Abstract

In this paper, we test the use of Markov-switching (MS) GARCH (MSGARCH) models for trading either oil or natural gas futures. Using weekly data from 7 January 1994 to 31 May 2019, we tested the next trading rule: to invest in the simulated commodity if the investor expects to be in the low-volatility regime at t + 1 or to otherwise hold the risk-free asset. Assumptions for our simulations included the following: (1) we assumed that the investors trade in a homogeneous (Gaussian or t-Student) two regime context and (2) the investor used a time-fixed, ARCH, or GARCH variance in each regime. Our results suggest that the use of the MS Gaussian model, with time-fixed variance, leads to the best performance in the oil market. For the case of natural gas, we found no benefit of using our trading rule against a buy-and-hold strategy in the three-month U.S. Treasury bills.

Highlights

  • Energy futures, such as oil and natural gas, are a widely used means for hedging the commodity price risk and for investing and speculation

  • We found increasing literature related to testing the use of MS, GARCH, and MSGARCH models in European and emerging stock markets [41,44,45]

  • From the perspective of the use of MS and MSGARCH models in other types of time series, we found the works of Alexander and Kaeck [54] and Ma and Deng [55]

Read more

Summary

Introduction

Energy futures, such as oil and natural gas, are a widely used means for hedging the commodity price risk and for investing and speculation. Given their close relationship with economic activity and general prices, energy commodities (especially oil) have been a source of portfolio diversification. The diversification benefit for a portfolio is observed only with agricultural commodities [4] and other types of alternative assets, such as real estate [5], hedge funds [6], volatility futures [3], or clean-energy (technology) stocks [7]. The issue of a diversified portfolio has been tested in several academic reviews, such as in [2,8,9] which are some of the most recent ones

Objectives
Methods
Results
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call