THE INFLUENCE OF SAMPLE SIZE AND SELECTION OF FINANCIAL RATIOS IN BANKRUPTCY MODEL ACCURACY

Abstract

This paper aims to clarify the influence of changing both the sample size and selection of financial ratios in bankruptcy models accuracy of companies listed in the industrial sector of Jordan. The study sample is divided into three sub-samples counting 6, 10 and 14 companies respectively; each sample is composed of bankrupt companies and the solvent ones during the period from 2000 to 2013. Financial ratios were calculated and categorized into two groups. The first group includes: liquidity, profitability, debt, and activity, while the second group includes ten most popular financial ratios found to be useful in earlier studies and expected to predict financial distress. The results show that when 18 models built us ing discriminant analysis, the model based on most popular financial ratios, found to be useful in earlier studies, has the highest classification accuracy with 100% and consistently for all the samples before bankruptcy. The prediction accuracy varies among models when increasing the sample size from 6 to 14 companies for the models that developed from the financial ratios of the first group.