Stylized Facts in Financial Markets: A Comparative Study of Gold and Equities

Abstract

This thesis investigates the applicability of eight stylized facts in financial markets by comparing the behavior of gold and the equities. Using daily, weekly, monthly and annual log-returns data from 2000 to 2023, we apply various statistical tests to analyze properties such as non-stationarity, stationarity, asymmetry, heavy tails, aggregational Gaussianity, lack of autocorrelation, volatility clustering, and the leverage effect. The findings reveal that while most of the stylized facts are applicable on both asset classes, distinct differences exist. For instance, significant autocorrelation in daily returns and the leverage effect are observed in the equities but not in gold, highlighting differences in market dynamics and risk profiles. These results suggest that financial models must be tailored to account for asset-specific characteristics, particularly when managing risks associated with extreme market events. Future research could extend this comparative analysis to other asset classes and explore high-frequency data to capture more granular market behaviors.