MULTIFRACTAL ANALYSIS OF CONSUMER BEHAVIOR DURING COVID-19

This study utilizes multifractal detrended fluctuation analysis (MF-DFA) to examine the nonlinear dynamics inherent in consumer spending and their correlation with financial market volatility in South Korea during the COVID-19 pandemic period (January 2019–August 2022). We analyzed high-frequency daily credit card transaction data, categorized into total, offline, and online expenditures across 17 distinct regions, to characterize the temporal complexity of consumption patterns. Prior to multifractal estimation, seasonal patterns — particularly quarterly trends — were removed using a three-month moving average subtraction method. Furthermore, we employed iteratively amplitude-adjusted Fourier transform (IAAFT) surrogate series to test the null hypothesis of linearity, confirming that the observed multifractal properties are intrinsic to the spending dynamics rather than artifacts of random processes or seasonal effects. We also find that the correlation between the complexity of consumer spending (quantified by the multifractal spectrum width, [Formula: see text]) and KOSPI market volatility weakened significantly during the peak pandemic year (2020) compared to the pre-pandemic period. However, this link resurged during the recovery period (2021–2022.8), particularly for online transactions, indicating a structural re-coupling between consumer behavior and financial market dynamics as the economy adapted to post-pandemic conditions. Furthermore, the study documents notable regional heterogeneity in transaction complexity and shows a weakening of the moderating influence of regional Gross Domestic Product (GRDP) on the relationship between spending complexity and market volatility during the most acute phase of the pandemic crisis.