Printer Friendly

Momentum anomaly in emerging stock markets: some empirical evidence from the Colombo Stock Exchange.

1. Introduction

Capital asset pricing model (CAPM) of Sharpe (1964), Lintner (1965), and Mossin (1966) states that expected returns on securities have positive linear relations with their beta coefficients. Thus the beta is the sole factor that explains the cross-section of expected returns. Though early studies by Black et al. (1972) and Fama and MacBeth (1973) provide evidence in favor of the CAPM, subsequent empirical studies found evidence against the CAPM. These findings are referred to as CAPM anomalies. Anomalies suggest that there are variables other than the beta which explain returns of securities in both cross-sectional and time series studies. The most important cross-sectional anomalies include the size effect, debt-to-equity (D/E) ratio anomaly, the earnings-to-price (E/P) ratio anomaly, book-to-market (B/M) ratio anomaly, cash flow to price (CF/P) ratio anomaly and Contrarian and momentum effects.

In recent years a number of researchers have provided evidence that cross-sectional stock returns are predictable based on past returns. For example, De Bondt and Thalar (1985, 1987) document that stocks which performed poorly over 3 to 5 year holding periods achieve higher returns in the next 3 to 5 years than stocks which performed well in the same period. After De Bondt and Thalar many researchers have examined the long-term contrarian strategies globally. (1) Jegadeesh (1990), Lehman (1990) and Chang et al. (1995) also report price reversals at monthly and weekly intervals. But perhaps the most puzzling results are the intermediate-horizon return continuation reported by Jegadeesh and Titman (1993). Forming portfolios based on the past 3 to 12 month returns they show that past winners on average continue to out-perform the past losers over the next 3 to 12 months.

Price momentum effect has been extensively studied in the US (see, for example, Conrad and Kaul (1998); Lee and Swaminathan (2000); Chordia and Shivakumar (2001) and Jegadeesh and Titman (2001) and in other developed markets (see, for example, Bildik and Gulay (2002); Chui et al. (2000); Geert Rouwenhorst (1998 and 1999); Nijman et al. (2002) and Shen et al. (2005)) and in certain other countries. Some of these studies are discussed in the next section. However, no study has examined this anomaly in the Sri Lankan context.

The main objective of this paper is to examine the medium-term momentum effect as a risk measurement variable that measures the cross-sectional variability of stock returns on the Colombo Stock Exchange (CSE). During the last two decades, the CSE has emerged as a stock market in South Asia employing the most sophisticated technologies with a considerable amount of foreign interest. During the period 1991 to 2009, the average percentage turnover attributed to foreigners was 36. During this period, in 1996 and 2008, turnover attributed to foreigners exceeded that attributed to the domestic investors. In 1996 foreigners accounted for approximately 55 per cent of the total turnover whereas in 2008 they accounted for 54 per cent of the total turnover. The above interest mainly emanated as a result of the tax incentives given by the Sri Lankan government to foreign nationals. The above figures indicate that empirical research relating to the CSE has implications for both domestic and foreign investors. However, at present there is a dearth of empirical research into many aspects of the CSE in which local as well as foreign investors may be interested. In this background, the main focus of this study is whether investors in the CSE can earn higher returns by employing certain portfolio strategies. Research into other aspects will be reported in a series of future papers.

The remainder of this paper is organized as follows. Section 2 reviews related literature, while basic methodology and data are described in Section 3. Section 4 contains empirical results for momentum and contrarian strategies. The last Section concludes the paper.

2. Literature Review

De Bondt and Thalar (1985) investigated return patterns over long periods of time and found that contrarian strategies were profitable over 3 to 5 year horizons. Thereafter, Jegadeesh (1990) and Lehmann (1990) provided evidence for the existence of short-term return reversals. In contrast, Jegadeesh and Titman (1993) uncovered that strategies which buy past period winner-stocks and sell past period loser-stocks (momentum strategy) generate significant positive returns (about 1 per cent per month) for 3 to 12 month holding periods. The extended study of Jegadeesh and Titman (2001) reconfirms that momentum effect is not a result of data mining effort. Further, Conrad and Kaul (1998), Lee and Swaminathan (2000), Chodia and Shivakumar (2002) have found significant momentum profits on the NYSE over holding periods ranging from 3 to 12 months.

