Financial distress prediction of public listed companies in Malaysia using artificial neural network

According to the statistical records of Malaysian bankruptcies taken from Trading Economics (2019), there is an increasing trend of bankruptcy rate in Malaysia since 1998. Therefore, financial distress prediction of Public Listed Companies in Malaysia plays a crucial role for investors, financial managers, and also creditors for them to make the right decision to invest, manage and lends money respectively. The financial distress of the company needs to be identified to avoid the company being declared as bankrupt. There are various methods or techniques that have been introduced over the year where commonly used are Multiple Discriminant Analysis and Linear Regression. The accuracy of the method is really important to estimate the prediction. In this study, Artificial Neural Network (ANN) was used to predict the financial distress of the companies which are listed in Bursa Malaysia. Data of 120 companies from sectors in the PN17 & GN3 list were used in this study. Liquidity and profitability ratios were used as main financial ratios with seven inputs nodes and thirteen hidden nodes. One output node with sigmoid function as activation function is used to determine whether the company is facing financial distress or not. Five subsamples were used to estimate the accuracy of the method where the performance of each subsample is measured using Mean Square Error (MSE). The results show the financial distress company predicted by ANN is quite similar to the financial distress company listed in PN17 & GN3 by Bursa Malaysia. As the accuracy of ANN in each subsample is above 50 percent and nearly hundred, ANN is proven to be a good method to be used for financial distress prediction in Malaysia.