Background

Turkey is located at the junction of Europe, Asia and Middle East. This strategic geographical location, combined with massive domestic market and stable macroeconomic policy has enabled it to become the 16th largest economy of the world. It is a country with which large exchanges and global players wish to partner, and do business (Borsa Istanbul A Story of Transformatıon 2013; Ernst & Young, Attractiveness Survey Turkey 2013). The regulatory and supervisory authority is Capital Markets Board of Turkey (CMB) in charge of the securities markets in Turkey. All the exchange operating is done under Borsa Istanbul (BIST) in Turkey. The products of BIST are equities, exchange traded funds, warrants, options, futures, certificates, debt instruments and lease certificates (Borsa Istanbul 2016). Markets of BIST are equity market, debt securities market, derivatives market, precious metals and diamond market, and market surveillance activities (Borsa Istanbul 2016).

The day of the week effect is one of the market anomalies that has tendency to show more performance, and may give possibility for investors to make extra stock returns. Market anomalies have been reported in the developing markets, as well as in the developed markets. Getting high stock returns and predicting the behaviour of stock prices are important for investors and subjects of studies. One of the methods of the predicted returns is the detection of the day of the week effects in the stock markets. “Week-day effect” is a specific anomaly in the behaviour of asset prices and financial indexes (Carlucci, Júnior, Lima, & Gaio, 2014). In the literature section the studies which were found the existence of the day of the week effects are indicated.

Literature

The existence of the day of the week effect was found from 1950’s to 1970’s for Standard & Poor’s Index (Cross, 1973; French, 1980; Gibbons & Hess, 1981; Keim & Stambaugh, 1984; Lakonishok & Levi, 1982; Rogalski, 1984). Additionally, in later studies, the day of the week effect was tested for different markets and periods. These studies were grouped according to markets.

The day of the week effect in international markets

The day of week effect was investigated previously in various studies which are summarized in Table 1.

Table 1 Summary of the studies about the day of the week effect in international markets
Full size table

A study by Bayar and Kan Another study (Bayar & Kan, 2012) investigated the day of the week effects in stock market returns denominated in both U.S. dollars and local currencies in most of the nineteen countries for the period from July 1993 to July 1998. In local currency terms, the highest returns were on Tuesday and then Wednesday; the least returns were on Thursday and then Friday. In U.S. dollar currency, the highest returns occurred on Wednesday and then Tuesday. The lowest returns were found towards the end of the week, on Thursday and then Friday (Bayar & Kan, 2012).

The day of the week effect in Turkey’s market

The studies that examined the day of the week effect in Borsa Istanbul in Turkey is summarized in Table 2. The day of the week effect existed for BIST-100, BIST-50, BIST-30 and sector indexes on several studies (Berk Oğuzsoy & Güven, 2003; Cinko & Avci, 2011; Demirer & Karan, 2002; Dicle & Hassan, 2007). These studies examined whether stock returns on the days of the week are equal or not. The day of week effects were found different days for example daily returns on Friday, Tuesday and Monday were the highest; daily returns on Monday, Thursday and Friday were the least return in BIST.

Table 2 Summary of the studies about the day of the week effect in Turkey’s market
Full size table

On the other hand some studies were found the day of the week effect for the years 1990s. Friday and Monday effects were observed for the period from 1988 to 1996 (Metin, Muradoglu, & Yazici, 1997), and stock returns were higher in the second part of week and lower in the first two days of the week from 1988 to 1999 (Bildik, 2001).

The objective of this paper is to examine Monday effect in stock index in Turkey with an updated and extended data set that contains daily stock returns from 2010 to 2014. Previous studies were conducted before 2008. Moreover, 289 companies in Borsa Istanbul classified according to sectors were analysed.

The day of the week effects were found to be varying in the previously mentioned studies. For this reason, this study aimed to reexamine the existence of the day of the week effect in BIST when Monday returns kept constant with a larger amount of data in the BIST.

Method

Research design

The method of least squares was used to obtain the coefficients of the returns of the days of the week. Descriptive statistics were used to define the features of stock returns in the analysis. In order to determine whether there were any significant differences between the means of stock returns, the one-way analysis of variance (ANOVA) was used. This method was used for re-examines the Monday effect in the U.S. stock market, using daily returns (Mehdian & Perry, 2001)

Daily returns of BIST were calculated from the Eq. 1.

$$ {\mathrm{R}}_{\mathrm{t}}\kern0.5em = \ln\ \left[{\mathrm{E}}_{\mathrm{t}}/{\mathrm{E}}_{\mathrm{t}\hbox{-} 1}\right] $$
(1)

Where Rt was daily logarithmic returns of the index at the time t, Et was daily closing values at the time t, Et-1 was daily closing values at the time t-1 in the equation.

