THE DAY-OF-THE-WEEK EFFECT IN SELECTED BALANCED FUNDS IN THE PHILIPPINES

Transcription

1 THE DAY-OF-THE-WEEK EFFECT IN SELECTED BALANCED FUNDS IN THE PHILIPPINES ABSTRACT Catherine Kalayaan S. Almonte The returns of selected balanced funds in the Philippines were tested for conformity with the dayof-the-week effect. Annual and cumulative analyses (Almonte, 2012) were conducted for the years 2006 to The results of the Kruskal-Wallis Test indicate that, for at least one fund, Thursday returns are higher than the returns of other days in the week. Therefore, similar with the implication of maximizing earnings (Almonte, 2004; 2012), an investor might one to time his or her redemptions (i.e. selling transactions) on a specific day where returns are high (in this case, on a Thursday). Keywords: market efficiency, day-of-the-week effect, mutual funds 1. INTRODUCTION Rational investors would like to get the best possible return out of their investment (Markowitz, 1952). While some actively invest, others prefer the passive form of investing. Placing money in mutual funds is a form of passive investing because mutual funds allow investors the luxury of being diversified at a relatively low amount of investment, and the burden of selecting what securities to purchase and when to trade them lies on an investment professional (Mishkin & Eakins, 2006). Given that fund investors do not have any control with regards to the specific assets to invest in and at what price to trade them, what they do have control over are: (1) what particular fund(s) to invest in, (2) the amount of investment to place, and (3) when to buy (or sell) shares of the fund. Thus, this paper examines if it is possible to time one s investment (Almonte, 2004) in mutual funds by determining the existence of the day-of-the-week effect Framework and Literature The day-of-the-week effect is a phenomenon wherein Friday stock returns are greater compared to Monday returns (Weekend Effect, n.d.; Hirt & Block, (2012)). Using Almonte s (2012, p ) definition,... a day-of-the-week effect is defined as at least one [40] trading day in a week wherein returns for that day are statistically significantly different from at least one other day (note: this definition is also consistent with Almonte, 2004). This phenomenon challenges Fama s (1970) weak form of market efficiency (Aly, Mehdian, & Perry, 2004; Almonte, 2004; 2012). Previous research provides different results regarding the existence of the phenomenon (Almonte, 2004; 2012): a study about Egypt s stock market supported Fama s (1970) theory (Aly et al., 2004) while other studies of different Asian equities markets challenged it (Poshakwale, 1996; Almonte, 2004; 2012; Hussain, Hamid, Akash, & Khan, 2011; Selvakumar, 2011). Furthermore, another paper by Muhammad & Rahman (2010) on the Malaysian stock market concluded that Monday returns were lower compared to other trading days. Publicly available studies about Philippine mutual funds are relatively few. Two notable works are: (1) an industry study (Valderrama & Bautista, 2003) and (2) a paper that distinguished mutual funds from common trust funds (Valderrama, 2004). To date, no published study has been found about the day-ofthe-week effect as applied to balanced mutual funds. As such, this paper determines the presence of the day-of-the-week effect as applied to Philippine balanced mutual funds. Since balanced funds partially comprise of stocks

