WSEAS Transactions on Business and Economics
Print ISSN: 1109-9526, E-ISSN: 2224-2899
Volume 13, 2016
Biases in Market-Based Measure of the Conditional Conservatism
Authors: Olga Fullana, Mariano González, Juan M. Nave, David Toscano
Abstract: In this paper we analyse the effects of outlier observations and endogeneity on the market-based measurement of conditional accounting conservatism. To address it, we apply a reverse engineering approach by using two alternative samples to estimate a measure of country-specific conditional conservatism – one including outliers and another without the multivariate influential observations identified. In the same way, on the sample without outliers we use two alternative estimation techniques – one affected by endogeneity and another specially designed to deal with the endogeneity problem. We apply this reverse engineering approach to the estimation of a comparative model of the conditional conservatism in order to analyse the effect of the International Financial Reporting Standards first adoption on the country-specific conditional conservatism. We report for both cases the two alternative results whose differences are only due to the outlier bias and the endogeneity bias, respectively. Our results prove the presence of these biases when outliers are not correctly identified and when the Ordinary Least Squares estimation technique is conducted. Moreover, these biases are large enough to result in misleading conclusions.
Keywords: Earnings conservatism, endogeneity, GMM-SYS, IFRS first adoption, market-based accounting research, multivariate outlier detection, OLS, panel data
Pages: 529-547WSEAS Transactions on Business and Economics, ISSN / E-ISSN: 1109-9526 / 2224-2899, Volume 13, 2016, Art. #49