Identifying Chinese Hospitality Firm Failures and Differences from Results on Developed Countries: Significant Variables and Predictive Models

This research made an early investigation on firm failure prediction (FFP) in hospitality industry of China using support vector machine (SVM) and classical statistical models, and an early comparison on difference of identified significant variables in FFPs for developed and developing countries. The findings indicate that: (a) working capital turnover, equity turnover, ratio of owners´ equity, equity to fixed-assets, interest coverage ratio, fixed-assets turnover, and total assets turnover are significant ratios in distinguishing failed hospitality firms from non-failed ones in China; (b) SVM is the most preferred model in FFP of Chinese hospitality firms from a integrated view of k-times in-sample assessment and k-times out-of-sample assessment on various variable sets; and (c) Both MDA and logit can produce the most accurate in-sample or out-of-sample performance with the optimal variable sets. This study also finds that two chief differences exist between hospitality FFPs in developed countries and developing countries, which refers to: the number of significant financial ratios and the type of significant financial ratios.