IMPACT OF CAPITAL STRUCTURE ON BUSINESS PERFORMANCE OF ENERGY COMPANIES LISTED ON VIETNAM STOCK MARKET

Abstract

The article assesses the impact of capital structure on the performance of energy enterprises listed on the Vietnamese stock market, with data of 35 energy enterprises from 2012 to 2020 through using the XGB (Extreme Gradient Boosting) algorithm in python-based machine learning along with techniques to extract general insights from a machine learning models such as Xgboost feature importance and Shap (Shapley additive explanation). Research results show that enterprises with large capital size (SIZE) operate less efficiently, while enterprises with large net fixed assets over total assets (GROW) operate more efficiently. It is important for energy enterprises to reduce long-term debt (LDR), increase short-term debt (LDR) and reduce total debt to total assets (TDR) to maintain optimal capital structure while ensuring liquidity (LIQ) to improve business performance.