Multi-frequency downside risk interconnectedness between soft agricultural commodities

Abstract

In this article, we explore multiscale extreme risk interdependence between four soft agricultural markets - coffee, cocoa, cotton and orange juice. Wavelet correlation and cross-correlation are used to investigate this interlink, and dynamic conditional Value at Risk is used to measure extreme risk. Wavelet correlation results suggest a very weak connection between the markets in the short-term and midterm horizons, which means that investors who operate in the short term or midterm do not have to apply hedging measures against extreme risk. However, the situation is different in the long term, where relatively high correlations are found on the highest wavelet scale in all pairs, except coffee-cocoa. Complementary cross-correlation analysis indicates a lead-lag relationship between the markets. The results are mostly in line with expectations, as bigger markets lead smaller markets. Only in the cases of cocoa-cotton and cocoa-orange juice does the opposite happen.

Keywords

• conditional value at risk
• extreme risk interdependence
• wavelet correlation
• wavelet cross-correlation