Using The Artificial Neural Network (ANN) to Assess Bank Credit Risk: A Case Study of Indonesia

Ever since the Asian Financial Crisis, concerns have risen over whether policy-makers have sufficient tools to maintain financial stability. The ability to predict financial disturbances enables the authorities to take precautionary action to minimize their impact. In this context, the authorities may use any financial indicators which may accurately predict shifts in the quality of bank exposures. This paper uses key macro-economic variables (i.e. GDP growth, the inflation rate, stock prices, the exchange rates, and money in circulation) to predict the default rate of the Indonesian Islamic banks’ exposures. The default rates are forecasted using the Artificial Neural Network (ANN) methodology, which incorporates the Bayesian Regularization technique. From the sensitivity analysis, it is shown that stock prices could be used as a leading indicator of future problem.