Asymmetric gasoline-oil price nexus: recent evidence from non-linear cointegration investigation

ABSTRACT

In this paper, weekly data from June 1986 to October 2018 is used to investigate the asymmetric relationship between oil prices (WTI Spot Price) and gasoline prices (New York Harbor Conventional Gasoline Regular Spot Price). The novel non-linear Autoregressive Distributed Lags (NARDL) approach of cointegration was adopted to examine the asymmetric association. The empirical results support long- and short-run asymmetry. Furthermore, the direction of causality between oil and gasoline prices is examined using the Toda and Yamamoto non-causality test. The results show bidirectional causality between oil prices (positive and negative innovations) and gasoline prices.

KEYWORDS:

• Gasoline prices
• oil prices
• cointegration
• NARDL
• asymmetry

JEL CLASSIFICATION:

• E44
• G12
• Q43

Acknowledgments

The author wishes to express his gratitude to two referees and the Editor of this journal whose comments and suggestions improved the merit of this work. Needless to say, the usual disclaimer applies.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

¹ Pesaran, Shin, and Smith (2001) provided two sets of critical values (upper bound and lower bound) to verify the cointegration test. When the computed F-statistic exceeds the upper bound critical value, it indicates support for cointegration. Alternatively, when the F-statistic falls below the lower bound critical value, it indicates no cointegration. Finally, if the F-statistic falls within the bounds; then, the test is inconclusive.

² For more details, see Toda and Yamamoto (1995).

³ The findings are not reported in this paper for space limitations but can be made available upon request.

⁴ a measure of testing model specification, and follows ${\chi }^2$ distribution with one degree of freedom.

⁵ a measure of testing the serial correlation of the residuals, and has a ${\chi }^2$distribution with two degrees of freedom.

⁶ measures of the model stability, where the results are reported as stable and marginally stable.

⁷ In explaining how would gasoline prices cause oil prices, we consider the derived demand concept of microeconomics. Oil is an input used in refiners to produce gasoline; hence, demand for oil is a derived demand from demand for gasoline. As such, when gasoline prices increase, customers respond negatively, leading to an effect on demand for oil.