Both momentum and contrarian strategies have also been found to work in international markets. Rouwenhorst (1998) examined stock returns of twelve European markets between 1980 and 1995 and found that an internationally diversified portfolio of past medium term winners outperformed a portfolio of medium term losers by 1 per cent per month. Similarly, Chui et al. (2000) found evidence supporting momentum profits in Asian markets except Japan and Korea. (2)

Shen et al. (2005) examined the performance of momentum strategies in eighteen developed capital markets using country indices instead of individual security returns and found momentum profits for medium-time horizons. Nijman, Swinkels, and Verbeek (2002) found momentum profits in eighteen European countries except for Sweden and Austria. Chui et al. (2000) examined the profitability of momentum strategies in eight different East and South East Asian Countries. They found positive momentum profits over the entire sample period for the whole sample except for two countries, Indonesia and Korea.

There is some empirical evidence on short and medium-term contrarian strategies too. Lehemann (1990) examined the profitability of short-term trading strategies on the NYSE and American Stock Exchange (AMEX). He found that considerable return reversals in one week's time to make short-period contrarian strategies profitable. Chang, Mcleavey and Ruhee (1995) examined the short-term abnormal returns to contrarian investment strategies applied to stocks listed on the Tokyo Stock Exchange (TSE). They found statistically significant contrarian profits only in the first year.

Wang et al. (2009) examined the impact of market states on the profitability of momentum strategies using weekly data from the Taiwan Stock exchange over the 10-year period 1997-2006. The results indicate that market conditions in the formation period are positively associated with the profitability of the momentum strategies.

Bildik and Gulay (2002) found significant contrarian profits on the Istanbul Stock Exchange. Their analysis of contrarian strategies showed that the holding period returns of past period losers outperform the past period winners in all 1 to 12 months strategies.

In most of the studies, researchers have imposed a time lag of one month between the end of the portfolio formation period and beginning of the holding period in order to avoid potential micro structure biases, thin trading problem and problems due to bid-ask spread (see, for example, Jegadeesg and Titman, 1993; Lee and Swaminathan, 2000; Nijman et al. 2002 and Chui et al., 2000). All of them found that the magnitude of the momentum effect increases when a time lag of one month was imposed between the formation and holding periods.

Momentum and Risk

Most of the authors have distinguished momentum effect from risk measurement by CAPM or Fama-French factor model. Jegadeesh and Titman (1993) report that CAPM fails to explain the momentum effect. They report that CAPM beta of a portfolio of past losers is higher than the beta of a portfolio of past winners. Further Jegadeesh and Titman (2001) and Lee and Swaminathan (2000) report that Fama-French three factor model does not explain the momentum portfolio returns. Rouwenhorst (1998) reports that momentum strategy cannot be accounted for by a simple adjustment for beta risk when the betas of the winner and loser portfolios are very similar.

Past studies on momentum effect reveal that momentum effect prevails in most of the developed markets as an anomaly. However, there is a lack of evidence for developing markets including the CSE.

3. Data and Methodology

Data used in the study have been taken from the CSE data library. The sample period covers 16 years from January 1992 to December 2007. The sample of the study includes all the voting stocks in the Main Board and the Second Board of the CSE. In accordance with Bildik and Gulay (2002), stocks which have less than 12 month data are excluded from the sample. therefore, the total sample includes 256 companies. This sample also includes stocks of delisted companies.

Using individual stock returns, monthly average percentage returns were computed. Percentage monthly returns were adjusted for dividends, right issues and bonus issues on the basis of the reinvestment assumption.

The stocks selected for the strategies implemented in this study are based on their returns over the past 3, 6, 9 and 12 months. The selected stocks were held for 3, 6, 9 and 12 months. This gives a total of 16 standard strategies. Computations are performed in two ways. Firstly, without imposing a time lag between formation period and the holding period and secondly by imposing a time lag of one month between the end of the formation period and the beginning of the holding period in order to avoid possible micro structure biases, thin trading problem and bid-ask spread.

In order to increase the power of statistical tests, the strategies examined include portfolios with overlapping holding periods. Therefore, in any given month t, a series of portfolios were held. These were selected in the current month as well as in the previous K-1 months, where K is the holding period. For example, the monthly return for a three-month holding period is based on an equally-weighted average of portfolio returns from this month's strategy, last month's strategy, and the strategy from two months ago.

