eThesis - School of Business electronic theses

» hse_ethesis_14058.pdf   size:2 MB (1192012)

The purpose of this study is to synthesize the findings of prior bankruptcy prediction research studies by compiling and classifying the independent variables used as predictor variables in the studies. The objective is to find out the popularity of the different types of the predictor variables by classifying the variables into the categories describing the fincancial function of the variables, and by assessing the popularity of the significant variables in the categories. This work studies elementary theories on firm failure and bankruptcy to discuss and seek justitication for what might be the reasons for using the most popular financial function measures in the bankruptcy prediction.

Bankruptcy prediction research literature covers vast amount of studies in which various different predicton models are developed for predicting bankruptcy. Usually these studies use a prediction model with a set of some financial and/or non-financial variables that are presumed to be relevant proxies for financial distress and eventually business failure and bankrupcty. However, there seems to be no consensus or unified theory on how the variables predicting bankrupcty should be selected, thus the numerous bankruptcy prediction research studies include vast number and various different types of variables that are presumed to be applicable in predicting bankruptcy.

This study includes a systematic literature review where 51 bankruptcy prediction research studies were collected from well-recognized scientific journals. The studies included into the review were such that included a single or multiple bankruptcy prediction models, the detailed description of the independent variables, and the information about the statistical significances of the independent variables. The variables were then classified according to their financial function and a meta-analysis were conducted on those variables which were significant in bankruptcy prediction, to find out the popularity of the different variable categories.

The findings of this study suggest that the most popular predictor variables included into the banktuptcy predicton models are accounting-based financial ratios, particurarly ones measuring liquidity, profitability, and financial leverage, and that there exists also theoretical foundation for using these variables in the bankruptcy prediction.