Failure Predictions for the Small-Medium Enterprises in the UK (original) (raw)

Abstract

Most of the business failure prediction studies focus on large firms. This study develops failure prediction models specifically for SMEs in the UK using financial ratios. We have shown that financial ratios are one of the best warning signals to predict business failure, due to their quantitative nature, simplicity, and their ability to explain and evaluate business conditions and performance. We use a range of financial ratios representing profitability, liquidity, activity, and leverage, and select the most significant ratios in predicting failures. Using MDA, Logistic Analysis which are both the most popular methods, and Neural Network as a non parametric technique, on UK data consists of 560 SMEs (under the UK definition of SMEs, less than 250 employees, less than £25.9m turnover and less than £12.9m total assets), over the period 2000-2007, three-year failure prediction model has been developed for the SMEs and other two models have been developed for small firms in one hand and for medium firms in the another. The results were tested using a holdout sample of 187 (represents 33% of the original sample) to measure the prediction level of the models. For the three models, the small, the medium, and SME, the predictable variables were as follows: the small firms’ model (profit margin, net assets turnover, creditors’ payment days, and solvency ratio), the medium firms’ model (return on assets, net assets turnover, creditors payment days, liquidity ratio, and solvency ratio), and finally the SMEs model (return on assets, net assets turnover, creditors payment days, liquidity ratio, and solvency ratio). The results show that the predictive variables in small model and medium model are almost the same, (there is only one more predictive variable in the medium firms’ model than the small one), and these results show the similarity of the both firms’ characteristics as they both capture almost the same predictive variables. The overall accuracy of the medium model was 76%, which was a little higher than the small 70%, which show that the quality of financial data in medium firms is higher than the smaller ones. Also there was consistency in the different methods as they have produced almost the same predictive variables in the three models. The results show that SMEs failure prediction model developed is capable of predicting failure with an accuracy level of 74%.

Ibrahim Khanji hasn't uploaded this paper.

Let Ibrahim know you want this paper to be uploaded.

Ask for this paper to be uploaded.