This paper follows the same methodology used by Jegadeesh and Titman (1993). At the end of each month, from January 1992 to December 2007, all eligible stocks in the sample were ranked based on their past J month returns, for example, month -5 to month 0, if J is defined as six. From here the stocks were grouped into five equally weighted portfolios based on these ranks. Portfolio P1 represents the stocks with the highest ranking period returns and Portfolio P5 represents the stocks with the lowest ranking period returns. The top quintile portfolio is called the "winners" quintile and the bottom quintile is called the "losers" quintile. In each month t, the strategy buys the winner portfolio and holds this position for K months. Each port-folio is held for K months following the month in which it was ranked. For example, a portfolio is held from month 1 to month 3, if K is defined as three (K3). Hence, under this strategy, the weights of 1/K of the stocks in the entire portfolio in any given month are revised and carried over to the rest of the holding period from the previous month. The profits of the above strategies were calculated for a series of momentum portfolios (P1-P5) that were rebalanced monthly to maintain equal weights.

Risk adjustment

Returns of each of the momentum portfolios were adjusted for systematic risk based on the approach followed by Basu (1977) using the following equation:

[R.sub.pt] - [R.sub.ft] = [[alpha].sub.p]+[[beta].sub.p]([R.sub.mt] - [R.sub.ft])+[e.sub.pt]

Where,

[R.sub.pt] = continuously compounded return on momentum portfolio p in month,

[R.sub.ft] = monthly risk free rate at time t and this is represented by the interest rate on 3- month Treasury bills.

[R.sub.mt] = continuously compounded return on "market portfolio" in month t. The return is based on the All Share Price Index (ASPI).

[[alpha].sub.p] = the intercept of the regression to measure excess returns (Jensen's alpha) of portfolio P.

[[beta].sub.p] = the beta of portfolio P that is defined by the CAPM.

Hypotheses

1. If the pattern of the past period stock returns continues in the same direction over the next period, then a momentum portfolio can be formed by deducting returns of the loser portfolio (low return stocks) from returns of winner portfolio (high return stocks) in the holding period. Accordingly, the null hypothesis ([H.sub.0]) and the alternative hypothesis ([H.sub.1]) are developed as follows:

[H.sub.0]:E([R.sub.W,t+j]-[R.sub.L,t+j]) = 0

[H.sub.1]:E([R.sub.W,t+j]-[R.sub.L,t+j]) > 0

Where, [R.sub.W,t+J] = Winners' returns in the next period (holding period)

[R.sub.L,t+J] = Losers' returns in the next period (holding period) t+J = Holding period (months)

J = Number of months

The above null hypothesis explains that winners and losers have the same expected returns in the holding period while the alternative hypothesis explains that expected returns of winners are higher than those of losers in the holding period.

2. If the CAPM explains the momentum portfolio returns[[alpha].sub.p] of the equation 1 should equal to zero. Therefore, the null ([H.sub.0]) and alternative ([H.sub.1]) hypotheses are as follows;

[H.sub.0]:[[alpha].sub.p] = 0

[H.sub.1]:[[alpha].sub.p] [not equal to] 0

The null hypothesis explains that CAPM explains the returns of portfolios while the alternative hypothesis states that after adjusting for CAPM risk there is an excess return.

4. Empirical Results

Table 1 presents the results for all the portfolios formed. Each month's Stocks of are ranked and grouped into five portfolios on the basis of their returns over the previous three, six, nine and twelve months. The means for all the quintile portfolios with extreme winners (P1) and extreme losers (P5) and mean returns for winner minus mean returns for loser momentum portfolios (P1- P5) are also reported.
Table 1 Mean returns of momentum portfolios without a time lag between
portfolio formation and holding periods This table presents average
monthly and annual returns in percentages for portfolios which are
formed based on J-month lagged returns and held for K months. The
values of K and J for different strategies are indicated in the first
column. In each month t stocks are ranked in descending order on the
basis of J months lagged returns and five equally-weighted portfolios
are formed. An equally weighted portfolio of stocks in the highest
return portfolio is named as the winner (P1) portfolio and an equally
weighted portfolio of stocks in the lowest return portfolio is the
loser (P5) portfolio. The return of the winner portfolio minus that of
the loser portfolio shows the momentum effect. All figures are
percentages. The sample includes all the stocks traded on the Main
Board and on the Second Board of the CSE excluding those having
absolute returns greater than 50 percent. The sample period for the
study is from January 1992 to December 2007. The t-statistics are
reported below means.

