財務危機預測模型在各產業都是一重要課題,然過去研究多將營建產業樣本排除在外,原因是營建業獨特的產業特性與會計原則。本研究的目的為利用一創新方法結合兩種財務危機預測模型,以預測台灣營建業之財務危機機率。 過去研究之財務危機預測模型多以會計資訊建立,然會計報表易受人為因素操控且無法立即反應危機症狀。市場基礎模型利用公司的股票資訊預測財務危機,它的假設-公司的所有資訊都顯示在股票價格中-常與現實不符。 事實上,同時考慮市場與會計資訊最能完整反映公司的財務危機。本研究提出一結合會計與市場資訊之整合財務危機預測模型。方法是將市場基礎模型計算出的違約機率當成一變數輸入會計基礎模型之中。 本研究由台灣經濟新報取得上市與曾經上市之台灣營建公司資訊。結果顯示,與會計基礎模型(AUC: 0.6912)及市場基礎模型(Merton Model AUC: 0.7038 與 Barrier Model AUC: 0.7079)比較,整合模型(Hybrid 1 AUC: 0.7232與Hybrid 2 AUC: 0.7364)的預測能力較好。
Building default-predicting models is an important issue in all areas of business, yet past researches mostly excluded the construction industry from their sample due to its distinct characteristics and accounting principles. This study aims at predicting construction contractor default in Taiwan using an accounting-based model, market-based models, and a hybrid approach. Furthermore we compare the results of these different approaches using the Area Under Curve (AUC) to identify the most suitable model for predicting default in the construction industry in Taiwan. Default-predicting models are in large built by accounting information, yet accounting sheets are subject to manipulation and unable to show immediate symptoms. Market-based models use market information to predict default and are based on the presumption that all information is reflected in the stock price, which is only consistent in an efficient market. Both accounting-based and market-based models face some kind of limitation, thus in this study we try putting both accounting and market information into account to predict company default. We propose two hybrid default-predicting models that combine accounting and market information by inputting the default probability calculated from the market-based models as a variable into the accounting-based model. The input data of listed and unlisted construction firms in the Taiwan market was collected from the Taiwan Economic Journal. Results show that the hybrid models (Hybrid 1 AUC: 0.7232 and Hybrid 2 AUC: 0.7364) outperform both the accounting-based model (AUC: 0.6912) and the market-based models (Merton Model AUC: 0.7038 and Barrier Model AUC: 0.7079) in predicting construction contractor default.