企業在發生危機前,勢必會提早反應且存在異常報酬,此外,當公司面臨財務危機時,經理人是否會誘使企業透過操縱盈餘以掩飾其財務之困難。當公司發生問題時,基金公司或投資人若能從異常報酬或盈餘管理層面看出端倪,將能即早做因應。本研究以台灣2006年至2020年間發生財務危機之公司為研究樣本,以二元羅吉斯迴歸模式探討異常報酬、盈餘管理與財務危機之關係。 本研究首先篩選出225件之有發生危機事件公司作為研究樣本,並以未發生財務危機之公司作為對照樣本,由於本研究以營業額相近做為配對公司,加上發生財務危機下市後,有許多資料遺漏,因此最後的樣本數為126家公司,而配對公司為158家。 研究發現當發生危機事件前一日存在顯著之負向異常報酬,且在發生危機事件後一天之負向之異常報酬反應最大,且持續10天之久,且研究發現下市危機公司之負向異常報酬皆比上市、上櫃、興櫃危機來的大。本研究以四因子模型探討在有無發生財務危機之公司是否存在長期之異常報酬,我們發現財務危機公司在發生財務危機之前一個月、前三個月及前六個月皆存在負向之異常報酬。 最後由二元羅吉斯迴歸探討異常報酬、盈餘管理與財務危機之關係,當公司存在異常報酬時,則公司發生財務危機之機率愈大,而盈餘管理與財務危機皆呈現無顯著相關,表示當公司發生財務危機時,經理人可能會被迫轉為保守操作,以降低公司對於財務之盈餘管理操控。
Before a crisis occurs, companies are bound to react early and have abnormal returns. In addition, when companies face financial crisis, will managers induce companies to manipulate earnings to cover up their financial difficulties? When there is a problem in the company, if the fund company or investors can see the clues from the abnormal return or earnings management level, they will be able to respond as soon as possible. This study uses Taiwanese companies with financial crisis from 2006 to 2020 as a research sample, and uses a binary logistic regression model to explore the relationship between abnormal returns, earnings management and financial crisis. In this study, 225 companies with crisis events were selected as research samples first, and companies without financial crisis were used as control samples. Since this study used similar companies in operating income as matching companies, and after the financial crisis occurred, the company was delisted. There are many data omissions, so the final sample size is 126 companies compared to 158 matching companies. This study found that there was a significant negative abnormal return on the day before the crisis event, and the negative abnormal return response was the largest one day after the crisis event, and lasted for 10 days. The study also found that the negative abnormal returns of the companies in the delisting crisis companies are larger than those in the listing, OTC and emerging companies. This study uses a four-factor model to explore whether there are long-term abnormal returns in companies with or without financial crisis. We find that financial crisis companies have negative abnormal returns one month, three months and six months before the financial crisis. Finally, the relationship between abnormal returns, earnings management and financial crisis is discussed by binary Logis regression. When the company has abnormal returns, the probability of the company's financial crisis is greater, and there is no significant correlation between earnings management and financial crisis. When the company is in financial crisis, managers may be forced to turn conservative operations to reduce the company's financial earnings management control.