DYNAMIC CONNECTEDNESS OF OIL TO AGRICULTURAL COMMODITIES: COMPARISM OF TIME-VARYING VAR AND DIEBOLD-YILMAZ METHODS.

Abstract

By using Time-Varying Parameter VAR (TVP-VAR), Diebold-Yılmaz, and Partial Correlation Network methodologies to analyze the time-varying variance-covariance mechanism of daily data for the period 20 May 1987 to 13 December 2023. This study investigates the dynamic connectivity of oil to agricultural commodities. Both West Texas Intermediate (WTI) oil and Brent oil were considered since they are two most popular oil markets in the world. Global
agricultural commodities are considered, and these are as wheat, corn, soyabean, cotton, sugar, coffee, cocoa, live cattle, and lean hogs. The results show that the assets under study exhibit distinct patterns of volatility interdependence. It is found that, using all available techniques, cotton, sugar, cocoa, and lean hogs are truly identified as net shock receivers. Findings showed that, the result of Diebold-Yılmaz were larger own-forecast errors (90.00, 61.59, 97.67,…….,89.84) and that of TVP-VAR were (58.49, 61.45, 68.98,…, 80.30); The fact that WTI and Brent crude oil are not listed convincingly as shock transmitters or shock receivers based on these different methods imply that researcher should be careful when rendering policies on them using one approach. It was discovered that, most of the agricultural commodities were shock recipients; investors should diversify their portfolio across commodities to minimize risk, as the connectedness between commodities varies across methods.