재무분석가의 장기이익예측 특성 및 결정요인

This paper examines properties of analysts’ long-term earnings forecasts and the determinants of long-term forecast issuance in Korea. Although long-term earnings forecasts are one of the most important inputs in calculating the fundamental value of a firm, prior studies focus primarily on issues related to short-term earnings forecasts such as biases in analysts’ short-term forecasts and their incentives to provide such biased forecasts and shed little light on why they issue long-term forecasts and factors that determine long-term forecasts bias. This paper fills that void. We expect that analysts who issue only long-term forecasts are different from those who issue short-term forecasts and there will be some fundamental differences in their characteristics. It is a natural conjecture that long-term forecasts entail more complex tasks that require better firm-specific and domain-specific knowledge and they are more likely to be made by more capable analysts. However, the accuracy of forecasts made by more capable analysts could be lower if their forecasts are opportunistically biased due to their monetary incentives such as sales commissions and other investment banking related business. Assuming that analysts who issue only long-term forecasts are less subject to those incentives, we hypothesize that analysts issuing only long-term forecasts provide more accurate long-term forecasts and less optimistically biased. Concerning the determinants of long-term only forecasts, we hypothesize that they are more concentrated on small-sized growing firms that their stock prices are currently undervalued than their book values but relatively more profitable. Using analysts’ earnings forecasts for non-financial firms listed in Korea Stock Exchange (KSE) and KOSDAQ for the period from 2000 till 2007, we find the following results. First, long-term only forecasts are more accurate than long-term forecasts with short-term forecasts. In addition, long-term forecasts are primarily issued for growth stocks with relatively high ROAs but low market-to-book ratio. This result suggests that long-term only forecasts are issued in order to satisfy the information demand by less informed investors with a long-term investment horizon. Second, long-term only forecasts are issued by more capable top analysts, who are relatively less subject to financial incentives since they have higher reputation. Overall, the result is consistent with the fact that information demand is more important in generating long-term earnings information and reputable analysts play an important role in disseminating more accurate long-term earnings information that helps make information environment better. This study has several contributions to the literature. First, this is the first study that investigates the properties and determinants of long-term earnings forecasts. Prior studies investigate only analysts’ short-term earnings forecasts and their incentives and not many studies examine characteristics of long-term earnings forecasts, especially those forecasts without short-term forecasts. By improving our understanding of the properties of long-term earnings forecasts, the result of this paper provides a better way to get more accurate fundamental value of a firm and, ultimately, helps make more efficient resource allocations. Second, we find an important, proactive role of investors who demand long-term financial information by showing that investor-initiated information demand, not incentive-driven analysts’ supply of long-term earnings information, helps make better information environments. Third, the result of this paper confirms the findings of prior studies on analysts’ incentives in a sense that the quality of analysts’ provided information is better when there are less financial incentives, especially in the case of long-term forecasts where there are more uncertainties and individual analysts’ capability is not well observed.