Transmission of risks between energy and agricultural commodities: Frequency time-varying VAR, asymmetry and portfolio management
Document Type
Article
Publication Date
3-1-2023
Abstract
This paper examines energy and agricultural commodities' short-run and long-run connectedness by using the Time -varying parameter vector autoregressions (TVP-VAR). It applies the frequency version of the TVP-VAR model, which is a modified version of the dynamic TVP-VAR model. The frequency decomposition definition also decomposes into short-run and long-run connectedness. We further the analysis by investigating the effect of asymmetry in returns on connectedness. It also examines how portfolio management strategies would lead to a maximization of profits with minimal risks. Empirical evidence indicates that only 32.52% and 31.38% of connectedness in oil and gas, respectively, are transmitted to agricultural commodities, which suggests their weak tendencies in influencing agricultural commodities; the total connectedness index hovers around 40-60% in the 2018-2019 period; however, it dropped below 40% in 2020-2021 when the COVID-19 pandemic contributed to disintegrate the connectedness between energy and agricultural commodities but increased further during the 2022 Russia-Ukraine saga. The findings also indicate that corn, wheat, and flour are net transmitters of risks to oil and natural gas in the long and short-run, and wheat-flour pairwise connectedness is the strongest in the connectedness. Asymmetry is also pronounced in the network of connectedness. Portfolio analyses indicate that investors require a low proportion of energy in a portfolio of energy-agricultural com-modities to achieve an optimum profit. The findings will offer exciting insights into the connectedness of agri-cultural and energy commodities, particularly during periods of high price uncertainty.
Keywords
Agricultural commodity, Asymmetry, Frequency TVP-VAR, Optimal weight, Risk
Divisions
deputyacademic,aei
Publication Title
Resources Policy
Volume
81
Publisher
Elsevier
Publisher Location
THE BOULEVARD, LANGFORD LANE, KIDLINGTON, OXFORD OX5 1GB, OXON, ENGLAND