PL EN


2013 | Volume 2 | 4 | 28-46
Article title

BUSINESS FAILURE PREDICTION MODELS BASED ON EXPERT KNOWLEDGE

Authors
Selected contents from this journal
Title variants
Languages of publication
EN
Abstracts
EN
This paper presents two business failure prediction models developed with multivariate linear discriminant analysis and multivariate logistic regression. The financial ratios as predictors for the models were selected based on results from previous empirical research. It was assumed that companies can be categorized into three classes – healthy (group 1), crisis-resistant (group 2) and insolvency endangered (group 3) – which are describing different economic conditions. Data for model building were obtained by a survey of 35 professionals from management consulting and banking industry. The results show consistency with findings of prior research. High values for equity-ratio, EBIT/total assets, operating cashflow/financial liabilities and percentage sales development are positively related to financial health. Within model building several problems occurred, which influenced classification accuracy. Non-normality of data had an impact on discriminant analysis, but also on logistic regression. Successful preliminary analyses of suitable predictors are not a guarantee that model fit including statistically significant variables will provide a superior prediction model. This indicates that model building is heavily dependent on the quality of metrics used. Logistic regression was less sensitive to outliers in terms of prediction sign within classification formula. It was also shown that crisis indicators used in practice are similar to those proposed by empirical research and literature.
Contributors
author
  • Redakce CJSSBE, Univerzitní servis, s.r.o., Trnkovo nám. 1112/2, 152 00 Praha 5, Czech Republic
References
Document Type
Publication order reference
Identifiers
YADDA identifier
bwmeta1.element.desklight-3a65e9d1-6930-48f9-97f4-d75a04ca4958
JavaScript is turned off in your web browser. Turn it on to take full advantage of this site, then refresh the page.