Differential Earnings Management to Avoid Earnings Declines And Losses Across Publicly and Privately-Held Banks

There is a documented empirical regularity that publicly-held firms report fewer small losses and fewer small declines in earnings than expected. This paper betters our understanding of this observed phenomenon by testing for this regularity on a sample of public and private banks during 1987-1998. We argue that the incentives to manage the earnings stream to achieve simple benchmarks such as positive earnings and increases in earnings should be higher for publicly-held banks. Consistent with our predictions we find that private banks with earnings near zero are significantly more likely to report losses than public banks with earnings near zero even after controlling for bank size, cash flows, and differences in the types of loans written. Similarly, private banks with changes in earnings that are near zero are significantly more likely than public banks to report a decline in earnings than an increase. In addition, we also find that the length of the string of consecutive earnings increases, even after controlling for the length of the string of consecutive increases in cash flows, is greater for public banks. These three tests all suggest that public banks face a greater incentive than private banks to achieve simple earnings benchmarks