Equation 2 was used to analyse for there were any significant differences between the returns in different days of the week in BIST.

$$ {\mathrm{R}}_{\mathrm{t}}\kern0.5em = {\upbeta}_0\kern0.5em + {\upbeta}_1\mathrm{Tuesday} + {\upbeta}_2\mathrm{Wednesday} + {\upbeta}_3\mathrm{Thursday} + {\upbeta}_4\mathrm{Friday} + {\mathrm{E}}_{\mathrm{i}} $$
(2)

Where Rt was daily returns of the index, β0 was coefficient of fixed variable in the model that was coefficient of Monday, β1−β4 were coefficient of regression for the days, the days of Tuesday, Wednesday, Thursday and Friday were dummy variables in the equation

The t-tests were performed to check the means of other day returns which were statistically different from Monday returns. If there were any differences on Monday, this would contradict with the efficient-market hypothesis. The natural logarithm was used to eliminate the negative effect of extreme value in the data set. The hypotheses in the research are stated below, and sub-hypothesis is showed Table 3.

Table 3 The sub-hypothesis of the t-test in the analysis
Full size table

Data design

In the analysis, listed companies in equity market on the BIST were used in the period of 02.01.2010-22.10.2014. Stock returns of 289 listed companies were calculated according to daily historical stock prices of companies and were classified according to the sectors.

Results and Discussion

The daily stock returns of 289 companies that were constantly traded between 02.01.2010 and 22.10.2014 were studied, and it was analysed if the other days of the week had any effects on Monday when the daily stock returns of Monday were kept constant. The results of the statistical analysis indicated that, the day of the week effects in BIST were found. The stock returns on Monday were affected by the other days. These effects varied according to stocks that were showed in Appendix 1. The average returns of stocks on Monday had all days effects.

A group of stocks which are significant had the all-day effects on Monday stock returns. The closing day returns of these stocks were presented according to the sectors (Table 4). Monday returns were affected negatively by all other days for these stocks that were industrial textile, media, ceramics, iron steel and tourism. All day effect that was found for seven stocks was not common for all stocks. According to the analysis, there were the days of the week effects which are significant on the average stock returns and the effect of each day on stock returns were different (Table 5). According to statistically significant results, 7% of the stocks have the day of effect on Tuesday, 15% on Wednesday, 19% on Thursday, and 17% on Friday.

Table 4 The seven stocks that have the all-day effects which are significant on Monday mean returns
Full size table
Table 5 The quantity of stocks which have the day effects on stock returns
Full size table

The day of the week effects on the stock returns were grouped according to the sectors, and the number of these stocks were classi fied as negative or positive and showed in Table 6. There were negative effects on the sectors of automotive & parts, holdings, cement and concrete, iron steel, industrial textile, ceramics, tourism, livestock, and retail trader.

Table 6 The quantity of stocks which had the day of the week effects on the stock returns based on the sectors
Full size table

Conclusion

This study examines the Monday effect in listed companies in Borsa Istanbul that is one of the stock market of Turkey. The day of the week effects were found to be varying in the previously mentioned studies. For this reason, this study aimed to re-examine the existence of the day of the week effect in BIST when Monday returns kept constant with a larger amount of data in the BIST. This study covers all companies continuously listed in BIST and it is limited for the 5 years from 2010 to 2014. It is aim to present a statistical evidence of the day of the week effect in Turkey’ market. The results of the statistical analysis of the daily returns on Monday were affected by the other days. These effects were mostly negative and varied according to the stocks and sectors. The sector of food showed positive effects on Monday returns compared to the other sectors on Tuesday, Thursday and Friday. On the other hand, positive effects were found on Monday returns on Tuesday and Thursday. These positive effects were not in the same sectors except for the sector of food. The sectors of airlines & ground handling services, construction, petroleum, pharmaceutical and health, furniture, integrated textile, and others were found to have the least effects. The sectors of automotive & parts, holdings, cement & concrete, and industrial textile were found to have the most effects. These statistical results confirmed the day of the week effects on Monday return for the period from 2010 to 2014 in the BIST. The results of the present study show that the day of the week effect, which is one of the indicator of inefficient stock market, may give possibility to estimate the stock returns in the Turkish emerging stock market. Furute research could be determined the relation between company size and the day of week effect.