Determinants of Gold Price in the Presence of Structural Breaks: An Error Correction Model Approach

Abstract

This study investigates the long-run and short-run determinants of global gold prices within a comprehensive econometric framework. Motivated by gold’s dual role as both a financial hedge and a safe-haven asset, the research addresses two core questions: (i) what long-run equilibrium relationships exist between gold prices and macroeconomic, financial, and uncertainty-related variables? and (ii) how do short-run adjustments and structural breaks shape these relationships over time? Using monthly data spanning 1985–2024, the analysis incorporates U.S. real interest rates, expected inflation, inflation volatility, world income, world inflation and its volatility, exchange rates, credit risk premiums, gold’s systematic beta, and crisis episodes. The empirical strategy combines graphical inspection, unit root and structural break tests, Autoregressive Distributed Lag (ARDL) modelling, and an Error Correction Mechanism (ECM), with the Engle– Granger two-step approach employed for robustness. Diagnostic and stability tests were also conducted to validate the models. The findings reveal limited evidence of stable long-run cointegration across the full sample, although specific drivers such as U.S. inflation volatility, world income, exchange rates, and credit risk premiums exert significant long-term effects under alternative specifications. In the short run, gold prices are strongly influenced by exchange rate movements, financial risk, and selected structural breaks, confirming its sensitivity to macrofinancial shocks. Importantly, the results demonstrate that gold’s safe-haven role is episodic rather than universal, activated primarily during systemic crises such as the Global Financial Crisis and COVID-19 pandemic.