Financial Distress Prediction Models for Small and Medium Enterprises in Pakistan

Abstract

This study aimed to develop and evaluate financial distress prediction models for small and medium enterprises (SMEs) in Pakistan. Financial ratios and firm-specific variables were used as predictors in logistic regression and neural network models. Data was collected from 250 SMEs, with 125 financially distressed and 125 non-distressed firms over the period 2015-2019. The logistic regression model achieved an accuracy of 82.4%, while the neural network model had 86.8% accuracy in classifying distressed and non-distressed SMEs one year prior to distress. Profitability, liquidity, and leverage ratios were found to be significant predictors of financial distress. The study provides valuable insights for SME stakeholders in assessing potential financial distress risks.

Keywords: Financial Distress, Small and medium enterprises, Prediction Models, Logistic Regression, Neural Networks.