Optimistic Bias in Analysts’ Earnings Forecasts and Asymmetric Market Reactions to Earnings Surprises; [재무분석가 예측치의 낙관적 편향과 어닝 서프라이즈에 대한 시장 반응의 비대칭성*]

초록

This study empirically investigates how the stock market responds to analysts’ earnings forecasts in the context of the Korean capital market. Prior research highlights that earnings surprises play a critical role in corporate performance evaluation and stock valuation. While exceeding analysts’ forecasts typically leads to favorable market reactions, these forecasts are known to contain optimistic biases, often driven by analysts’ incentives to maintain relationships with management. As a result, reported earnings surprises may not fully reflect investors’ perceived surprises. Building on this premise, the study formulates and tests three hypotheses: (1) stock market reactions are asymmetric, with stronger responses to positive surprises than to negative surprises of equal size; (2) small positive surprises trigger disproportionately stronger market reactions due to the impact of forecast bias; and (3) small negative surprises may lead to upward price movements if investors reinterpret them as positive signals after adjusting for optimistic bias. Using a data set of 8,950 firm-year observations from publicly listed companies in Korea between 2012 and 2023, the empirical analysis confirms these hypotheses. The results show that market reactions to positive earnings surprises are significantly stronger than reactions to negative surprises of equal size. Moreover, the strongest reactions occur in cases of small positive surprises, and in the case of small negative surprises, stock prices sometimes move in the opposite direction than implied by the reported surprise. This study contributes to the literature by offering evidence that the Korean stock market disproportionately rewards earnings surprises, taking into account the perceived optimism embedded in analysts’ forecasts. The findings provide new insights into how investors interpret earnings information in an environment where forecast bias is prevalent, thereby offering both theoretical and practical implications. © 2026, Korean Accounting Association.

키워드