Implications of Data Restrictions on Performance Measurement and Tests of Rational Pricing

Earnings quality in UK private firms: comparative loss recognition timeliness

“…Scaling problems arise from near-zero observations in the scaling variables. 31 Kothari et al (1999) demonstrate that truncating skewed distributions such as earnings can bias test statistics, but we believe that there are sufficient errors (as distinct from genuinely extreme observations) in the data to warrant it. We repeat the analyses with 0.5% and 2% exclusion criteria, and our conclusions are unaltered.…”

Limited attention, information disclosure, and financial reporting

“…As emphasized by Kothari (2001), some findings of apparent market inefficiency in existing literature may be artifacts of flaws in methodology. Potential sources of spurious effects include problems in risk measurement (see, e.g., Franks et al, 1991), data issues such as survivorship bias, and the effects of skewness of financial variables (Kothari et al, 2002). This said, there appears to be a set of 'anomalies' (patterns of return predictability) that have so far proven stubbornly hard to explain from an efficient markets perspective, including post-earnings announcement drift, the accruals anomaly, and stock return momentum.…”

The demand for accounting information: young NASDAQ listings versus S&P 500 NYSE listings

“…8 Many prior studies winsorize all variables, dropping observations in the upper and lower tails of the distribution at some arbitrary cutoff points. Because I am concerned that firms in the extreme tails might have the characteristics that I seek to study (Kothari et al 2001), at the data diagnostic stage, I employ two outlier analyses. I employ both the Cook D ðdistance D i ¼ 3ðb y ðIÞ À b y j Þ 2 =ps 2 Þ where p is the number of parameters) and the Rstudent method to eliminate outliers (Welsch 1980).…”