Professional Documents
Culture Documents
Khokan Bepari
Faculty of Agricultural Economics & Rural Sociology
Bangladesh Agricultural University
Mymensingh -2202, Bangladesh
E-mail: khokan552@yahoo.com
Abstract
The presence of the seasonal or monthly effect in stock returns has been reported in
several developed and emerging stock markets. This study investigates the existence of
seasonality in return series of Dhaka Stock Exchange (DSE) of Bangladesh. The study uses
the monthly return data of the DSE all share price index (DSE All Index) for the period
from 1993 to 2006 for the analysis. After examining the stationarity of the return series, we
specify a “combined regression-time series model” with dummy variable for months to find
the monthly effect in stock returns in DSE. The results confirm the existence of seasonality
in stock returns in DSE but do not support the “tax-loss-selling” hypothesis. Instead of
“July effect” (January effect), we find an “April effect” in DSE. The results of the study
invalidate the paradigm of the efficient market hypothesis in DSE meaning that, investors
can time their share investments to improve returns.
Introduction
The topic of capital market efficiency is one of the most searched areas in finance. Following Fama
(1965), number of studies was conducted to test the efficient market hypothesis (EMH). Recently,
however, researchers have collected evidence against the EMH. One of the significant anomalies of
EMH is the seasonal effect. Testing for a seasonal effect in monthly returns has been given
considerable attention in the literature.The existence of the seasonal effect negates the weak form of
the EMH and implies market inefficiency. If the seasonal effects are prominent and systematic in the
stock markets, then investors can have useful clues regarding their investment decisions and
speculators and portfolio managers can engage in playing games in derivative markets such as futures,
options, and mutual funds portfolio rebalancing. The different patterns identified in stock returns
include the day of the week effect, the day of the month effect, the end of the year effect, etc. French
International Research Journal of Finance and Economics - Issue 24 (2009) 168
(1980), Gibbons and Hess (1981), Keim (1983), Gultekin and Gultekin (1983), Harris (1986), Ariel
(1987), Broca (1992), etc., report these seasonal patterns.
Several studies have investigated the seasonal behavior of monthly stock market returns of a
number of countries. But most of the studies were conducted on developed countries. The “January
effect” and the “tax-loss –selling” hypotheses are two important hypothesis tested in the literature.
Increasing attention is being accorded by the capital market analysts, portfolio managers, finance
professionals and researchers to the emerging capital markets (ECMs). No such study has, thus far,
been conducted on the DSE return series of Bangladesh. The tax system in Bangladesh differs from
that of the USA and many other developed and developing countries in that the tax year ends in June in
Bangladesh and not in December like in the USA. Thus the “January effect” of most of the stock
markets may take “July effect” in DSE. The taxpayers have to pay capital gain tax on the sale of
shares. The capital losses can be set off against the capital gains. The ‘tax-loss-selling’ hypothesis may
provide an explanation, different from others, for the seasonality in stock returns in Bangladesh.
Information hypothesis (Kiem, 1983) may provide yet another explanation for the seasonality in stock
return.
The objective of this study is to investigate the existence of seasonality in stock returns in
Dhaka Stock Exchange (DSE).We use monthly closing share price data of the DSE all share price
index (DSE All) from 1993 to 2006 for this purpose. We use a combined regression –time series model
(Pindyck and Rubinfeld, 1998) with dummy variables for months to test the existence of seasonality in
stock returns. The results of the study confirmed the monthly effect in stock returns in DSE, but runs
counter to the ‘tax-loss selling’ hypothesis. Instead, we find an “April effect”. These findings have
important implications for financial managers, financial analysts and investors for devising investment
strategies in DSE.
of predicting the next 12-months' returns. A summary of earlier empirical work in this area can be
found in Kendall (1953).
