발생액과 현금흐름이 현금배당 공시시점의 주가반응에 미치는 영향

Earnings can be decomposed into two components: cash flows and accruals. The cash flows are regarded as a reliable information on firm performance and as persistent for longer period of time. In contrast, accruals are regarded not as reliable as cash flows and not persistent. As a result, when investors observe information on earnings, they respond differently depending on the perceived reliability of the information. This study investigates how investors respond to the cash flows components and accruals components of earnings separately when dividends are announced in the stock market. Dividends policy is a very important decision among corporate financial decisions. In return for the financial investments by shareholders, firms pay dividends to the shareholders. Dividends payments provide important information to the shareholders on the future prospects of the firm as well as the current financial status and operating performance. These perspectives on dividends are collectively called as signaling hypothesis of dividends (Miller and Rock 1985). As a result, when dividends are announced, stock market strongly responds to the announcements (Aharony and Swary 1980; Asquith and Mullins 1983; Woolridge 1983; Grullon et al. 2002). Especially, when firms announce dividend increase or initiation (decrease or suspension), stock price significantly increases (decreases). The magnitude of the reaction is greater when firms announce dividend decrease or suspension than that when firms announce dividend increase or initiation. The findings suggest that investors regard the dividend decrease or suspension as more reliable information than dividend increase or initiation. In accounting area, studies link dividends to earnings. They find that the magnitude of the market reaction at the time of dividends announcements is positively associated with the level and quality of earnings (Healy and Palepu 1987; Manakyan and Carroll 1991; Nissim and Ziv 2001; Mikahil et al. 2003; Koch and Sun 2004). When the pattern of earnings stream supports the change of dividends, the market regards the dividends information as more credible and responds more strongly. Investors may treat the components of earnings differently. Healy and Palepu(1993), Sloan (1996) and Subramanyam (1996) argue that accruals may reflect the information component (which conveys inside information to external financial statement users) and earnings management component (which distort true financial performance of the firm). Depending on the market’s perception on the reliability of accruals, the market treats accruals as important information or evidence of earnings management. If investors regard the information as the former (later), market is (not) likely to respond to the component of accruals at the time of the dividend announcements. We empirically test this prediction whether the market responds differently to the two components of earnings at the time of dividends announcements. Our samples are 2,550 firm-year observations that announce cash dividends during the period from year 2001 to year 2010 in Korean Stock Market. Among them, 1,077 observations increased their dividends while 522 observations decreased their dividends. There were no change of the dividends from pervious years for the remaining 951 observations. In terms of sample size, this study is one of the rare works that use extensive samples collected over long period in Korean stock market. Most prior Korean studies, that mostly use small sample of dividends, even had a trouble to find the significant market reaction at the time of dividends announcements. Empirical results using these samples are summarized as follows. First, as documented in prior studies, we find that earnings of the year is positively associated with the magnitude of the market’s reaction (measured by the cumulative abnormal returns during the event period) at the time of cash dividends announcements. Second, when we divide the earnings into accruals and cash flows, we find that both components are significantly associated with the magnitude of the market’s reaction. In addition, the magnitudes are not significantly differ between cash flow component and accruals component. It implies that the market regard the accruals as a credible signal of private information rather than the output of discretionary earnings management by insiders. Third, when divide the samples to the subsamples, the firms with dividend increase and the other firms with dividend decrease, we find that all the significant results are due to the subsamples with dividend decrease. For the samples with dividend increase, we fail to find any significant relations, suggesting that investors trust dividend decrease as more credible signal than dividend increase. Otherwise, it could be the evidence of asymmetrically stronger investor response to bad news disclosure compared with good news disclosure (Kahneman and Tversky 1979). Fourth, above explained results are not sensitive in various analyses, suggesting the robustness of our findings. These findings provide valuable insights into the market’s behavior at the time of dividends announcements.