Dynamic Analysis of ETF Returns under the Influence of Macroeconomic Variables and Investor Behavior: A GMM-PVAR Approach

This study analyzes the dynamics of ETF returns in response to macroeconomic variables and investor behavior using a Panel Vector Autoregression (PVAR) model with a Generalized Method of Moments (GMM) estimator. Monthly data for the period 2013–2023 are employed. Macroeconomic variables (interbank interest rate, inflation, gold price, and exchange rate) and behavioral variables (investor sentiment and risk aversion) are analyzed within a structural panel vector autoregression (PVAR) framework. Impulse-response function results indicate that a positive shock to risk aversion leads to an immediate decline in ETF returns, whereas a shock to investor sentiment has a strengthening, persistent effect. The interest rate also exerts a significant negative pressure on returns. Furthermore, ETFs play an active role in shock transmission, such that increases in ETF returns reduce risk aversion and strengthen investor sentiment. Macroeconomic variables also exhibit distinct responses, including defensive behavior in gold prices, short-term reactions in inflation, and an upward trend in exchange rates. FEVD results indicate increasing long-run interdependence among variables and highlight the prominent roles of interest rates, risk aversion, and ETF returns in explaining fluctuations in other variables. These findings emphasize the dynamic and endogenous nature of the relationships between behavioral factors, macroeconomic variables, and ETF returns, as well as the importance of ETFs as a channel for financial shock transmission.