Abstract:
Since the identification and documentation of a day-of-the-week effect, it has captured imagination of the investing public and the attention of researchers. Indeed a significant amount of research has been dedicated towards the day-of-the-week effect. Until recently, the results of such research were consistent in that the evidence seemed to indicate that a day-of-the-week effect may indeed exist throughout the world.
More recent studies have, however, produced different results and a second body of evidence is developing which indicates that the day-of-the week effect is dwindling. Attempts by researchers to attribute the day-of-the week effect to the settlement practices of various exchanges have met with limited success.
This study argues that one would expect traded prices on an exchange to incorporate an adjustment for the delay between transacting and settlement. A model is formulated to adjust the mean daily returns on the exchange for the particular exchange’s settlement practice. This model is tested against historic price data from the Johannesburg Stock Exchange.
The evidence presented does not support the notion that the traded prices are adjusted for the delay between the transaction and the settlement, the overall conclusion is that the settlement effect may represent a Johannesburg Stock Exchange inefficiency, though the size and significance of the effect has decreased in recent times
