Narrow Results

This paper investigates the effects of SFAS 133 on earnings volatility, earnings predictability, and equity volatility in bank holding companies (BHCs). In contrast to large BHCs' assertion prior to the adoption of SFAS 133, the three income-affecting portions (i.e. hedge ineffectiveness gains/losses, gains/losses excluded in the assessment of effectiveness, and effects from canceled forecasted transactions once designated as cash flow hedge) did not increase earnings volatility in the top 30 BHCs for the first 12 quarters after adoption. Also, the three income-affecting portions did not deteriorate analysts' forecast performance. In addition, there is no evidence that volatility of stockholders' equity significantly increases due to SFAS 133. Further investigation of notes to financial statements reveals that some BHCs adjusted their usage of derivatives prior to SFAS 133 to mitigate the impact of SFAS 133 on earnings volatility.

We argue that coinsurance among a firm’s business units changes the properties of reported earnings through less volatile operations (financial synergies) and fewer estimation errors in the accrual process (accounting synergies). Consistent with a coinsurance effect, we find that diversified firms have on average higher earnings quality compared to industry-matched portfolios of focused firms. Specifically, diversification leads to more predictable earnings, superior mapping of accruals to cash flows, and lower absolute abnormal accruals. In addition, we find higher earnings quality for diversified firms with less correlated segment earnings and that the coinsurance effect is stronger for firms that operate in more volatile and uncertain environments. We contribute by identifying the coinsurance effect of diversification as a new determinant of earnings quality. Our findings complement prior literature on agency-related disadvantages of diversification for earnings quality by highlighting coinsurance related benefits of diversification for earnings quality.