CEO Incentives And Future Earnings

So far, the domestic scholars are most focus on study about incrementalinformation of historical accounting data, but pay less attention to study onforecasting information. As the development of our country’s capital market, thedemands which investors ask for forecasting information are more and more. On thestudy about incremental information of earnings forecasting, accounting and financescholars tend to write about predictions of future earnings and effects of executiveincentives on future performance as if they were two distinct subjects. However, theconnection between earnings prediction and performance consequences of executiveincentives is fundamental: how successfully a company solves agency problemsthrough incentive contracts effectively determines management behavior(e.g.,investment and financing decision, cost and revenue management, and riskmanagement), which ultimately determines earnings and cash flows that companiesgenerate for shareholders. Since there are few domestic researches on this fieldsystematically, it gives us much more space for our examination and test to fill thisgap in the literature.Based on this, after the implementation of "Administration Division for theListing Company Equity Incentive Measures", we select A-share listed companies assample, which announced to implement equity incentive plan from1stJanuary,2006to31stDecember,2009. We examine the sample from four aspects as follows:whether CEO incentives have effects on earnings persistence; whether CEOincentives provide incremental information for earnings prediction; whether theincremental predictive power of CEO incentives is associated with earningsmanagement; whether financial analysts incorporate CEO incentives into earningsprediction in order to improve the accuracy of forecasting results.On the basis of related theories and literature, referring to the research results ofdomestic and oversea scholars, we find that CEO incentive is an importantshareholder value driver through logistic analysis. Following the research methods ofearnings prediction in domestic and oversea scholars’ study, we propose thecorresponding hypotheses, and then elaborate model selection and variables calculation. After that, we do a systematic research on sample from variables selection,descriptive statistics, correlations analysis and regression analysis. Through analyzingthe results of regressions, we make judgments on the corresponding hypotheses. Atthe same time, we give explanations on the reasons which lead to these results.Our research finds when we sort each year’s observations into quintiles based onCEO incentives, the coefficients of current earnings and next year’s earnings increasewith CEO increase. The result holds for forecasting horizons ranging from one year tothree years. We add CEO incentives as an additional predictor in the earningsprediction model, and find that this variable obtains a positive and highly significantcoefficient. The result remains when we expand the prediction model to control forsize, growth, risk and other correlated variables. CEO incentives are significantlypositively related to future cash flow from operations and non-discretionary accruals,but insignificantly related to future discretionary accruals. The coefficient of CEOincentives and analyst forecast on current year’s earnings is statistically insignificant,but the CEO incentives variable is positively significantly related to forecast error forone-year-ahead earnings.Therefore, we find that companies with higher CEO equity incentives havehigher earnings persistence and that in these companies current year’s earnings aremore predictive of future earnings. In an earnings prediction setting, CEO incentivesare shown to provide information about future earnings that is incremental to currentearnings and earnings components. The predictive power of CEO incentives for futureearnings remains robust after we control for a number of correlated variables. Further,we find that CEO incentives are predictive of “real” future earnings, as represented byoperating cash flow and non-discretionary accruals, but not of future discretionaryaccruals. Interestingly, we find that financial analysts do not incorporate informationfrom executive incentives in their forecasts of future earnings. Including CEOincentives can potentially improve analyst forecasts.