Predicting hospitality financial distress with ensemble models: the case of US hotels, restaurants, and amusement and recreation

The importance of industry-specific characteristics in financial distress is widely acknowledged, but often overlooked by researchers studying the hospitality industry. The primary objective of this paper is to investigate the key determinants of US hospitality firms’ financial distress between 1988 and 2010 using ensemble models. The data used in this study come from the Compustat database produced by Standard and Poor’s Institutional Market Services. The data were collected from three hospitality-related segments, 5812 eating Places, 7011 hotels and motels, 7990 amusement and recreation services not elsewhere classified according to Standard Industrial Classification. In the restaurant-stacking model, debt-to-equity ratio, growth in owners’ equity, net profit margin, and stock-price trend were chosen as financial distress predictors. In the hotel stacking model, debt-to-equity ratio, stock-price trend, and account receivable turnover were selected as financial distress predictors. In the amusement and recreation-stacking model, debt-to-equity ratio, growth in owners’ equity, net profit margin, and management practice were defined as significant financial distress predictors. Although many researchers have stressed that an ensemble method, which combines the characteristics and advantages of particular models, may improve the performance or interpretability of predictive methods, few hospitality financial distress prediction studies employed ensemble methods. This study demonstrates its originality in this perspective.