                    P1            P2                          P4

J=3   mean             0.44         0.720          0.52          0.70
K=3   t-value          1.41  [2.43.sup.a]          1.77          1.26
J=6   mean             0.88          0.73          0.70          0.56
K=6   t-value  [4.17.sup.a]  [3.39.sup.a]  [3.18.sup.a]  [2.53.sup.a]
J=9   mean             1.07          0.95          0.76          0.61
K=9   t-value  [6.30.sup.a]  [5.31.sup.a]  [4.52.sup.a]  [3.33.sup.a]
J=12  mean             1.10          0.89          0.65          0.61
K=12  t-value  [7.15.sup.a]  [5.94.sup.a]  [4.40.sup.a]  [4.40.sup.a]

                P1-P5

J=3   0.63          -0.19
K=3   1.92          -0.95
J=6   0.38           0.50
K=6   1.64   [3.78.sup.a]
J=9   0.29           0.78
K=9   1.55   [7.72.sup.a]
J=12  0.30           0.80
K=12  1.88  [10.46.sup.a]

Note:(a) signifies statistical significance at the one per cent level.


According to Table 1, a 3-months/3-months strategy shows a negative momentum effect where past period losers outperform past period winners by 0.19 per month. Apart from that all the other strategies reported in the table reflect positive momentum effects where past period winners outperform past period losers. This is indicated by the statistically significant means.

The most successful momentum strategy is the portfolio with stocks based on their returns over the past 12 months which were held for next 12 months. This strategy yields 0.80 per cent over the loser portfolio per month.

Except for the 3-months/3-months strategy, holding period returns for all the other strategies are gradually declining from portfolio P1 to portfolio P5. Therefore, these findings reveal that investors can use profitable buy and hold strategies on the CSE based on the momentum patterns of portfolio returns.

Because the bid-ask bounce and the thin trading problem could intensify the continuation effect, Table 2 reports the average returns if the portfolio holding period is delayed relative to portfolio formation period by one month. For the shorter ranking and holding intervals, delaying the portfolio formation increases the return difference between the winners ( P1) and losers (P5). These findings support with the findings of Rouwenhorst (1998) and Jegadeesh and Titman (1993). According to the table all the strategies show positive momentum effects and except 3-months/3-months strategy all the means are statistically significant at the one per cent level. When there is a time lag between the formation period and the holding period, the most successful momentum strategy is to select stocks based on their returns over the past 9 months and then hold the portfolio for 9 months. This strategy yields an excess return of 0.87 per cent per month.
Table 2: Mean returns of Momentum portfolios imposing a time lag
between portfolio formation and holding periods This table presents
average monthly and annual returns in percentages for Portfolios which
are formed based on J-month lagged returns and held for K months. The
values of K and J for different strategies are indicated in the first
column. Momentum portfolios are formed after imposing a lag of one
month between formation period and holding period. In each month t
stocks are ranked in descending order on the basis of J months lagged
returns and five equally-weighted portfolios are formed. An
equally-weighted portfolio of stocks in the highest return portfolio
is named as the winner (Pi) portfolio and an equally weighted
portfolio of stocks in the lowest return portfolio is the loser (P5)
portfolio. Winner minus loser portfolio is the momentum portfolio. All
the figures are in percentages. The sample includes all the stocks in
the Main Board and the Second Board of the CSE excluding those having
absolute returns greater than 50 per cent. The sample period for the
study is from January 1992 to December 2007. The t-statistics are
shown below the mean return for each portfolio.

