Inventory Accounting Method and Earnings-Price Ratios

Abstract

Lee (1988) finds LIFO firms have higher earnings-price (EP) ratios than non-LIFO firms despite the income-reducing effects of LIFO, a result contrary to economic intuition and which Lee describes as a ?puzzle.? This paper attempts to resolve the puzzle identified by Lee using an approach that involves improved measurement of those variables which theory suggests should determine EP ratios. We examine EP ratios for a sample of firms for the years 1982 through 1993. Although EP ratios are consistently higher for LIFO firms during this period, we find the difference is reduced substantially when firms with losses are excluded. When the regression model used by Lee (1988) is estimated, the coefficient on the LIFO indicator variable is positive for both the full sample of firms and for the subsample with loss firms excluded. A regression model is then estimated which includes improved proxies for risk and expected growth. The improved proxies are analysts' expectations of future growth rather than realized growth, beta computed using a procedure designed to reduce measurement error rather than the usual OLS beta, and leverage as a supplemental risk measure. Further, we control for expected earnings changes, since transitory earnings shocks that are not expected to persist in future earnings can affect EP ratios. Results of this model indicate that EP ratios are positively related to risk and negatively related to expected earnings growth. After controlling for differences in risk, expected earnings growth, and transitory earnings, we find that EP ratios for LIFO firms are actually lower than those of non-LIFO firms, a result consistent with economic intuition and the result expected by Lee (1988).