Machine Learning Assessment of Gold’s Safe-Haven Role During Market Crises

For a long time, during periods of financial tightening, gold has been regarded as a defensive asset, although its safe-haven attribute varies with market conditions. This study employed the daily price data of gold and the S&P 500 index from 1990 to 2023 to analyze the dynamic relationship between gold and the stock market. Linear regression, decision tree regression, and random forest regression models are adopted to test whether gold can have safe-haven properties during market contraction periods. Through data cleaning and daily earnings calculation, the results show that during periods of market stability, the correlation between gold and the stock market is not obvious. However, during major crises such as the 2008 global financial crisis and the COVID-19 pandemic, the correlation between gold and the stock market has significantly increased. Among the several types of models tested, random forest regression performed the best because this model can more effectively capture nonlinear changes in market behavior, which are often overlooked in traditional linear changes.