ABSTRACT
This article aims to understand the gas-pricing mechanisms in the world’s major markets and hence draw implications for gas-pricing reform in Asia. It adopts a newly proposed time-varying Granger causality test to investigate the connection between crude oil and natural gas prices. The empirical results suggest the necessity to establish gas trading hubs and hence adopt hub-based pricing in Asia and Europe so that gas pricing can fully reflect the fundamentals in gas markets and help achieve more efficient gas allocation. The resultant growth in gas consumption and potential replacement of dirtier fuels such as coal and oil is important for emission reduction and hence climate change action in the world, particularly in Asia, the world’s most dynamic region.
Acknowledgements
The research presented in this paper benefited from a University Postgraduate Award (UPA) and an Australian Government Research Training Program (RTP) scholarship. The authors also acknowledge helpful comments and suggestions from the editors Professor Helen Ross and Assoc. Prof. Thilak Mallawaarachchi two anonymous referees, Muhammad Mohsin and participants in a work-in-progress departmental seminar (Economics, UWA, Australia) and a workshop on ‘Meeting Environmental Objectives through Energy Sector Reforms in Asia and the Pacific: Role of Energy Pricing and Reforms for Emissions Reduction’ that the ADBI and CAREC Institute jointly organised on 22–24 June 2020.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
1 Other examples include Asche, Osmundsen, and Sandsmark (Citation2006), Ramberg and Parsons (Citation2010), Wakamatsu and Aruga (Citation2013), Atil, Lahiani, and Nguyen (Citation2014), Brigida (Citation2014), Asche, Oglend, and Osmundsen (Citation2015), and Zhang and Ji (Citation2018).
2 The Brent and WTI crude oil prices are two popularly quoted commodity prices in the world.
3 We use the imported liquefied natural gas price of Japan as a proxy of Asian gas market.
4 Full sample covers all observations in the database. The window size refers to the number of observations included in each sub-sample.
5 For the sake of saving space, the graphic presentation of these results is not reported here but available upon request from the authors.