An improvement on Basu-based conservatism measures

Abstract

This study puts forward an alternative Basu-based measure for measuring conditional accounting conservatism. While the effectiveness of the Basu model persists to be an open question, Basu-based measures have been widely employed to measure conditional conservatism. Therefore, it is worthwhile trying to find an improvement for Basu-based measures to increase their effectiveness. Khan and Watt first improve the original model by constructing a measure for conditional accounting solely in terms of firm-specific news but its advantage over the Basu model is deteriorated by neglected firm-specific factors. The study follows up on this initial improvement and suggests decomposing stock returns used to represent general economic news in the Basu mode into two component stock returns representing firm-specific and market-wide news respectively. Their estimation using the linear regression of calculating beta coefficients of CAPM results in a new potential method of conditional conservatism measurement. This intrigues several future studies on Basu-based measures.