2 (Mishkin & Eakins, 2006; Philippine Investment Funds Association website, a), it is expected that the day-of-the-week effect will be present similar to the results generated by Almonte s (2004; 2012) researches on equities Basics of Mutual Funds A Mutual fund is an investment company that pools the funds of many individual and institutional investors to form a massive asset base (Philippine Investment Funds Association website, c). There are five classes of funds according to investment objective: (1) money market funds, (2) bond funds, (3) hybrid funds, (4) equity funds, and (5) index funds (Mishkin & Eakins, 2006). Money market funds invest in money market securities, bond funds invest in bonds, hybrid fund invest in stocks and bonds, equity funds invest in stocks, and index funds invest in securities that mimic a particular index (Mishkin & Eakins, 2006). Investors in mutual funds buy and sell shares of the fund based on the fund s net asset value per share (NAVPS; Mishkin & Eakins, 2006) An Overview of Mutual Funds in the Philippines Philippine mutual funds have been around since the 1950s but their existence was almost threatened because of very high fees charged to investors, lack of rules and regulations, and an unstable political climate (Philippine Investment Funds Association website, b). Furthermore, a law pertaining to investment companies, R.A or the Investment Company Act (ICA), has been established for years but it was not until 1989 when the Securities and Exchange Commission (SEC) came up with its implementing rules and regulations (Philippine Investment Funds Association website, b). The improvements in the law gave new life to the industry and the number of funds has been growing for the past several years [41] (Philippine Investment Funds Association website, b). Currently, local funds are categorized according to investment objective (consistent with the fund classification mentioned by Mishkin & Eakins, 2006). The classes of funds are: (1) money market funds, (2) bond funds, (3) balanced funds, and (4) stock funds (Philippine Investment Funds Association website, a). Notably, there are minor distinctions from the Mishkin & Eakins (2006) classification: (1) hybrid funds are called balanced funds (although these funds are not strictly balanced in the sense that 50% is allocated for bond investments while the remaining 50% is allocated for equity investments) and (2) index funds are merged with stock funds (Philippine Investment Funds Association website, a). Based on the Philippine Investment Funds Association website ( 2012a): (1) there are 48 mutual funds in the Philippines (money market funds account for 10.42%, bond funds at 43.75%, balanced funds at 27.08%, and stock funds at 18.75%); (2) Peso-denominated funds account for 66.67% while foreign currency denominated funds account for 33.33% of total funds; and (3) funds denominated in both Peso and foreign currency (a) 47.62% of bond funds are Peso-denominated while 52.38% are foreign currency denominated, (b) 69.23% of balanced funds are Peso-denominated while 30.77% are foreign currency denominated, and (c) 88.89% of stock funds are Peso-denominated while 11.11% are foreign currency denominated. H1 2. HYPOTHESIS The day-of-the-week effect is present in selected Philippine balanced funds. 3. METHODOLOGY Out of the 13 balanced funds in the Philippines, nine were denominated in Philippine Peso and four in foreign currency (Philippine Investment Funds Association website, a). Three Peso-denominated balanced

3 funds were selected for the study: (1) Philam Fund,, (2) Sun Life Prosperity Balanced Fund,, and (3) MFCP Kabuhayan Fund, It should be noted that ATR KimEng took over MFCP s management on August 31, 2003 and the fund was re-named ATRKE Philippine Balanced Fund, on September 12, 2011 (ATR KimEng Asset Management website, ails.aspx?id=mfcp, n.d.). The mutual fund data was sourced from Technistock. Daily returns, based on the NAVPS, were computed. As with Almonte (2004; 2012), (1) codes were used to represent a particular trading day: 1 = Monday, 2 = Tuesday, 3 = Wednesday, 4 = Thursday, and 5 = Friday; and (2) the daily returns were tested if they conformed with a normal distribution (since the returns of the three funds for all periods do not follow a normal distribution (Tables 1A and 1B), the day-of-the-week effect was tested via the Kruskal-Wallis Test). Just like in Almonte (2012), (1) the analyses were conducted annually and cumulatively; (2) XLSTAT 2011 was used for all statistical calculations; (3) if the hypothesis was supported, another test was conducted (Steel-Dwass-Critchlow-Fligner Procedure) to determine which day substantially differed from another day; and (4) selected summary statistics were examined the method of evaluation was loosely comparable to what was made by Agathee (2008). This study was generally patterned after the papers done by Almonte (2004, 2012). Table 1. A. TEST FOR NORMALITY Jarque-Bera Test at = Philam Fund, JB (Observed Value) JB (Critical Value) DF p-value < < < JB (Observed Value) JB (Critical Value) DF p-value < < < < MFCP Kabuhayan Fund, JB (Observed Value) JB (Critical Value) DF p-value < < < [42]

