Modelling the volatility of crude oil returns: Jumps and volatility forecasts

• Published date: 01 January 2021
• DOI: http://doi.org/10.1002/ijfe.1826
• Author: David Roubaud,Elie Bouri,Anupam Dutta
• Date: 01 January 2021

RESEARCH ARTICLE

Modelling the volatility of crude oil returns: Jumps and

volatility forecasts

Anupam Dutta

1

| Elie Bouri

2

| David Roubaud

3

1

School of Accounting and Finance,

University of Vaasa, Vaasa, Finland

2

USEK Business School, Holy Spirit

University of Kaslik, Jounieh, Lebanon

3

Montpellier Business School,

Montpellier, France

Correspondence

Elie Bouri, USEK Business School, Holy

Spirit University of Kaslik, Jounieh,

Lebanon.

Email: eliebouri@usek.edu.lb

Abstract

We contribute to the scarce literature on the oil market volatility index (OVX)

by examining the presence of time-varying jumps in OVX and by assessing the

ability of OVX to predict the conditional variance of crude oil returns. Using a

GARCH-jump model, we find evidence that OVX is characterized by jump

behaviour that tends to vary over time. Further analysis indicates that account-

ing for the jump behaviour of OVX helps improve the conditional variance

forecasts of crude oil returns. Since the studied features of OVX play a crucial

role in asset pricing and risk analyses, our findings have policy implications

related to refining volatility prediction models and risk measures.

KEYWORDS

conditional variance, GARCH-jump, OVX, time-varying jumps, volatility forecasts

1|INTRODUCTION

Crude oil plays a decisive role in economic and financial

stability, and its price volatility is crucial to asset pricing,

portfolio modelling and risk management (e.g., Bouri,

Lien, Roubaud, & Shahzad, 2018). Considering the eco-

nomic importance of crude oil, academics are constantly

engaged in finding appropriate methods of modelling oil

price volatility. However, looking for such methods is a

challenging task given that oil prices can become highly

volatile due to the adverse effects of natural calamities

(such as tsunami, drought etc.), and tend to behave dif-

ferently under diverse market conditions. For example,

Choi and Hammoudeh (2010) argue that downturns in

global stock markets could lead to a significant drop in

oil prices, whereas oil prices usually increase during bull-

ish periods. Moreover, oil price volatility is influenced by

the volatilities of other assets such as exchange rates and

the gold market.

Released by the Chicago Board Options Exchange

(CBOE), the crude oil implied volatility index (OVX) is a

reliable gauge for oil price risk. Like the US equity mar-

ket volatility index (VIX), OVX measures market

expectations of 30-day implied volatility. Importantly,

OVX conveys useful information for forecasting crude oil

price volatility, which makes it an important volatility

instrument for investors and policymakers (Becker, Clem-

ents, & White, 2006; Chatrath, Miao, Ramchander, &

Wang, 2015; Corrado & Miller Jr., 2005; Troster, Bouri, &

Roubaud, 2019). Nevertheless, while the properties of VIX

are well-documented in the existing literature,

1

the key

features of OVX have not received as much attention,

especially regarding the presence of jumps, and the fore-

casting ability of OVX and its jump behaviour. In this

study, we extend the scarce related literature by examin-

ing the presence of time-varying jumps in OVX and by

assessing the ability of OVX to predict the conditional var-

iance of crude oil returns.

The contributions of this empirical research are two-

fold. Firstly, we investigate the presence of time-varying

jumps in the crude oil volatility index. Such jumps, which

are common phenomena in financial and commodity

markets, may occur due to terrorist attack, natural disas-

ter, recession or political violence. The importance of ana-

lyzing the jump behaviour is illustrated in earlier studies

(see Pan & Duffie, 2001; Liu, Longstaff, & Pan, 2003;

Received: 25 April 2019 Revised: 9 October 2019 Accepted: 18 June 2020

DOI: 10.1002/ijfe.1826

Int J Fin Econ. 2021;26:889–897. wileyonlinelibrary.com/journal/ijfe © 2020 John Wiley & Sons, Ltd. 889