Big Bath Accounting in Norway : empirical evidence on earnings management surrounding CEO turnovers in Norwegian firms

Abstract

In this thesis, we investigate the empirical relationship between CEO turnovers and big bath accounting by analyzing 5 979 Norwegian firms from 1999 to 2013. Using OLS and Fixed Effects regressions on panel data, we investigate if incoming CEOs make reporting choices that reduce earnings in their initial year, i.e. take a big bath. The big bath allows incoming CEOs to artificially enhance earnings in later periods, in addition to setting an initial low performance benchmark. We follow existing literature and analyze the turnovers’ effect on earnings in combination with revenue, write-downs and discretionary accruals. We distinguish between non-routine and routine turnovers, as well as divide our sample according to firm size. Our analysis suggest a clear correlation between low earnings and non-routine CEO turnovers, while it does not provide evidence of big bath accounting in routine turnovers. We also find indications of earnings reducing discretionary accruals and higher write-downs in the turnover year. As non-routine turnover often occurs in relation to bad firm performance, it is hard to determine if low earnings cause the turnover or if the turnover causes the low earnings. The issues with reversed causality make it hard to conclude with certainty that earnings are intentionally managed down in the turnover year rather than a result of bad firm performance. However, our finding of decreasing earnings combined with increasing revenue in the turnover year is consistent with big bath accounting, and calls for further investigation.