회계부정과 장기시장반응

Abstract

This study examines the market reactions of firms identified by the FSS(Financial Supervisory Services) as fraudulently manipulating their financial statements. Prior studies focused on the short period market reactions. However, this study investigates long-term market reactions to the FSS's enforcement actions released. My main research objective is to examine long-term decrease in the information content of earnings following FSS's enforcement actions released. Also I examine whether firm-level actions(especially company name change and CEO turnover) to restore investor reliability of the financial statements process affect the duration of the decline in the information contents of earnings. I measure the ERC(earnings response coefficient) over several quarters to examine whether accounting fraud result in a long-term decline in information content of earnings. I also examine cross-sectional variation in the duration of the decline in the information content of earnings across characteristics after FSS's enforcement actions released. I test for differential impacts for firms caused by income overstate. And I also test for differential impacts for firms which change company name or CEO after FSS's enforcement actions released. I use a sample of 113 firms identified by the FSS(Financial Supervisory Services) as fraudulently manipulating their financial statements between 2002 to 2008. I measure CAR(Cumulative Abnormal Return) surrounding the earnings announcement for firms measured over a five-day window, where the abnormal return is the firm's return less the value-weighted index return (or equally weighted index return). I measure quarterly UE(unexpected earnings) for each announcement date, scaled by price as of the end of the quarter for which earnings are announced. The empirical results summarized as follows: First, the ERC is lower for several periods than pre-fraud released quarter. However, the ERC return to pre-fraud released quarter after average third quarter. This result indicates that the decline in the information content of earnings following fraud released is not long-lasting. Second, the ERC is more decline for firms that related income overstated fraud. Third, the duration of the decline in the information content of quarterly earnings is shorter for firms that change their name. Fourth, CEO turnover doesn't affect to the duration of the decline in the information contents of earnings. I believe that this empirical evidence signifies a meaningful policy implication. For example, fraud firms can not change their corporate names for a certain period for investor protection.