盈餘之落後反應與非線性反應對Basu(1997)不對稱及時性衡量之影響

Abstract

This dissertation consists of two parts. The first study examines the manifestation of conservatism based on Basu (1997) measure under the price-lead-earning attribute. I construct a multi-period model to capture lagged earnings responses to good news and those to bad news to extend the Basu measure of asymmetric timeliness to a multi-period concept. Results show that differences between earnings responses to good news and those to bad news exist and persist for several periods. With varying incremental earnings responses to bad news, I argue that comparison of conservatism across firms based on Basu’s (1997) single period measure may not fully capture the effect of accounting conservatism. The second study examines the manifestation of conservatism based on Basu (1997) measure under the nonlinearity in earnings responses. I find that due to nonlinear earnings responses to news of large absolute values, a linear measure of conditional conservatism may not fully capture the effect of accounting conservatism. Both results indicate that prior evidence which suggests firms with different characteristics result in different extent of conservatism and the impacts of differences in conservatism across firms lead to differential economic consequences such as information quality, economic decisions and assets valuation needs to be re-examined. Considering both attributes can improve the effectiveness of comparing the extents of conservatism based on the asymmetric timeliness of earnings across firms.