在商業模式日新月異的時代，如何透過財務報表及公開資訊察覺具有財務困難之公司至關重要，因此本研究先以Altman (1968)的Z-Score建立財務指標模型，再結合Kim and Zhang (2014)的不透明度模型(Opacity)及Roychowdhury (2006)實質盈餘管理模型建立財務與盈餘管理指標模型，並透過羅吉斯迴歸進行分析探討，最後以經濟成長率之變動為依據區分樣本，探討總體經濟環境對預測模型之影響。 實證結果顯示企業的流動性、累積至今之盈餘為預測財務危機重要之指標；而財務與盈餘管理指標模型顯示異常營業現金流量、異常裁決性支出於危機前一年及危機前二年呈現負向顯著，應計盈餘管理於危機前一年至危機前三年呈現正向顯著。與實質盈餘管理相比，應計盈餘管理於危機前三年即呈現顯著，顯示管理當局會優先選擇對公司損害較小之方法美化財務報表。考量總體經濟環境後，企業的流動性仍為重要之指標，而盈餘管理指標方面發現總體經濟環境較佳時，管理當局透過盈餘管理美化財務報表之程度較大。;This paper uses logistic regression to conduct a financial distress prediction model. There are five financial ratios in the financial model; the financial and earnings management model consists of financial model including five financial ratios, opacity indicators, and real earnings management indicators. Last, the discussion on the impact of macroeconomic on prediction. The empirical results show that liquidity and retained earnings are the most important ratios in financial model. Ab_CFO and Ab_DISEXP are significant in the year before the crisis events and the second year before it. Opacity is significant during the year before the crisis events to the third year before it. In addition, accrual earnings management is significant one year earlier than actual earnings management, which means the manager gives priority to choosing lesser hazard methods to manipulate financial statements. After considering macroeconomic, liquidity is an important ratio too. We find the manager uses earnings management to manipulate financial statements to a greater extent when the macroeconomic is better.
