Financial Distress Prediction Model on Public Firms－The Application of Adjusted Merton Model

[[abstract]]本研究主要在探討調整Merton之信用風險模型是否能增進信用風險模型之違約預測能力。研究結果顯示，不論在KMV模式(流動負債加1/2長期負債)、各年度相同門檻的最適違約門檻模式、或是各年度不同門檻的最適違約門檻模式下，透過台灣經濟新報社(TEJ)的台灣企業信用風險指標(TCRI)，將信用風險因素調整進資產價值，其模型的違約預測能力明顯優於未調整資產價值之信用風險模型。以調整後資產價值為分析時，各年度不同門檻的最適違約門檻模式明顯優於KMV模式，各年度不同門檻的最適違約門檻模式優於各年度相同門檻的最適違約門檻模式。即各年度不同最適違約門檻之修正 Merton 模型，較原始Merton 模型與KMV模型更適用於台灣上市櫃公司對於信用風險的評估。在股價報酬率變異數方面，用歷史標準差估計的違約模型在2002年至2005年具有較佳的預測力，而以GARCH(1,1)標準差估計的模型，在2001年及2006年具有較佳的違約預測能力。[[abstract]]This research is mainly discussion about the prediction power of the credit risk in the adjusted Merton model. The result shows that, the prediction power of the model after adjusting the risk factors into the firm’s asset value by using the TEJ and TCRI index was excellent then the unadjusted model regarding of the KMV model or the optimal default threshold model, which has the same default threshold in each year or not. In the view of the adjusted firm’s asset value, the optimal default threshold model, which has the different threshold was better then the same threshold in each year and the KMV model. Thus, the adjusted Merton model, which has the different threshold in each year was more suitable for Taiwan public corporation about the evaluation of credit risk then the original Merton model and KMV model. In the side of the variance of the stock returns, the model with the historical volatility has better prediction power in 2002 to 2005; if the model with the GARCH(1,1) volatility, it has better prediction power in 2001 to 2006