Artificial Neural Network for Financial Distress Prediction on Energy Companies Listed in Indonesia

Energy companies have specific characteristics and risks. Because of the COVID-19 epidemic, which has caused financial concern for both the world and Indonesia, the performance of Indonesian energy sector enterprises has declined since 2020. As a result, export activity to other countries decreased, resulting in negative EPS for 24 firms. The goal of this study is to use data mining to create a financial distress prediction model for energy companies listed on the Indonesia Stock Exchange for the years 2018-2021 using an Artificial Neural Network (ANN). Using financial ratios as input parameters, the optimal model with the lowest prediction error rate should be constructed, specifically the current ratio, return on assets, debt-to-assets ratio, and operating cash flow. According to the findings, 4 inputs, 5 hidden layer neurons, and 1 best model output were directly tested on the sample utilized; the findings indicate that 26 energy firms in Indonesia are in financial trouble, whereas 58 are not. This discovery may act as a warning to businesses to increase revenue in order to prevent financial difficulties. Traditional financial distress models, which have established researchers' recommendations for measuring financial hardship, can be used to improve and corroborate the AAN 4-5-1 model.