A study of long-lived asset impairment under U.S. GAAP and IFRS within the U.S. institutional environment

Abstract

This paper explores whether differences in accounting standards influence reporting behavior within the U.S. institutional environment where both IFRS and U.S. GAAP are used for reporting purposes. We focus on the accounting for impairment of long-lived assets, an area where significant differences exist between U.S. GAAP and IFRS. We identify all U.S.-listed firms who have recognized long-lived asset impairment losses during the 2004–2012 period. From these firms, we identify firms following IFRS, then develop a matched sample of U.S. GAAP firms, using a propensity score matching procedure. We examine the relation between impairment loss and unexpectedly high or low earnings in the year of impairment using a two-stage Heckman regression model, controlling for industry, country, year of write-down, and firm-level economic factors. We find that the association between impairment losses and unexpectedly high and low earnings is significantly greater for U.S. GAAP firms as compared to IFRS reporting firms, implying differences in accounting standards influence firm financial reporting. Our findings are robust to alternative measures of country level institutional factors and macro-economic variables, as well as inclusion of asset impairment reversals.