FINANCIAL RATIOS, MACROECONOMIC VARIABLES AND THEIR INTERACTION EFFECTS IN MODELS OF FIRM-SPECIFIC FINANCIAL DISTRESS (BANKRUPTCY, LOGIT)

Empirical financial distress research has focused on using financial ratios as predictors of firm-specific financial failure. Little attention has been given to the use of ratio trends, macroeconomic variables and variable interaction effects. Also, previous empirical models have attempted to distinguish between failed firms and either pairwise matched or randomly selected nonfailed firms; no attempt has been made to discriminate between failed and financially weak firms. The main thrust of this dissertation is to measure the effects of including additional variables and their interactions, and also to see if a model could be developed that could distinguish not only between bankrupt and nonbankrupt firms, but also between bankrupt and financially weak firms.; Logit analysis was used to test a series of 28 models for measurable effects when trend, macroeconomic and interaction variables were added to two basic sets of financial ratios. The estimation results showed that, as a group, the financial ratios interacting with the trends in the financial ratios were consistent in adding explanatory power to the models. The macroeconomic variables had stronger and more consistent effects in models to distinguish between bankrupt and financially weak firms. Other variables were less consistent in their statistical significance and contribution to classification accuracy.; Classification accuracy improvements from the original estimations were difficult to translate into improved forecasting results (i.e., when estimating in one time period and using a holdout sample from another time period). Large models containing numerous financial ratios, trends, macroeconomic variables and interaction effects were inferior to simple models containing only financial ratios. However, some small improvements over the basic financial ratio models were made using a factor analyzed model to reduce the large number of explanatory variables.; The results of the statistical estimations and forecasts suggest that macroeconomic variables and various interaction effects can be used to improve both explanatory power and forecasting ability of financial distress models, especially in models to distinguish between bankrupt and financially weak firms. However, model specification and variable selection will have to be improved to produce substantial and consistent gains over existing models.