An Investigation into The Predictive Power of Overnight Gaps on The Johannesburg Stock Exchange

Stock market gaps occur nearly every day, yet very little is known about their influence on subsequent pricing behaviour, particularly in developing economies like South Africa. The aim of this research is to comprehensively identify and analyse the relationship between overnight price gaps and subsequent intraday returns for publicly traded South African companies. These theoretical findings will also be applied practically, with a simulated trading strategy created and tested based on the theoretical and statistical findings. The primary method of identifying the underlying relationships at play is a collection of multiple linear and multiple logistic regression models, created using data spanning 20 years and 371 companies and split into training and testing sections to ensure accurate and bias-free results. The robust set of statistical tests and analyses performed indicates a persistent and highly significant inverse relationship that exists between large overnight gaps and subsequent intraday returns. This significant relationship was also applied to a very successful mean-reversion based trading strategy, with a sustained average annual outperformance of the JSE AllShare Index observed under even the highest transaction costs of 1.3% per trade. At the lower transaction cost level associated with CFD trading, an average annual outperformance of 166% was recorded. The theoretical and practical implications of these findings are in stark contrast to the widely accepted efficient market hypothesis and provide compelling new evidence of the exploitable nature of a relatively under researched market anomaly created by the inefficiencies associated with overnight price gaps. These results also pave the way for further analyses of gap behaviour, and collectively these findings add new and meaningful results to the body of knowledge on market anomalies.

Grantor

Department of Finance and Tax

Year dc:date.issued

2023

Author dc:creator

• de Jager, Johannes

Advisor dc:contributor.advisor

• van Rensburg, Paul

• Finance and Tax

Handle dc:identifier.uri

http://hdl.handle.net/11427/39335

OAI identifier oai:identifier

oai:open.uct.ac.za:11427/39335