Best proxy to determine firm performance using financial ratios: A CHAID approach

Abstract The main purpose of this study is to investigate the best predictor firm performance among different proxies. A sample 287 Czech firms was taken from automobile, construction, and manufacturing sectors. Panel data acquired Albertina database for time period 2016 2020. Three proxies performance, return assets (RoA), equity (RoE), capital employed (RoCE) were used as dependent variables. Including three firm’s 16 financial ratios measured based on previous literature. machine learning-based decision tree algorithm, Chi-squared Automatic Interaction Detector (CHAID), deployed gauge each proxy’s efficacy examine proxy performance. partitioning rule 70:30 maintained, which implied that 70% dataset training remaining 30% testing. results revealed (RoA) detected be a robust predict targeted indicators. methodology will useful policy-makers, stakeholders, academics managers take strategic business decisions forecast