summary and critical discussion of the article: Basu, Sudipta (1997): The conservatism principl

4641 Words Dec 10th, 2014 19 Pages
List of abbreviations AND SYMBOLS II
1. Introduction 1
2. The conservatism principle and the asymmetric timeliness of earnings – a summary 2
2.1. The author’s motivation 2
2.2. The asymmetric sensitivity of earnings to returns 2
2.3. Earnings-return association versus cash flow-return association 5
2.4. The asymmetric persistence of earnings changes conditional on news 7
2.5. Conservatism and the asymmetric effect on the earnings response coeffcients 9
2.6 Further testing 11
3. International differences in the effects of asymmetric timeliness on earnings 11
3.1 Common-law versus code-law countries 11
3.1.1 Classification-differences of bad news as a bias to measured conservatism 13
4. Conservatism and its impact on
…show more content…
After that he comes to his final hypothesis: Basu here predicts a higher earnings response coefficient (ERC hereafter) for positive earnings changes than for negative earnings changes. Following his examinations of his four major hypotheses, he conducts some additional tests to back-up his prior findings. Hereupon he presents the reader with several alternative explanations of the rise of conservative practices to close his work with a critical view. In the end he concludes, summarizing his positively tested predictions and therefore punctuating the empirical evidence of conservatism in the research field of earnings quality.

2.2. The asymmetric sensitivity of earnings to returns
To illuminate the conservatism principle, Basu formulates and tests a set of predictions.
The first prediction argues that earnings reacts more sensitively and timely to publicly available bad news than to good news. News, in this context representing unexpected stock returns, therefore have a differentiated effect on earnings depending on its binary characteristic. To test his predictions he regresses annual earnings on current annual returns with the help of Beaver et al. (1980) ‘reverse’ regression. With earnings set as the dependent variable and returns as the independent one, the regression takes place in the following model:

Xi,t/Pi,t = α0 + α1DRi,t + β0Ri,t +

Related Documents