Analyzing the Effect of Fund Costs on Open-End Funds Performance Using XAI Technique

High fund costs, which are discussed as main factors reducing the competitiveness of domestic open-end fund products, have lowered the actual return on investment, leading to stagnation in the open-end fund market. Unlike previous studies, this research focuses on fund management and sales costs to predict the out-performance of domestic open-end funds. Using the latest data, we empirically analyzed the potential occurrence of agency costs between investors and fund family across all types of open-end funds in Korea. Considering the nonlinear relationship between fund costs and returns, we applied XGBoost to predict fund performance using fee rates and used the SHAP method to extract and visualize the feature attribute of cost variables from the XGBoost model. Additionally, we conducted OLS regression analysis to confirm the directional effect of fund costs on returns. The results showed that the effect of fund costs on performance varies by fund type. These findings provide significant insights and empirical evidence supporting the need to improve the fund cost policy in the domestic open-end fund market.