Predicting Construction Contractor Financial Distress in Taiwan – Integration of Accounting-based & Market-based Models
Date Issued
2010
Date
2010
Author(s)
Wang, Wen-Pei
Abstract
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.
Subjects
construction contractor
financial distress prediction
market-based model
hybrid model
Type
thesis
File(s)![Thumbnail Image]()
Loading...
Name
ntu-99-R97521705-1.pdf
Size
23.32 KB
Format
Adobe PDF
Checksum
(MD5):bfb0ad3056ea996be7fc6491844e6f7b