                    P1                          P3            P4

J=3   mean             0.67          0.84          0.64          0.45

K=3   t-value  [2.14.sup.a]  [2.75.sup.a]  [2.11.sup.a]          1.49

J=6   mean             0.97          0.88          0.74          0.50

K=6   t-value  [4.74.sup.a]  [3.99.sup.a]  [3.34.sup.a]  [2.22.sup.a]

J=9   mean             1.13          1.00          0.74          0.58

K=9   t-value  [6.47.sup.a]  [5.55.sup.a]  [4.32.sup.a]  [3.22.sup.a]

J=12  mean             1.10          0.89          0.65          0.62

K=12  t-value  [6.92.sup.a]  [5.65.sup.a]  [4.18.sup.a]  [3.85.sup.a]

                P1-P5

J=3   0.41          0.25

K=3   1.27          1.42

J=6   0.26          0.71

K=6   1.08  [5.36.sup.a]

J=9   0.26          0.87

K=9   1.40  [8.61.sup.a]

J=12  0.30          0.80

K=12  1.80  [9.85.sup.a]

Note:(a) signifies statistical significance at the one per cent level.


The fundamental assumption underlying the results in Tables 1 and 2 is that the momentum effects over the 16 year period are stationary. To determine the validity of this assumption the 16 year period is divided into two non-overlapping subperiods from January 1992-December 1999 and January 2000-December 2007.

Table 3 documents the mean returns of the 3, 6, 9, and 12 month strategies for two sub-periods, the first period from January 1992 to December 1999 and the second period from January 2000 to December 2007. The table indicates that all the four strategies result in positive momentum profits for both sub-periods. Apart from the 3-months/3-months strategy all the other strategies generate statistically significant momentum effects. However, in the first sub-period momentum effects are mainly due to the fact that losers generate large negative returns. In contrast, in the second sub-period, momentum effects are mainly due to the fact that winners and losers both generate high positive returns. This contradictory return behavior over the two sub sample periods may be due to stagnated market activities in the first period and increasing market indices or trading efficiency in the second period. However these momentum profits in both periods prove the pervasive character of the momentum profits of the CSE.
Table 3: Sub-period mean returns of momentum portfolios
This Table presents average monthly returns in percentages for
portfolios that are formed based on J-month lagged returns and held
for K months. The values of K and J for different strategies are
indicated in the first column and second row of each column,
respectively. Holding period portfolios are formed after imposing a
time lag of one month between formation period and holding period. In
each month t stocks are ranked in descending order on the basis of J
months lagged returns and five equally-weighted portfolios are formed.
An equally-weighted portfolio of stocks that has the highest return is
named as the winner (P1) portfolio and an equally-weighted portfolio
of stocks that has the lowest return is the loser (P5) portfolio.
Table shows the momentum effects on a sub-sample basis. Total sample
period (sixteen years) is divided into two periods from January
1992-December 1999 and January 2000 to December 2007. The sample
includes all the stocks on the Main Board and on the Second Board of
the CSE excluding those having absolute returns greater than 50 per
cent. The t-statistics are reported below the mean values.
Sub-sample
period:
1992 to
1999

                           P1               P2              P3

J=3         mean             0.001          -0.13         -0.38
K=3         t-value           0.02          -0.28         -0.87
J=6         mean              0.33           0.00         -0.18
K=6         t-value           1.05           0.00         -0.52
J=9         mean              0.48           0.15         -0.06
K=9         t-value           1.74           0.55         -0.22
J=12        mean              0.46           0.02         -0.28
K=12        t-value           1.73           0.08         -1.26

Sub-sample
period:
2000 to
2007

                           P1               P2              P3

J=3         mean              1.31           1.79          1.64
K=3         t-value   [3.38.sup.a]   [4.59.sup.a]  [4.04.sup.a]
J=6         mean              1.58           1.71          1.62
K=6         t-value   [6.37.sup.a]   [7.10.sup.a]  [6.82.sup.a]
J=9         mean              1.77           1.83          1.53
K=9         t-value   [9.02.sup.a]   [9.38.sup.a]  [7.58.sup.a]
J=12        mean              1.73           1.74          1.56
K=12        t-value  [11.16.sup.a]  [12.38.sup.a]  [9.53.sup.a]

Sub-sample
period:
1992 to
1999

                 P4             P5             P1-P5

J=3                -0.57          -0.26          0.27
K=3                -1.37          -0.56          1.01
J=6                -0.43          -0.70          1.03
K=6                -1.28          -1.95          5.06
J=9                -0.30          -0.53          1.01
K=9                -1.15          -1.79          7.35
J=12               -0.30          -0.57          1.02
K=12               -1.34  [-2.26.sup.a]  [9.99.sup.a]

Sub-sample
period:
2000 to
2007

                      P4             P5         P1-P5

J=3                 1.45           1.08          0.24
K=3         [3.45.sup.a]   [2.40.sup.a]          0.99
J=6                 1.39           1.45          0.43
K=6         [5.10.sup.a]   [4.10.sup.a]  [2.56.sup.a]
J=9                 1.45           1.03          0.74
K=9         [6.73.sup.a]   [5.05.sup.a]  [4.99.sup.a]
J=12                1.51           1.15          0.58
K=12        [8.19.sup.a]   [6.13.sup.a]  [4.78.sup.a]

Note: (a) signifies statistical significance at the one per cent level.


