Does the day of the week effect exist once transaction costs have been accounted for? Evidence from the UK

Abstract

This article investigates the day of the week anomaly in the FTSE 100 Share Index over an 11-year time period from 1 January 1986 to 31 December 1997. Its focus is to assess whether the day of the week effect continues to persist once transactions costs are considered. Unlike previous literature it uses the bid–ask spread as a proxy for transactions costs. It finds that once returns become robust to transactions costs the anomaly appears to fade away. It then extends the research by looking at the time-varying volatility of stock returns with use of a GARCH model. The GARCH results further support the fact that transaction costs appear to die away the day of the week anomaly in the UK Stock market.

Acknowledgements

This article has benefited by helpful comments and suggestions of M. Dixon and J. V. Healey, the authors are grateful for their input. The usual disclaimer applies.

Notes

¹ Partial surveys may be found in Thaler (1987a, 1987b).

² Selective papers include Fama (1965), Cross (1973), French (1980), Lakonishok and Levi (1982), Keim and Stambaugh (1984) and Connolly (1989, 1991).

³ A considerable body of empirical evidence documents the impact of seasonal and holiday patterns on security returns. Although important, the aim of the current article is to focus on the day of the week effect.

⁴ This is the standard formula that is used for computed daily returns from daily stock prices. For more details on this see Arsad and Coutts (1997).

⁵ See Fama (1965), Badrinath and Chatterjee (1991), Aggarwal et al. (1989) for further details.

⁶ This method was employed due to the departure of the residuals from normality. The residuals were found to be non-normal with the use of the test proposed by Jarque–Bera (1987).