To investigate to what extent transaction mechanism matters, we examine the daily returns of 29 foreign exchange rates in the New York market. This paper finds that the day-of-the-week effect existed in the 1980s for some, not all, currencies. The fact that the day-of-the-week effect existed for only some currencies suggests that the US transaction mechanism alone cannot explain the anomaly. Furthermore, this paper finds that the day-of-the-week effect disappears for almost all currencies in the 1990s. This latter result is consistent with previous studies on anomalies in the stock markets.
Following French (1980), numerous studies have confirmed day-of-the-week effects, such as the Monday effect.2 However, no complete explanation of these day-of-the-week effects has been presented so far.3 Therefore, further research efforts investigating the anomalies in various markets surely help us to understand the cause of the anomalies. For example, if the settlement procedure in the US stock market causes the observed anomaly, other US asset markets or non-US stock markets may show different anomalies. So, it is natural to investigate the anomalies in other markets.
Although foreign currency markets are as important as stock markets for portfolio investment, the anomalies have received relatively little attention in studies of exchange rate behavior. One of the few stylized factors is that that returns on several foreign currencies to an American investor are generally high on Monday and Wednesday and low on Tuesday and Friday. This was first found by McFarland et al. (1982) and have been confirmed by Jaffe and Westerfield (1985), So (1987), and Cornett et al. (1995).
In this paper, expanding the investigation of the day-of-the-week effect in foreign currency markets, we test the significance of such effect in 29 foreign currencies markets. That is, we test day-of-the-week effects for 15 European currencies, 2 Oceanic currencies, 4 American Continental currencies, 7 Asian currencies, and 1 African currency. It is notable that, as we use price data in the New York foreign exchange market, the transaction or clearing system is basically the same across the currencies. If we find that there are anomalies regarding only some currencies, the US settlement mechanism or US information release timings are not the sole cause of the anomalies. The anomalies may be attributed to the different quantity and quality of the information on issuing countries.
Furthermore, by dividing the sample into two sub-periods (i.e. the 1980s and the 1990s), our investigation provides some evidence regarding the relationship between the currency market efficiency and the financial developments. As Hiraki and Maberly (2003) pointed out, the day-of-the-week return patterns are not stationary but dynamic, and related to market microstructure. Therefore, as the globalization of financial markets has been accelerated since the late 1980s, it is particularly interesting to investigate the evolution of the day-of-the-week effect in world currency markets.
This paper is structured as follows. Section 2 explains our data. Section 3 shows the results. The last section contains our conclusions.
The fact that the anomalies are found only for European currencies in the 1980s suggests that the US settlement mechanism or US news release timing alone cannot explain the anomalies and that there were some European-specific factors that caused the anomalies in the 1980s. The anomalies have almost disappeared in the 1990s. Our results are consistent with the continuing development of foreign currency markets.
As mentioned in the Introduction, there are few studies that investigate the day-of-the-week effect in the foreign currency markets, and those previous studies dealt mainly with the 1980s experiences. This paper updates the data sets to cover the drastic changes in the 1990s. It is interesting to investigate whether anomalies existed in the 1990s, because the capital markets liberalization and economic globalization have unprecedentedly advanced in the 1990s. As far as we know, the previous studies commonly found the day-of-the-week effects in foreign currency markets. For example, the newest available paper, Aydoǧan and Booth (1999), using Turkish data for the 1986–1994 period, found that the day-of-the-week effects still existed in the Turkish foreign currency markets in the early 1990s. Contrary to previous studies, we found that the day-of-the-week effect in the foreign currency markets disappears in the 1990s. Our findings have not been previously recognized in the literature.
However, our findings that the day-of-the-week effect existed in the 1980s and disappears in the 1990s are completely parallel to what have been found regarding US and other countries’ stock markets. For example, Chang et al. (1993), Dubois and Louvet (1996), and Kamara (1995), among others, documented that the day-of-the-week effect in the US stock markets disappears in recent years. More recently, Kamath and Chusanachoti (2002) found that the day-of-the-week effect in the Korean stock markets existed in the 1980s, but disappears in the 1990s. Regarding Japanese stock markets, Hiraki et al. (1998) and Hiraki and Maberly (2003) found that the Japanese Tuesday effect disappears after the introduction of index futures.
Finally, while the main purpose of the paper is to test whether the day-of-the-week effect depended on currencies, further research is necessary to formulate a more complete explanation of why the day-of-the-week effects existed in some European currencies in the 1980s and disappears in the 1990s.