4 Table 1. B. TEST FOR NORMALITY Jarque-Bera Test at = to to to to 2010 Philam Fund, JB (Observed Value) JB (Critical Value) DF p-value < < < < JB (Observed Value) JB (Critical Value) DF p-value < < < < MFCP Kabuhayan Fund, JB (Observed Value) , , , JB (Critical Value) DF p-value < < < < RESULTS AND ANALYSIS Based on Table 2A, for Philam Fund,, Mondays, Tuesdays, and Wednesdays had the most number of lowest mean returns (three out of five years) while Thursdays had the most number of highest mean returns (four out of five years). For Sun Life Prosperity Bond Fund,, Tuesdays has the most number of lowest mean returns (four out of five years) while Thursdays had the most number of highest mean returns (four out of five years). For MFCP Kabuhayan Fund,, Mondays, Tuesdays, and Wednesdays had the most number of lowest mean returns (two out of five years) while Thursdays had the most number of highest mean returns (three out of five years). For 2006, all three funds had Tuesdays and Wednesdays as the days with the least mean returns while Thursdays had the highest mean returns. For 2007, all funds indicated Fridays as the day with the lowest mean returns while Thursdays had the highest mean returns. For 2008, all funds specified Thursdays as the day with the lowest mean returns while Wednesdays had the highest mean returns. For 2009, two out of three funds identified Mondays and Tuesdays with the lowest mean returns while all funds specified Friday as the day with the highest mean returns. For 2010, all funds had Tuesdays as the day with the lowest mean returns while Thursdays had the highest mean returns. Based on Table 2B, for Philam Fund,, Mondays had the most instances of lowest mean returns (four out of four times) while Thursdays had the most instances of highest mean returns (four out of four times). For Sun Life Prosperity Balanced Fund,, Mondays, Wednesdays, and Fridays had the most instances of lowest mean returns (four out of four times) while Thursdays had the most instances of highest mean returns (four out of four times). For MFCP Kabuhayan Fund,, Mondays, Tuesdays, and Wednesdays had the most instances of lowest mean returns (three out of four times) while Thursdays had the most instances of highest mean returns (four out of four times). For 2006 to 2007, all funds had Wednesdays as the day with the lowest mean returns while Thursdays had the highest mean returns. For 2006 to 2008, the three funds identified Mondays as the day with the lowest mean returns while Thursdays had the highest mean returns. For 2006 to 2009 and 2006 to 2010, all funds specified Mondays and Tuesdays as the days with the lowest mean returns while Thursdays had the highest mean returns [43]

5 Table 2. A. SELECTED SUMMARY STATISTICS Day-of-the-Week Effect Philam Fund, Mean Standard Deviation Mean Standard Deviation Mean Standard Deviation Mean Standard Deviation Mean Standard Deviation Mean Standard Deviation Mean Standard Deviation Mean Standard Deviation Mean Standard Deviation Mean Standard Deviation MFCP Kabuhayan Fund, Mean Standard Deviation Mean Standard Deviation Mean Standard Deviation Mean Standard Deviation Mean Standard Deviation [44]

6 Table 2. B. SELECTED SUMMARY STATISTICS Day-of-the-Week Effect 2006 to to to to 2010 Philam Fund, Mean Standard Deviation Mean Standard Deviation Mean Standard Deviation Mean Standard Deviation Mean Standard Deviation Mean Standard Deviation Mean Standard Deviation Mean Standard Deviation Mean Standard Deviation Mean Standard Deviation MFCP Kabuhayan Fund, Mean Standard Deviation Mean Standard Deviation Mean Standard Deviation Mean Standard Deviation Mean Standard Deviation [45]