As per the results reported in Table 4, median returns and annual mean returns clearly show the cross-sectional variation of portfolio returns during the holding period. Median return of the winner portfolio is 0.95 per cent per month while the loser portfolio records a median return of 0.20 per month. Annual mean returns across portfolios show the same pattern.
Table 4: Risk-adjusted portfolio returns
This Table presents median, inter-quartile range, annual returns,
systematic risk ([[beta].sub.p]) and Jensen's a for portfolios which
are formed based on (5-month lagged returns and held for 6 months.
Holding period portfolios are formed after imposing a month time lag
between formation period and holding period. In each month t stocks
are ranked in descending order on the basis of 6 month lagged returns
and 5 equally weighted portfolios are formed. An equally-weighted
portfolio of stocks in the highest return portfolio is named as the
winner (Pi) portfolio and an equally weighted portfolio of stocks in
the lowest return portfolio is the loser (P5) portfolio. The sample
includes all the stocks in the main Board and the second Board of the
CSE excluding those having absolute returns greater than 50%.The
sample period for the study is from January 1992 to December 2007.
The t-statistics for systematic risk ([[beta].sub.p]) and Jessen's a
are reported below their respective values.

                            P1               P2

Median returns                   0.95             0.69

Inter-quartile range            -7.66            -8.85

Average annual rate             11.64            10.56
of returns

Systematic risk                  0.69             0.76
([[beta].sub.p])
                      ([23.23.sup.a])  ([27.78.sup.a])

Jensen's [alpha]                -0.06            -0.15
                              (-0.57)          (-1.53)

                                             P4

Median returns                   0.83             0.25

Inter-quartile range            -8.49           -10.73

Average annual rate              8.88              0.6
of returns

Systematic risk                  0.78             0.80
([[beta].sub.p])
                      ([29.22.sup.a])  ([29.08.sup.a])

Jensen's [alpha]                -0.29            -0.52
                      ([-3.00.sup.a])  ([-5.32.sup.a])

                            P5

Median returns                  -0.20

Inter-quartile range           -11.24

Average annual rate              3.12
of returns

Systematic risk                  0.81
([[beta].sub.p])
                      ([26.92.sup.a])

Jensen's [alpha]                -0.77
                      ([-7.08.sup.a])

Note:(a) signifies statistical significance at the one per cent level.


Table 4 also reveals that inter-quartile range of returns has an increasing trend from the winner portfolio to the loser portfolio. It implies that time series variation of portfolio returns gradually increases form P1 toP5 which is further evident by the similar pattern of systematic risk factor ([beta]) from portfolio P1 to P5. The Jensen's [alpha] explains the reason for the momentum effect on the CSE. Further, the CAPM clearly explains the variation of the winner portfolio measured by [alpha] is not statistically significantly different from zero (P1=-0.6). However, the loser portfolio (P5) records significantly negative excess returns of -0.77 per cent per month after adjusting for systematic risk. The difference between excess returns of P1 and P5 is 0.71 per cent and this value is exactly equal to the return of winner portfolio minus that of the loser portfolio reported in Table 2.

Therefore, it is clear that the momentum effect on the CSE is due to under performance with loser portfolio rather than as a result of positive excess returns with the winner portfolios. These results reveal that the momentum effect prevails on the CSE when systematic risk is adjusted for.

4. Conclusion

We examined the existence of a momentum effect on the CSE during the period from 1992 to 2007. The sample of the study includes all the voting stocks on the Main Board and the Diri Savi Board of the CSE traded during this periods. The study adds some important findings to the existing literature as there is little evidence on this concept for developing countries, particularly for Sri Lanka.

