Hamada [14], that a firm's operating leverage (the ratio of fixed to variable operating costs) is a variable affecting systematic risk.Pettit and Westerfield [30] assumed a discounted cash flow valuation model and separated the individual security return into cash flow and capitalization rate components. They proceeded to analytically develop a two factor model of beta.2 Although the approach which they used is promising, the model which they developed was not readily testable.The purpose of this paper is to provide a theoretical basis for empirical research into the relationship between systematic risk and financial (accounting) variables. Section II develops the assumptions and relationships of the capital asset pricing model. Systematic risk is defined as the /B parameter from this model. In the following sections we will show that, given the assumptions, there is a theoretical relationship between a firm's systematic risk and the firm's leverage and accounting beta. Also, we show that systematic risk is not theoretically related (directly) to the earnings variability, dividends, size or growth of a firm. * Assistant Professor, University of Oregon. The author wishes to acknowledge the helpful comments of William Beaver. 'Beaver, Kettler and Scholes [4], Pettit and Westerfield [30] and Rosenberg and McKibben [31] conducted tests involving many financial and accounting variables. Ball and Brown [2], Gonedes [11, 12] and Beaver and Manegold [5] were primarily concerned with the relationship between systematic risk and "accodnting beta." 2 See Sharpe [33] for a general theoretical model for this approach. 617 618 The Journal of Finance