The Empirical Research On Impacts Of Investor Sentiments On Earnings Management

Earnings Management is not a simple accounting behavior. It also involves management and economic issues. Its nature,however, is the problem of corporate governance.So far,most of the studies on earnings management are based on the premise of rationality. However,Investors are not always rational in the reality of capital market. There often exists irrational behavior due to the investors cognitive biases,which called investor Sentiment.Recent work in finance has studied the impact of such sentiment on corporate actions,such as equity issues,dividend payouts,investment and acquisitions. However,there is scant empirical evidence on how managers respond strategically to investor sentiment via corporate reporting decisions.The nascent accounting literature on investor sentiment has examined the influence of investor sentiment on management forecast disclosures and disclosure of proforma earnings metrics. Surprisingly,there is little evidence on the effect of investor sentiment on earnings management,given the wide attention in the literature to managers’attempts at interfering with the financial reporting process.The impact of investor sentiment on earnings management is studied based on627samples from firms that are listed on the Shanghai a-share for the period2008-2013. We use four kinds of discretionary accruals models to estimate normal accruals and discretionary accruals which is a proxy for earnings management.The investor sentiment measure is the consumer confidence index.We find that there is a positive association between earnings management and sentiment,that is,managers respond to market-wide sentiment by boosting discretionary accruals in high-sentiment periods and reporting conservatively in low-sentiment periods. There is a Significant relationships between the likelihood of positive (negative) discretionary accruals and Investor Sentiment. The likelihood of managing earnings upwards increases with the level of investor sentiment,and the likelihood of taking negative discretionary accruals decreases with investor sentiment. We also find that managers’incentive to report positive discretionary accruals to avoid negative earnings surprises increases with sentiment,that is,the probabilities of positive discretionary accruals conditional on firms’meeting or beating analysts’ forecasts increases in periods of high sentiment. We probe further to understand what drives the documented positive association between earnings management and investor sentiment. If the time variation in discretionary accruals is at least partly driven by managers’responding to investor sentiment,then we would expect firms whose stock returns co-move more with changes in investor sentiment (or with higher sentiment betas)to boost earnings by a larger extent. Consistent with this intuition,we find that the positive association between discretionary accruals is stronger for firms with greater investor sentiment betas.The study makes several contributions to the literature. First,I extend the existing research on earnings management in response to investor sentiment by investigating whether managers strategically vary their accrual reporting choices in periods of high and low investor sentiment,and whether investor sentiment affects managers’ involvement in earnings surprise games. Second,the study adds to the limited literature on revenue management by documenting the time-varying use of discretionary revenues as an earnings management tool. Third,the results question the common assumption that reported earnings surprises cause stock price changes,but not vice versa.We show that the sensitivity of a firm’s stock returns to market-wide investor sentiment might affect future earnings for that firm via earnings management.