Financial academics and practitioners have recognized that the average stock returns are related to past performance and the cross-section of stock returns is predictable on the basis of past returns. A number of past researchers have reported that past winners outperform past losers in the subsequent period not only in the US market but also in some of the other markets. However, there is still insufficient evidence to support the above finding in developing markets.

Our analysis reveals the existence of intermediate period positive momentum effects on the CSE. Momentum effects are highly significant and comparatively massive with the imposition of a time lag of one month between portfolio formation periods and holding periods. Sub-sample results indicate that the momentum effects are visible during both sub-sample periods on the CSE.

The risk-adjusted portfolio returns indicate that the momentum effect on the CSE is due to the under-performance of losers rather than as a result of positive excess returns of winner portfolios. The momentum anomaly existing on the CSE could be used as a risk factor to absorb the cross-sectional variability of returns in the Sri Lankan market as in other stock markets. As an emerging stock market in South Asia with significant foreign investor interest, the results reported in this paper have significant implications for the portfolio decisions of foreign investors in addition to their importance to domestic investors.

Limitations of this paper are that we have not investigated how volatility and company size affect the momentum profits. These issues are left for future research. Further, future researchers could investigate the sensitivity of the results reported in this paper to different trading frequencies such as daily and weekly frequencies and different portfolio weighting schemes.

NOTES

(1.) See Chopra and Lakonishok (1992); Ritter, Ball and Khothari (1989); Lo and Mackinlay (1990); Gunaratne and Yonezawa (1997).

(2.) This study does not cover South Asian Countries.

REFERENCES

Ball, R. and Kothari, S.P. (1989), "Nonstationary Expected Returns: Implications for Tests of Market Efficiency and Serial Correlation in Returns," Journal of Financial Economics 25: 51-74.

Banz, R. (1981), "Relationship between Return and Market Value of Common Stocks," Journal of Financial Economics 9: 3-18.

Barberis, N., Shleifer, A. and Vishny, R. (1998), "A Model of Investor Sentiment," Journal of Financial Economics 49: 307-343.

Basu, S. (1977), "Investment Performance of Common Stocks in Relation to Their Price Earnings Ratios: A Test of the Efficient Market Hypothesis," Journal of Finance 32: 663-682.

Bildik, R., and Gulay, G. (2002), "Profitability of Contrarian Vs Momentum Strategies Evidence from The Istanbul Stock Exchange," www.efmaefm_org/bildikrecp/bildikrecep.pdf

Black, F., Jensen, M. and Scholes, M. (1972), "The Capital Asset Pricing Model: Some Empirical Tests," in Michael Jensen (ed.), Studies in the Theory of Capital Markets. New York: Praeger.

Chan, K.C., Jegadeesh, N. and Lakonishok, J. (1996), "Momentum Strategies," Journal of Finance 51: 1681-1713.

Chang, R.P., Mcleavey, D.W. and Ruhee, S. G. (1995), "Short-term Abnormal Returns of the Contrarian Strategy in the Japanese Stock Market," Journal of Business Finance and Accounting 22: 1035-1048.

Chodia T. and Shivakumar, L. (2002), "Momentum, Business Cycle, and Time Varying Expected Returns," Journal of Finance 57: 985-1018.

Chopra, N., Lakonishok, J. and Ritter, J.R. (1992), "Measuring Abnormal Performance: Do Stocks Overreact?" Journal of Financial Economics 31: 235-268.

Chui, A., Titman, S. and Wei, K.C.J. (2000), "Momentum, Ownership Structure, and financial Crisis: An Analysis of Asian Stock Markets," Working Paper, University of Texas at Austin.

Conrad, J. and Kaul, G. (1998), "An Anatomy of Trading Strategies," Review of Financial Studies 11: 489-520.

Daniel, K., Hirshleifer, D. and Subramanyam, A. (1998), "Investor Psychology and Security Market Under-and Overreaction," Journal of Finance 53: 1839-1886.

De Bondt, W. and Thaler, R. (1985), "Does the Stock Market Overreact?" Journal of Finance 40: 793-805.

DeBondt, W.F.M., and Thaler, R.H. (1987), "Further Evidence on Investor Overreaction and Stock market Seasonality," Journal of Finance 42: 557-581.

Fama, E.F. (1991), "Efficient Capital Markets: II," Journal of Finance 46: 1575-1617.

