Volatility spillovers across financial markets: the role of oil price uncertainty

ABSTRACT

This paper analyzes the state-dependent volatility transmission mechanism between oil, stock, dollar, and bond prices to further examine the role of oil price uncertainty in financial markets. To this end, we extend the Diebold and Yilmaz (2014) spillover framework by incorporating a Markov-switching model and a Bayesian MCMC algorithm. We find that oil prices spills the highest degree of volatility to other markets during crises. The interdependence between the stock and oil markets is solid and stable, regardless of the regime shift. In contrast, the effect of oil price uncertainty on the foreign exchange or bond market during crises is double that during non-crisis periods. This suggests that oil price is closely related to other asset classes and reinforces its role as a risk transmitter during a crisis.

KEYWORDS:

• Oil price
• spillovers
• Markov-switching
• bayesian estimation

JEL CLASSIFICATION:

• G15
• C32
• C11

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

¹ The DY method is based on the predictive error variance decomposition of system impulses in generalized VAR model. Therefore, the integration of Markov-switching and Bayesian MCMC methods with the DY method is doable. In contrast, estimating state-dependent parameters is not trivial in other approaches, such as the Multivariate GARCH model, due to dimensionality and identification restrictions.

² For technical details of each posterior sampling algorithm, refer to Kim and Lee (2020):

³ To ascertain the ARCH effects, we conduct Engle’s Lagrange multiplier test. In all the cases, we can reject the null hypothesis of no ARCH effect. The constant coefficients in the mean equation are either significant (oil and bond) or insignificant (stock and FX), and the GARCH parameters are well specified. Additionally, the correlation between estimated volatility and implied volatility indices is relatively high. For stock, the correlation between estimated volatility and VIX is 0.743. For oil, the correlation between estimated volatility and OVX is 0.742.

Additional information

Funding

This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. 2021R1G1A1093372).