Abstract
Textbooks and much empirical research assume that firms' cost structure, typically operationalized as the ratio of fixed to total operating costs, is a valid proxy for estimating the Degree of Operating Leverage (DOL). This study shows that this assumption can be conceptually incorrect and empirically misleading. Using a set of accounting identities, I demonstrate that DOL can be expressed as a function of the Margin of Safety (MOS), and thus as the ratio of the Contribution Margin Ratio (CMR) to Return on Sales (ROS). Building on these mathematical relationships, I develop a new estimator of operating leverage that estimates time-varying contribution margin ratios from rolling regressions of EBIT on Sales and combines them with observed ROS to obtain firm-year DOL measures. I compare this "TV-CMR" estimator to Lev's (1974) cost-structure proxy and to elasticity-based estimators of Mandelker and Rhee (1984) and O'Brien and Vanderheiden (1987) using a simulation framework and a battery of robustness checks. The results show that Lev's method performs poorly, misclassifying most high-and low-risk firms. Elasticity-based estimators perform substantially better, while the proposed TV-CMR estimator achieves the highest accuracy, the lowest measurement error, and the strongest identification of operating risk across almost all scenarios studied. I conclude by discussing how the TV-CMR estimator can help researchers estimate break-even points and firms' margins of safety using archival data, opening new possibilities for empirical research on cost structure, profitability, and firms' risk exposure.