While empirical evidences in this line are abounding on developed countries, research in
emerging capital market has only recently surfaced. Maghyereh (2003) using the standard GARCH,
exponential GARCH (EGARCH) and the GJR models for Amman Stock Exchange (ASE) of Jordan
found no evidence of monthly seasonality as well as January effect in the ASE returns. Pandey (2002)
and Lazar et al.(2005) using data from Bombay Stock Exchange’s Sensitivity Index document the
monthly effect in the stock returns in India. Their results confirm the existence of seasonality in stock
returns in India consistent with the ‘tax-loss selling’ hypothesis. Alagidede and Panagiotidis (2006)
examined both the day of the week and month of the year in the stock returns in Ghana employing
rolling techniques to asses the affects of policy and institutional changes thereby allowing deviations
from the linear paradigm. Contrary to a January return pattern in most markets, they find an April
effect for Ghana’s stock market. Al- Saad and Moosa (2005) applying structural time series model
incorporating stochastic dummies find that seasonality is present in the monthly stock returns derived
from a general index of the Kuwait Stock Exchange but it is deterministic as implied by the constancy
of the monthly seasonal factors over the sample period. Moreover, they find seasonality to take the
form of a July effect, as opposed to the better-recognized January effect. Doran et al.(2008.) find using
data form Chinese stock markets find that Chinese stock markets as a whole and highly volatile
Chinese stocks in particular outperform at the turn of the Chinese New Year, but not in January.
Although evidence contrary to the weak form of efficiency in the DSE has been documented in prior
studies (see, for instance, Rahman & Hossain, 2006; and Mobarek, & Keasey, 2000.), there is no
empirical evidence regarding the seasonality in the DSE, the premier capital market of Bangladesh.
The present study fills the gap.
Methodology
The seasonal effect is easily detectable in the market indices or large portfolios of shares rather than in
individual shares (Boudreaux, 1995). This study analyses monthly returns of the DSE All Index from
1993 to 2006. We measure stock return as the continuously compounded monthly percentage change in
the share price index as shown below:
rt = (ln Pt -ln Pt-1) ҳ 100 (1)
We first determine whether the DSE return series is stationary. One simple way of determining
whether a series is stationary is to examine the sample autocorrelation function (ACF) and the partial
autocorrelation function (PACF). We also use a formal test of stationarity, that is, the Augmented
Dickey-Fuller (ADF) test.
We use a month-of-the-year dummy variable for testing monthly seasonality. The dummy
variable takes a value of unity for a given month and a value of zero for all other months. We specify
an intercept term along with dummy variables for all months except one. The month of June, the
omitted month, is our benchmark month because tax year in Bangladesh ends in June. Thus, the
coefficient of each dummy variable measures the incremental effect of that month relative to the
benchmark month of June. The existence of seasonal effect will be confirmed when the coefficient of
at least one dummy variable is statistically significant (Pandey, 2002). Our initial model to test the
monthly seasonality is as follows:
y t = α1 + α2Jan + α3Feb + α4Mar+ α5Apr+ α6May + α7Jul+ α8Aug+ α9Sep+ α10Oct+ α11Nov
+ α12Dec + ε t (2)
The intercept term α1 indicates mean return for the month of June and coefficients α 2 …α 12
represent the average differences in return between June and each other month. These coefficients
should be equal to zero if the return for each month is the same and if there is no seasonal effect. εt is
the white noise error term. This approach, however, may be flawed because the residuals may have
serial correlation.
International Research Journal of Finance and Economics - Issue 24 (2009) 170
To deal with this problem we improve upon Equation (2) by constructing an ARIMA model for
the residual series µt. We then substitute the ARIMA model for the implicit error term in Equation (2)
to form a combined regression –time series model (Pindyck and Rubinfeld, 1998). The transfer
function model (Pandey, 2002 & Lazar et al, 2005 use similar model) is as follows:
y t = α1 + α2Jan + α3Feb + α4Mar+ α5Apr+ α6May + α7Jul+ α8Aug+ α9Sep+ α10Oct+ α11Nov
+ α12Dec + φ (B) θ (B)ηt
−1
(3)
where ηt is a normally distributed error term and it may have different variance from εt. We also check
for the ARCH effect in residuals.