Hong, H. and Stein, J. C. (1999), "A Unified Theory of under Reaction, Momentum Trading and Overreaction in Asset Markets," Journal of Finance 54: 2143-2184.

Guneratne, P. S. M., and Yonezawa, Y. (1997), "Return Reversals in the Tokyo Stock Exchange: A Test of Stock Market Overreaction," Japan and the World Economy 9: 363-384.

Jegadeesh, N. (1990), "Evidence of Predictable Behavior of Security Returns," Journal of Finance 45: 881-898.

Jegadeesh, N. and Titman, S. (1993), "Returns to Buying Winners and Selling Losers: Implications for Stock Market Efficiency," Journal of Finance 48: 65-91.

Jegadeesh, N. and Titman, S. (2001), Profitability of Momentum Strategies: An Evaluation of Alternative Explanations, Journal of Finance 56: 699-720.

Jegadeesh, N. and Titman, S. (2002), "Cross-Sectional and Time Series Determinants of Momentum Returns," Review of Financial Studies 15: 143-157.

Lewellen, J. (2002), "Momentum and Autocorrelation in Stock Returns," Review of Financial Studies 15: 533-563.

Lee, C.M.C. and Swaminathan, B. (2001), "Price Momentum and Trading Volume," Journal of Finance 55: 2017-2069.

Lehmann B. (1990), "Fads, Martingales and Market Efficiency," Quarterly Journal of Economics 105: 1-28.

Lintner, J. (1965), "Security Prices, Risk, and Maximal Gains from Diversification," Journal of Finance 20: 587-615.

Lo, A. and MacKinlay, A.C. (1990), "When Are Contrarian Profits Due to Stock Market Overreaction?" Review of Financial Studies 3(2): 175-205.

Mossin, J. (1966), "Equilibrium in a Capital Asset Market," Econometrica 34: 768-783.

Nijman T., Swinkels, L. and Verbeet, M. (2004), "Do Countries or Industries Explain Momentum in Europe?" Journal of Finance 11: 461-481.

Rouwenhorst, G. (1998), "International Momentum Strategies," Journal of Finance 53: 267-284.

Rosita P., Chang, McLeavey, D.W. and Rhee, S.G. (1995), "Short-Term Abnormal Returns of the Contrarian Strategy in the Japanese Stock Market," Journal of Business Finance & Accounting 22: 1035-1047.

Sharpe, W.F. (1964), "Capital Asset Prices: A Theory of Market Equilibrium under Considerations of Risk," Journal of Finance 19: 425-443.

Shen, Q., Szakmary, A.C. and Sharma, S.C. (2005), "Momentum and Contrarian Strategies in International Stock Markets: Further Evidence," Working Paper, http://www.sciencedirect.com/science/article/B6VGV-4FX23HP

Wang, K., Jiang, C., Huang, Y. (2009), "Market States and the Profitability of Momentum Strategies: Evidence from the Taiwan Stock Exchange," International Journal of Business and Finance Research 3: 89-102.

Zarowin, P. (1990), "Size Seasonality and Stock Market Overreaction," Journal of Financial and Quantitative Analysis 25: 113-125.

CHANDRAPALA PATHIRAWASAM pathi3@yahoo.com Faculty of Management and Economics Tomas Bata University in Zlin

YATIWELLA KORALALAGE WEERAKOON BANDA weerakonyatiwella@yahoo.com Department of Finance University of Sri Jayewardenepura, Nugegoda

GUNERATNE B. WICKREMASINGHE Guneratne.wickremasinghe@vu.edu.au School of Accounting and Finance & Centre for Strategic Economic Studies Victoria University

[c] Chandrapala Pathirawasam, Yatiwella Koralalage Weerakoon Banda, Guneratne B. Wickremasinghe
COPYRIGHT 2011 Addleton Academic Publishers
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2011 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Pathirawasam, Chandrapala; Banda, Yatiwella Koralalage Weerakoon; Wickremasinghe, Guneratne B.
Publication:Economics, Management, and Financial Markets
Date:Dec 1, 2011
Words:5864
Previous Article:An analysis on political risks and the flow of foreign direct investment in developing and industrialized economies.
Next Article:An international perspective on the methods and practices of student loans: its implications for India.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters