The impact of economic policy uncertainty on Vietnam's Stock Market volatility: Evidence from a Domestic EPU Index constructed via a Hybrid ML-LLM approach and ARDL models

Abstract

This study empirically investigates the impact of Economic Policy Uncertainty (EPU) on the volatility of the Vietnamese stock market, measured by the VN-Index, over the period 2008- 2025. By applying a Hybrid Machine Learning approach, combining a Logistic Regression classifier with Large Language Model (LLM)-assisted labelling, to a textual dataset of nearly 500,000 Vietnamese-language economic news articles, a domestic EPU index for Viet Nam was constructed. Employing ARDL and NARDL models, the empirical results suggest the existence of a robust long-run cointegrating relationship between EPU and VN-Index volatility. Notably, the findings reveal significant asymmetric market responses, whereby positive EPU shocks, representing increased policy uncertainty (bad news), amplify market volatility considerably more than the dampening effect produced by negative EPU shocks representing reduced uncertainty (good news). These results underscore the importance of transparent and consistent policy communication in stabilizing investor sentiment and mitigating systemic risk in the Vietnamese equity market.