Pairs trading of Chinese and international commodities

Adrian Fernandez-Perez, Bart Frijns, Ivan Indriawan and Yiuman Tse

Applied Economics, 2020, vol. 52, issue 48, 5203-5217

Abstract: We investigate the profitability of a pairs trading strategy using Chinese and international commodity futures contracts covering the period January 2004 to February 2018. We use a time-series approach where the commodity pairs share a similar underlying. An out-of-sample test is employed to infer the performance of these pairs, allowing us to determine the optimal open and close positions for the pairs trading strategy. Applying this strategy to a portfolio of commodities yields an excess return of 2.08% per annum and a Sharpe ratio of 0.79. For a portfolio of metal futures, this strategy yields 5.32% excess returns and a Sharpe ratio of 1.47, whereas for gold-only futures, this strategy yields 7.39% excess returns and 1.95 Sharpe ratio. This performance is superior to traditional strategies based on term structure, momentum, and value portfolios. Arbitrage opportunities in these commodity pairs remain even after accounting for transaction costs and are robust to data-snooping bias.

Date: 2020
References: Add references at CitEc
Citations: View citations in EconPapers (2)

Downloads: (external link)
http://hdl.handle.net/10.1080/00036846.2020.1761009 (text/html)
Access to full text is restricted to subscribers.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:taf:applec:v:52:y:2020:i:48:p:5203-5217

Ordering information: This journal article can be ordered from
http://www.tandfonline.com/pricing/journal/RAEC20

DOI: 10.1080/00036846.2020.1761009

Access Statistics for this article

Applied Economics is currently edited by Anita Phillips

More articles in Applied Economics from Taylor & Francis Journals
Bibliographic data for series maintained by Chris Longhurst ().