Nasruddin Zainudin

Modelling small and medium-sized enterprises' failure in Malaysia

Small and medium-sized enterprises (SMEs) are acknowledged as significant contributors to development and growth in the economy. Since failure among SMEs is common, this study intends to build an accurate model that could predict SME failure. An analysis on 132 companies in 2000 to 2010 shows that higher gearing and lower profitability are associated with higher failure risk. In addition, the incorporation of the companys age significantly improves the models predictive accuracy. Our result indicates that young SMEs rely heavily on debt, which leads them into distressed situations. To validate the predictive accuracy of the model, the area under the receiver operating characteristic (ROC) curve is utilised, suggesting that the inclusion of the non-financial variable significantly improves the model. The overall prediction accuracy rate ranges from 75% to 89% for the model developed with non-financial variables when applied to the one-year, two-year, three-year and four-year prior-to-default holdout samples.

Modelling Small Business Failures in Malaysia

Small and medium-sized enterprises (SMEs) are acknowledged as significant contributors to development and growth in the economy. Since failure among SMEs is common, this study intends to build a failure prediction model for SMEs in the manufacturing sector. By using 132 Malaysian SMEs during the period 2000 to 2010, the results show that higher gearing and lower profitability are associated with higher failure risk. Furthermore, the incorporation of firm’s age significantly improves the model’s predictive accuracy. Receiver Operating Characteristics curve demonstrates that the models possess better prediction ability than a random model. However, a model which includes financial and non-financial variables show superior performance. This model could detect failures as early as four years prior to the event. Overall, the accuracy rate that the model could correctly predict failures ranges from 75 percent to 89 percent.This model could be used as a refined tool to avoid possible adverse situations among the SMEs. A better credit decision making could take place that could potentially reduce non-performing loans.