Time-Frequency Analysis of the Interrelationship between the Global Macro Assets and Fear Indexes Using Wavelet-Based Tools

Understanding the interrelationships of global macro assets is crucial for the global macro investing. This paper investigates the local variance and the interconnection between stock, gold, oil, forex and implied volatility markets in time-frequency domains using wavelet methodology, including wavelet power spectrum, wavelet squared coherence and phase difference, wavelet multiple correlation and cross-correlation. The univariate analysis reveals that, in some crisis periods, underlying asset markets present the same pattern in terms of wavelet power spectrum indicating high volatility for the medium scale, and for other market stress periods, price volatility behaves differently. Moreover, unlike underlying asset markets, implied volatility markets are characterized by high power regions across the entire period even in the absence of economic events. Bivariate results show, differently from the other markets, a steady co-movement and causality between stock index and its corresponding fear index. Multiple correlation analysis indicates a strong correlation between markets at high scales with evidence of nearly perfect integration for a period longer than a year. In addition, the hedging strategies based on the volatility index lead to an increase in correlation between markets. On the other hand, the results from multiple cross-correlations reveal that the lead-lag effect starts from medium scale and that the VIX index pulls the rest of the markets