BUILDING A FORECAST MODEL OF VIETNAM'S STOCK MARKET UNDER THE IMPACT OF MACRO FACTORS

Abstract

The purpose of this research is to apply machine learning models to evaluate and compare the forecasting capabilities of the VNIndex, HNX Index, Upcom Index trends, and the profitability of the industries under the influence of macroeconomic factors. The study focuses on assessing and comparing the forecasting performance of various machine learning models, including Linear Regression, K-nearest Neighbors, Random Forest, Lasso Regression, and Ridge Regression. The dataset utilized comprises historical data of macroeconomic factors such as money supply M2, deposit interest rates, overnight lending rates, consumer price index (CPI), USD/VND exchange rate; stock indices (Dow Jones, Nikkei 225, S&P 500, VN-Index, VN30-Index, Upcom-Index); other indices (oil prices, gold prices); and profitability data of five industry groups (Banking, Securities, Real Estate, Steel, Retail). The research employs metrics such as R2, MSE, RMSE, and MAE to evaluate the forecasting performance of the models. The results reveal that the Linear Regression, Random Forest, and Lasso Regression models exhibit superior forecasting performance compared to the other models when the target variable is the profitability of stock indices. On the other hand, the Ridge Regression model demonstrates higher performance in forecasting the profitability of industries.