High-frequency Pairs Trading on a Small Stock Exchange
A. B. Mikkelsen, Frode Kjærland
Abstract
A. B. Mikkelsen, Frode Kjærland
Abstract
We study the performance of a high-frequency pairs trading (PT) strategy on the 100 most liquid stocks, in 15-min intervals, on a small commodity dominated stock exchange (Oslo Stock Exchange) using a comprehensive dataset from January 2012 to March 2016. We use both the distance and cointegration approach. Moreover, we let the formation (trading) period vary between 2 (1), 4 (2), and 6 (3) weeks, in order to test the impact on the PT profit. We find that the distance and cointegration approaches both have their strengths and weaknesses. In addition, we find that a shorter formation (and trading) period yields better results. As a further contribution to the literature, our findings imply that a simple static PT strategy, still is profitable using high-frequency data. Further, our results show better performance in a bull market than in a sideways-moving, volatile market.
OpenAlex reports 1 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
We study the performance of a high-frequency pairs trading (PT) strategy on the 100 most liquid stocks, in 15-min intervals, on a small commodity dominated stock exchange (Oslo Stock Exchange) using a comprehensive dataset from January 2012 to March 2016. We use both the distance and cointegration approach. Moreover, we let the formation (trading) period vary between 2 (1), 4 (2), and 6 (3) weeks, in order to test the impact on the PT profit. We find that the distance and cointegration approaches both have their strengths and weaknesses. In addition, we find that a shorter formation (and trading) period yields better results. As a further contribution to the literature, our findings imply that a simple static PT strategy, still is profitable using high-frequency data. Further, our results show better performance in a bull market than in a sideways-moving, volatile market.
Key concepts: Pairs trade, Cointegration, High-frequency trading, Algorithmic trading, Trading strategy, Econometrics, Stock exchange, Profit (economics)