감사보수와 감사시간이 재무분석가의 이익예측 오차에 미치는 영향

This study examines the association between audit fee and audit hour and properties of analysts' earnings forecasts. Specifically, we use abnormal audit fees and abnormal audit hours which started to be disclosed from year 2000. The abnormal audit fees imply the audit fees paid to auditors more than expected normal fees. Similarly, the abnormal audit hours imply the audit hours spend by auditors in order to audit the client firm more than expected normal hours. We focus on the accuracy (or error) of analysts' earnings forecasts among the properties of analysts' earnings forecasts because the accuracy is the greatest concern among the market participants. Auditors can provide high-quality audit service by exerting more audit efforts (measured by audit fee and hours). The higher quality audit service restricts opportunistic earnings management by managers. Thus, the more audit efforts may represent the higher quality of audit service. As a result, it is possible that analysts use high-quality financial information to generate earnings forecasts, which result in less forecast error. In contrast, the more audit efforts could represent more audit risk or uncertainty related to the audit process. In such a case, more audit effort may represent uncertain future of the firm. As a result, it is expected that analysts' earnings forecast error increases. Thus, it is an empirical issue to examine if the increased audit efforts are linked to higher or lower forecast error. Because analysts analyze the audited financial statements in order to issue earnings forecasts, if the financial statements are scrutinized by high (low) quality audit service, the financial statement would be more (less) reliable and could provide better (worse) picture of future prospect of the firm. Thus, when analysts issue forecasts, if they use the financial statements audited by high (low) quality auditors, it is expected that the forecasts issued by the analysts contain less (more) forecast errors. To examine this prediction, we use on average 675 firm-year observations collected over 2000 to 2004 period. Firms belong to financial industry are excluded. These samples are listed companies in Korean Stock Exchange. We collected the analysts' forecasts collected from Fn-Data Guide Pro database and use forecasts issued on March, June, September, and December of each year. The empirical findings are summarized as follows. First, the error of forecasts issued on June, September, and December are significantly lower (higher) for the firms with abnormally higher (lower) fees. In addition, firms with abnormal fees belong to top 10 percent have smaller forecast errors. Second, for forecasts issued on March and June, the firms with abnormally high (low) audit hours show greater (smaller) forecast error. In addition, firms with abnormal hours belong to top 10 percent have higher forecast errors. In sum, the results imply that abnormal fee has positive association with forecast accuracy, whereas abnormal audit hour has negative association with forecast accuracy. It implies that audit fee increases as audit quality increases. In contrast, the audit hour increases for risky firms or firms with greater future uncertainty. Auditors may spend more time to audit such firms but the financial statements of those firms are not as accurate or reliable as those of other firms. Our findings should also be useful to investor who use these forecasts to value firms, as it suggests that the earnings expectations of firms that employ a high audit quality and low audit risk to be more accurate.