Bankruptcy prediction methods in the changing economic environment of the Czech Republic

Paper is focused on the corporate bankruptcy prediction models and methods which can be used right now in the Czech Republic and which should provide reliable and sufficient results. Since 1960's there have been created plenty models but not all of them have enough accuracy for current market conditions. National as well as foreign imported methods predicting financial difficulties are used. The Czech Republic has been witnessing increasing number of insolvency proposals since 2008. There are several reasons like new insolvency law or global economic crisis. Changing economic environment brings the necessity of testing currently used methods and models because their reliability can decrease during the time period as many studies prove.

Key question is if we have sufficient tools for predicting corporate financial distress which could lead to insolvency and corporate bankruptcy. Consequences of corporate financial distress for other business partners can be devastating and finishing as well by an insolvency proposal. To avoid these consequences corporations need the tools providing reliable answers.

This research should answer the question which corporate bankruptcy prediction models provide the most reliable results and therefore they are recommended to be used in the changing economic environment. The accuracy of current used models and methods will be tested. On one hand we need to forecast the higher possibility of the corporate bankruptcy, on the other hand we do not want to stigmatize healthy companies as classifying them as unhealthy. Model cannot fail in any above mentioned way. It has to mark healthy companies as companies with low risk of bankruptcy and unhealthy entities as entities with higher risk of bankruptcy.

Reliability of current used tools predicting corporate bankruptcy will be tested with the use of historical accounting data. Analysed time period will contain only years since the new insolvency law is effective in the Czech Republic (2008). Analysed companies can be classified as industrial because most of the models have been created for this type of companies.  For other branches the models usually do not provide reliable answers. This statistical testing will prove which model has the highest accuracy right now. It is even not possible to reject the hypothesis that the new model should be created because the current tools do not provide reliable results any more.

This paper is one of the outputs of the research project TD 010093 registered at the Technology Agency of the CR (TACR).