Table 1: Descriptive statistics of the DSE all share share price index monthly return (1994-2007)
Month Mean Median Maximum Minimum Std. Dev. Skewness Kurtosis Jarque Bera Prob Obs
Jan 0.029 -0.61 34.228 -19.915 12.77 1.858 6.36 11.52 0.00 11
Feb -1.804 0.005 2.171 -9.793 3.52 -1.215 1.35 3.94 0.13 11
Mar -0.456 0.024 12.991 -32.564 31.66 0.002 5.308 3.11 0.15 14
Apr -5.835 -2.55 3.411 -24.07 8.76 -1.154 0.488 6.79 0.03 14
May 4.289 3.318 21.830 -8.302 8.05 0.578 0.491 4.45 0.10 14
Jun 3.943 4.12 10.914 -7.448 5.04 -0.623 0.453 4.68 0.10 14
Jul 1.003 0.288 17.795 -12.105 8.36 0.262 0.129 4.97 0.08 14
Aug 0.049 0.358 15.682 -15.712 8.00 -0.417 1.07 2.58 0.30 14
Sep -1.888 1.607 30.181 -5.985 8.81 2.332 6.306 19.06 0.00 14
Oct 5.34 1.656 56.997 -13.041 16.22 2.687 9.016 37.96 0.00 14
Nov 0.644 -0.63 24.173 -11.541 8.84 1.469 3.116 5.05 0.07 14
Dec -1.272 0.043 5.112 -26.995 7.98 -2.89 9.523 44.31 0.00 14
1993- 0.881 0.053 56.997 -32.564 9.86 1.231 8.298 230.39 0.00 162
2006
0
1 19 37 55 73 91 109 127 145 163
10 28 46 64 82 100 118 136 154
Months
Figure 2: Auto correlation & partial autocorrelation function of the DSE Gen index monthly return series
0.2
Autocorrelationandpartial
autocorrelationcoefficient
0.1
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
-0.1
-0.2
-0.3
La gs
We next examine the residuals obtained from the estimation of Equation (2). After
experimenting, we fit the ARIMA (5, 0, 5) model to the residual series. The results of the model are
given in Table 4. Actual and fitted residual series of Equation (2) are shown in figure 3. The Ljung-
Box Q-statistic to order of 36 is 27.302 & is insignificant with ρ- value of 0.851. Thus, we can
conclude that the residuals of the ARIMA (5,0,5) model are white noise.
60
50
40
30
20
Actual
10
Fitted
0
-10
-20
-30
-40
We next combine the ARIMA (5, 0, 5) model for the residual series of Equation (2) with the
regression model ((Equation (2)) to form a combined regression- time series model and estimate all
parameters simultaneously as given in Equation (3). The results of the estimation of Equation (3) are
given in Table 5. The R2 is.371 and the D-W statistic of 1.91 is very close to 2. The sample
autocorrelations for the residuals of the model (Equation (3)) are almost zero. Further, the Ljung- Box
Q-statistic is mostly insignificant. Thus, the residuals of the model are white noise. The residuals do
not show conditional autoregressive heteroskedasticity. A Lagrange Multiplier (LM) test for the
presence of the ARCH effects in the residuals (F-statistic of 0.573 and ρ-value of 0.718) reveals no
such effects.
We note from Table 5 that the coefficients of the dummy variables for the months of April,
August and September are statistically significant. The coefficient for the benchmark month of June is
2.105. Except for the month of May, coefficients are lower for all months as compared to the
benchmark month of June.
The coefficients for the intercept term i.e. for the month of June not statistically significant may
be interpreted to reject the “tax-loss-selling” hypothesis because tax year in Bangladesh ends in June.
Further evidence is found to reject the “July effect” (the so called “January effect” in countries where
tax year ends in December) as the coefficient for the month of July is not statistically significant.
However we find “April effect” in the DSE return series. The coefficient for the month of April is
negative and statistically significant at 5% significance level. One explanation for this may be that as
by the month of April most of the companies declare dividend and hold AGM, investors start selling
shares post- dividend and the share prices dip causing downward pressure in the index and returns.