7 The results presented on Table 3A indicate that only Philam Fund, and Sun Life Prosperity Balanced Fund, exhibited the phenomenon in As per Table 3B, only Sun Life Prosperity Balanced Fund, exhibited the day-of-the-week effect. The phenomenon was present during the following periods: 2006 to 2007, 2006 to 2009, and 2006 to Table 3. A. TEST FOR THE DAY-OF-THE-WEEK EFFECT Kruskal-Wallis Test at = Philam Fund, K (Observed Value) K (Critical Value) DF Asymptotic p-value (two-tailed) K (Observed Value) K (Critical Value) DF Asymptotic p-value (two-tailed) MFCP Kabuhayan Fund, K (Observed Value) K (Critical Value) DF Asymptotic p-value (two-tailed) Table 3. B. TEST FOR THE DAY-OF-THE-WEEK EFFECT Kruskal-Wallis Test at = to to to to 2010 Philam Fund, K (Observed Value) K (Critical Value) DF Asympto-tic p-value (two-tailed) K (Observed Value) K (Critical Value) DF Asympto-tic p-value (two-tailed) MFCP Kabuhayan Fund, K (Observed Value) K (Critical Value) DF Asympto-tic p-value (two-tailed) The data on Table 3C suggests that investors of Philam Fund, buy shares of the fund either on a Monday or a Wednesday (although Wednesday is preferred since it has the lowest mean rank of ) and sell on a Thursday (since it has the highest mean rank of ). On [46] the other hand, investors of Sun Life Prosperity Balanced Fund, are advised to buy on any day of the week except Thursdays. The preferred buying day appears to be Tuesday (given it has the lowest mean rank of ) while redemptions (i.e. sales) should preferably be

8 conducted during Thursdays (since this day has the highest mean rank of ). Table 3. C. SIGNIFICANCE OF THE-DAY-OF-THE-WEEK EFFECT FOR 2006 Steel-Dwass-Critchlow-Fligner Procedure; Bonferroni Correction Fund Philam Fund, Obs. Sum of Ranks Mean of Ranks Obs. Sum of Ranks Mean of Ranks 41 4, , , , , , , , , , Significant p-values and at 0.035, and at and at 0.033, and at 0.025, and at 0.013, and at According to Table 3D, investors of may invest on Mondays, Tuesdays, Wednesdays, and Fridays (although Fridays are preferred given its lowest mean rank of ) and sell on Thursdays (since it has the highest mean rank of ). Table 3E recommends that investors buy on Tuesdays (since it has the lowest mean rank of ) and sell on Thursdays (since it has the highest mean rank of ). Based on Table 3F, investors are advised to buy shares of the fund either on Tuesdays or Fridays (although Tuesdays are preferred given its lowest mean rank of ) and sell on Thursdays (given its highest mean rank of ). Table 3. D. SIGNIFICANCE OF THE-DAY-OF-THE-WEEK EFFECT FOR 2006 to 2007 Steel-Dwass-Critchlow-Fligner Procedure; Bonferroni Correction Fund Obs. Sum of Ranks Mean of Ranks 90 21, , , , , Significant p-values and at 0.047, and at 0.029, and at 0.033, and at Table 3. E. SIGNIFICANCE OF THE-DAY-OF-THE-WEEK EFFECT FOR 2006 to 2009 Steel-Dwass-Critchlow-Fligner Procedure; Bonferroni Correction Fund Obs. Sum of Ranks Mean of Ranks , , , , , Significant p-value and at [47]

