The Impact of Jumps and Leverage in Forecasting the Co-Volatility of Oil and Gold Futures

Abstract

This paper investigates the impact of jumps in forecasting co-volatility in the presence of leverage effects for daily crude oil and gold futures. We use a modified version of the jump-robust covariance estimator of Koike (2016), such that the estimated matrix is positive definite. Using this approach, we can disentangle the estimates of the integrated co-volatility matrix and jump variations from the quadratic covariation matrix. Empirical results show that more than 80% of the co-volatility of the two futures contains jump variations and that they have significant impacts on future co-volatility but that the impact is negligible in forecasting weekly and monthly horizons.

Forecasting volatility and co-volatility of crude oil and gold futures: Effects of leverage, jumps, spillovers, and geopolitical risks

“…While the direct effects from past volatility and returns are significant at the five percent level, the effects of past jumps are insignificant. The result is against the findings by Asai, Gupta, and McAleer (2019), and might be caused by the structure for guaranteeing positive definiteness of the covariance matrix. The sign of the t statistics for testing H kk,a 0 : a * k 2 k 2 ,a = 0 (k = 1, 2) indicates that a negative return increases future volatility, showing the existence of leverage effects.…”

“…Among them, we use the estimators of Christensen, Kinnebrock, and Podolskij 2010 As in Asai and McAleer (2017), we also apply thresholding of Bickel and Levina (2008) to guarantee the positive (semi-)definiteness of the estimators. Denote the estimators of QCov, ICov and jump component at day t as Ω t , C t and J t , respectively.…”

“…In addition, realizing the importance of jumps, that is, discontinuities, in the volatility processes of gold and oil prices (Sévi, 2014;Prokopczuk, Symeonidis, and Wese Simen, 2015;Demirer et al, 2019;Gkillas, Gupta, and Pierdzioch, 2019a), we also investigate the impact of jumps by simultaneously accommodating leverage effects, in addition to spillovers and geopolitical risks, in forecasting the volatilities and co-volatility of gold and oil markets, following the econometric approach of Asai and McAleer (2017) (applied to three stocks traded on the New York Stock Exchange (NYSE)). While some recent studies (see, for example, Demirer et al (2018), Baur and Smales (2018), Gkillas, Gupta, and Pierdzioch (2019b), Plakandaras, Gupta, and Wong (2019)) provide some, albeit weak, evidence of the role of geopolitical risks in predicting (in-and out-ofsample) gold and oil price volatility, to the best of our knowledge, this is the first paper to forecast volatilities and co-volatility of the two markets by accommodating geopolitical risks, jumps, leverage, and spillovers simultaneously in a model.…”

“…While some recent studies (see, for example, Demirer et al (2018), Baur and Smales (2018), Gkillas, Gupta, and Pierdzioch (2019b), Plakandaras, Gupta, and Wong (2019)) provide some, albeit weak, evidence of the role of geopolitical risks in predicting (in-and out-ofsample) gold and oil price volatility, to the best of our knowledge, this is the first paper to forecast volatilities and co-volatility of the two markets by accommodating geopolitical risks, jumps, leverage, and spillovers simultaneously in a model. In this regard, our paper can be considered to be an extension of Asai, Gupta, and McAleer (2019), in which the authors provided forecasts of the co-volatility of gold and oil using the information content of jumps and leverage.…”

The predictive power of oil price shocks on realized volatility of oil: A note

“…In the case of forecasting analysis, we use variants of the widely-studied HAR-RV framework of Corsi (2009) to model and forecast daily realized oil volatility. While the HAR-RV model apparently has a simple structure, it has become increasingly popular in the literature because it is able to capture long memory and multi-scaling behavior of commodity (oil) market volatility ( Asai et al, 2019 , 2020 ). In our application, the benchmark HAR-RV model is given by: where the index denotes -days-ahead realized volatility, with 1, 5, and 22 in our context.…”