The coefficients of dummy variables for the months of November and December are not
statistically significant mean that there is no year end effect in DSE return series. The coefficients of
dummy variable for the month of August and September are statistically significant. This could result
from several social, economic and political factors (such as flood that usually comes during August and
September) that may cause changes in the macroeconomic fundamentals (floods may slow down
economic activities and industrial production) affecting the stock market activities
Conclusions
This study investigates the nature of seasonality in the monthly stock returns of DSE All share price
index (DSE All). Descriptive statistics point to large dispersions in the monthly returns. Using a
combined regression-time series model incorporating dummy.variables for months, the analysis reveals
that seasonality is present in DSE All Index monthly return series. Contrary to the “Tax- loss –selling
“and “January effect” hypothesis reported in developed countries, the study finds an “April effect” in
the DSE All return series. The study further documents a statistically significant coefficient for the
month of August and September and it may be attributed to “information hypothesis”. The findings of
the study indicate that stock returns in DSE are not entirely random. As a consequence, investors may
have opportunities to improve their returns by timing their investments. However, further research is
needed involving other stock indices of DSE such as DSE General index and DSE 20 index before
making any firm conclusion in this regard.
175 International Research Journal of Finance and Economics - Issue 24 (2009)
References
[1] Aggarwal,R., Ramesh P.Rao and Takto Hiraki,“Regularities in Tokyo Stock Exchange Security
Returns: P/E, Size and Seasonal Influences,” Journal of Financial Research, vol. 13, (Fall
1990), 249-263.
[2] Alagidede, Paul and Theodore Panagiotidis (2006), “Calendar Anomalies in an Emerging
African Market: Evidence from the Ghana Stock Exchange," Discussion Paper Series 2006_13,
Department of Economics, Loughborough University, revised Jun 2006.
[3] Al-Saad, Khalid & Imad A. Moosa, 2005. "Seasonality in stock returns: evidence from an
emerging market," Applied Financial Economics, Taylor and Francis Journals, vol. 15(1), pages
63-71, January.
[4] Ariel, Robert A., “A Monthly Effect in Stock Returns,” Journal of Financial Economics, Vol.
18, (March 1987), 161-174.
[5] Berges, A., J. McConnell, and G. Schlarbaum, “An Investigation of the Turn-of-the-Year
Effect,the Small Firm Effect and the Tax-Loss Selling Pressure Hypothesis in Canadian Stock
Returns,” Journal of Finance, 39, (March1984), 185-192.
[6] Boudreaux, Denis O., “The Monthly Effect in International Stock Markets: Evidence and
Implications”, Journal of Financial and Strategic Decisions, Vol. 8, No. 1, (1995), 15-20.
Paper Available at SSRN: http://ssrn.com/abstract=876551
[7] Brown, Lawrence D. and Luo, Liyu, "The Predictive Value of the Signs of January Returns:
Evidence of a New January Effect" (March 30, 2004). Available at SSRN:
http://ssrn.com/abstract=525162
[8] Brown, P., D.B. Keim, A.W. Keleidon, and T.A. Marsh “Stock Return Seasonalities and the
Tax-Loss-Selling-Hypothesis: Analysis of the Arguments and Australian Evidence,” Journal of
Financial Economics, 12, (June1983), 105-127.
[9] Chan, K.C.(1986) Can Tax-Loss Selling Explain the January Seasonal in Stock Returns?
Journal of Finance, 41, 1115-1128.
[10] Doran, James & Jiang, Danling & Peterson, David, 2008. "Gambling in the New Year? The
January Idiosyncratic Volatility Puzzle," MPRA Paper 8165, University Library of Munich,
Germany
[11] Fama, Eugene F., “The Behavior of Stock Market Prices,” Journal of Business, Vol. 38,
(January 1965), 34-105.
[12] French, K.R. “Stock returns and the week end effect”. Journal of Financial Economics. (1980),
8:55-69
[13] Gibbons, M., and P. Hess (1981) Day of the Week Effects and Asset Returns. Journal of
Business 54: 579–596.
[14] Gultekin, M.N., and N.B. Gultekin, “Stock Market Seasonality: International Evidence,”
Journal of Financial Economics, 12, (December 1983), 469-481.