9 Table 3. F. SIGNIFICANCE OF THE-DAY-OF-THE-WEEK EFFECT FOR 2006 to 2010 Steel-Dwass-Critchlow-Fligner Procedure; Bonferroni Correction Fund Obs. Sum of Ranks Mean of Ranks , , , , , Significant p-values and at 0.002, and at CONCLUSION As per the annual analysis, the dayof-the-week effect was present only in 2006 (and only for two funds, Philam Fund, and ), Mondays and Wednesdays were common buy days (since the values were generally low) while Thursday was the sell day (since the values were generally higher than the rest of the days in the week). The cumulative analysis revealed that only Sun Life exhibited the anomaly (during the periods 2006 to 2007, 2006 to 2009, and 2006 to 2010). Tuesday was the common buy day while Thursday was the common suggested day to redeem (i.e. sell). Based on the annual and cumulative analyses (Almonte, 2012), Thursdays appear to have higher than average values as compared to other trading days. The results are similar to what was concluded in Almonte (2004, 2012). Hence, the research hypothesis is supported. There appears to be a day-ofthe-week effect in selected balanced funds in the Philippines. 6. REFERENCES Agathee, U.S. (2008). Calendar effects and the months of the year: Evidence from the Mauritian stock exchange. International Research Journal of Finance and Economics, (14), Almonte, C.K.S. (2004). The day-of-the-week effect in the Philippine stock market January 3, 2000 to July 23, Retrieved January 7, 2012, from rd/pdf/papers/working%20paper%20series % pdf. Almonte, C.K.S. (2012). Calendar effects in the Philippine stock market. Proceedings of the International Conference on Business, Economics and Information Technology, Aly, H., Mehdian, S., & Perry, M.J. (2004). An analysis of the day-of-the-week effects in the Egyptian stock market. International Journal of Business, 9(3), Retrieved June 20, 2004, from ATR KimEng Asset Management website. (n.d.). Funds: ATRKE Philippine Balanced Fund. Retrieved January 29, 2012, from ddetails.aspx?id=mfcp. Fama, E.F. (1970). Efficient capital markets: A review of theory and empirical work. Journal of Finance, 25(2), Retrieved June 3, 2006, from Hirt, G.A. & Block, S.B. (2012). Fundamentals of investment management (Tenth Edition). USA: McGraw-Hill. Hussain, F., Hamid, K., Akash, R.S.I., & Khan, M.I. (2011). Day of the week effect and stock returns: (Evidence from Karachi Stock Exchange-Pakistan). Far East Journal of Psychology and Business, 3(1), Retrieved January 7, 2012, from V3N1P3.pdf. Markowitz, H. (1952). Portfolio selection. The Journal of Finance, 7(1), Retrieved on June 12, 2007 from Mishkin, F.S. & Eakins, S.G. (2006). Financial markets and institutions (Fifth Edition). USA: Pearson Addison-Wesley. Muhammad, N.M.N. & Rahman, N.M.N.A. (2010). Efficient market hypothesis and market anomaly: Evidence from day-of-the-week effect of Malaysian exchange. International Journal of Economics and Finance, 2(2). Retrieved July 22, 2012, from ments/027_emh_ijef.pdf. Philippine Investment Funds Association website. (2012a). Facts & figures: NAVPS performance (as of 1/27/2012). Retrieved January 28, 2012, from [48]

10 Philippine Investment Funds Association website. (2012b). History of Philippine mutual funds. Retrieved January 28, 2012, from ml. Philippine Investment Funds Association website. (2012c). Mutual funds 101: What is a mutual fund? Retrieved January 29, 2012, from Poshakwale, S. (1996). Evidence on weak form efficiency and day of the week effect in the Indian stock market. Finance India, X(3), Retrieved January 7, 2012, from /41/MARKET_EFFICIENCY.pdf. Selvakumar, D.S. (2011). Existence of certain anomalies in Indian stock market International Conference on Economics and Finance Research, IPEDR Vol. 4 (2011). Retrieved on January 7, 2012, from Valderrama, H.S. (2004). Philippine mutual funds and common trust funds: How different are they? Retrieved January 27, 2012, from udy.pdf. Valderrama, H.A.S. & Bautista, C.C. (2003). The development and performance of the Philippine mutual funds industry. Retrieved January 27, 2012, from pdf. Weekend Effect. (n.d.). Retrieved January 25, 2012, from [49]