[15] Kendall, M.G., (1953), ‘The Analysis of Economic Time Series, Part I. Prices’, Journal of the
Royal Statistical Society, vol. 96, pp. 11-25
[16] Harris, L. “A Transaction Data Study of Weekly and Intradaily Patterns in Stock Returns”
Journal of Financial Economics. (1986) 101–116.
[17] Keim, D.B., “Size-Related Anomalies and Stock Return Seasonality: Further Empirical
Evidence,” Journal of Financial Economics, 12, (June 1983), 13-32.
[18] Lazar, D., Priya, Julia and Jeyapaul, Joseph, "SENSEX Monthly Return: Is there Seasonality?”
Indian Institute of Capital Markets 9th Capital Markets Conference, 2005
[19] Lewis, Mario, “Stock Market Anomalies: A Re-Assessment Based on the U.K. Evidence,”
Journal of Banking and Finance, Vol. 13, (1989), 675-696.
[20] Maghyereh, Aktham Issa (2003), "Seasonality and January Effect Anomalies in the Jordanian
Capital Market". Available at SSRN: http://ssrn.com/abstract=441081 or DOI:
10.2139/ssrn.441081
International Research Journal of Finance and Economics - Issue 24 (2009) 176
[21] Mobarek, Asma and Keavin Keasey. “Weak- form Market Efficiency of an Emerging Market:
Evidence from Dhaka Stock Market of Bangladesh.” Presented at ENBS Conference, Oslo,
2000. Available online at: http://www.bath.ac.uk/cds/enbs-papers-pdfs/mobarek-new.pdf
[22] Officer, R.R., “Seasonality in Australian Capital Markets: Market Efficiency and Empirical
Issues,” Journal of Financial Economics, Vol. 2, (March 1975), 29-52.
[23] Pandey, I.M. “Seasonality in the Malaysian Stock Market: 1992-2002”. Journal of Financial
Management and Analysis. 2002,15(2):pp.34-37
[24] Pandey I M, 2002. "Is There Seasonality in the Sensex Monthly Returns?" IIMA Working apers
2002-09-08, Indian Institute of Management Ahmedabad, Research and Publication
Department.
[25] Pindyck, Robert S. and Daniel L. Rubinfeld, Econometric Models and Economic Forecasts,
New York: McGraw-Hill, International Edition, (1998).
[26] Rahman, Shofiqur and Hossain, Farhad Farhad, "Weak-Form Efficiency: Testimony of Dhaka
Stock Exchange". Journal of Business Research, Vol. 8, pp. 1-12, 2006 Available at SSRN:
http://ssrn.com/abstract=940811
[27] Raj M.; Thurston D. “January or April? Tests of the turn-of-the-year effect in the New Zealand
stock market” Applied Economics Letters. Volume 1, Number 5, 1 May 1994, pp. 81-83(3).
[28] Reinganum, M.R. “The Anomalous Stock Market Behavior of Small Firms in January
Empirical Test for Year-End Tax Effect,” Journal of Financial Economics, 12, (June 1983), 89-
104.
[29] Rozeff, Michael S., and William R. Kinney, “Capital Market Seasonality: The Case of Stock
Market Returns, “Journal of Financial Economics, Vol. 3, (October 1976), 376-402.
[30] Smirlock, Michael, and Laura Starks, “Day of the Week and Intraday Effects in Stock Returns,”
Journal of Financial Economics, Vol. 17, (1986), 197-210.
[31] Tinic, Seha M., and Richard R. Wesst, “Risk and Return: January and the Rest of the Year,”
Journal of Financial Economics, Vol. 13, (December 1984), 561-574.
[32] Tinic, Seha M., Giovanni Barone-Adesi and Richard R. West, “Seasonality in Canadian Stock
Prices: A Test of the ‘Tax-Loss Selling’ Hypothesis,” Journal of Financial and Quantitative
Analysis, Vol. 22, (1987), 51-64.
[33] Wachtel, S.B., “Certain Observation on Seasonal Movement in Stock Prices,” Journal of
Business, 15, (1942), 184-193.