A mutual fund is a professionally managed firm of collective investments that collects money from many investors and puts it in stocks, bonds, short-term money market instruments, and/or other securities (Investopedia). As it is professionally managed this segment is very crucial to bring rationality in the market.
In Bangladesh, the mutual funds were first introduced by the state-owned investment agency Investment Corporation of Bangladesh (ICB) in 1980. It launched some 8 close-end and one unit fund till 2002, when it had to create 3 subsidiaries under an Asian Development Bank prescription, including the asset management company, which was entrusted with the responsibility to launch mutual funds under the SEC rules (AIMS 2007). This company has launched 7 close-end and 2 open-end mutual funds since 2009. Meanwhile, another state-owned lending agency, Bangladesh Shilpa Rin Sangstha (BSRS), now merged in to Bangladesh Development Bank, launched its solitary mutual fund in 1997, which is run under its own statute.
The first ever private asset management company AIMS of Bangladesh comes into play at 1999. They issued their first fund, AIMS First Guaranteed Mutual Fund, a closed end balanced fund in March 2000. Currently few other private fund managers doing this business namely RACE, LR Global, Brac-EPL etc. All together 24 close ended funds are traded in the Dhaka Stock Exchange (DSE) with quite few waiting in the queue for starting their operation.
In the market of active portfolio management the central concentration of related parties (manager, investor and researcher) is the performance of the active management versus passive management. This paper is trying to evaluate the performance of closed end mutual funds traded in the Dhaka Stock Exchange (DSE) in compare to benchmark index. The article is segmented in few parts. At first we focus on the earlier studies regarding the performance evaluation of mutual funds in literature review section. The following parts present the research questions and basic concepts of performance analysis. In part five, six, seven and eight presents the methodology, limitation, analysis of results and summary and conclusion respectively.
2. Literature Review:
In the area of Bangladesh's mutual fund performance analysis there are very few research conducted. As the market growing rapidly and related parties are also approaching promptly, it is very good time to evaluate the market activities. We conduct this research basically based on methods used and practiced in overseas countries. A wide range of technical and quantitative tools have been produced to assess and contrast performance of managed portfolios. The pioneering work on the performance evaluation of mutual funds was done by Sharpe (1966) who has developed a composite measure that considers return and risk. He evaluated the performance of 34 open-ended mutual funds during the period 1944-63 by the measures developed by him. He concluded that the average mutual fund performance was distinctly inferior to an investment in the DJIA (Dow Jones Industrial Average). It was also revealed in his study that good performance was associated with low expense ratio and only low relationship was discovered between fund size and performance. (Christensen, 2005)
A study performed by Treynor & Mazuy (1966) found no statistical evidence that investment manager of any 57 funds were not able to guess the market movements in advance. This study suggests that an investor in mutual funds was totally dependent on fluctuations in the general market. The study revealed that the improvement in rate of return was due to the fund managers' ability to identify under priced shares in the market. Jensen M C (1968) evaluated the ability of the fund managers in selecting the undervalued securities. He concludes that for the sample 115 mutual funds, the fund managers were not able to forecast security prices well enough to recover research expenses and fees.
E. Fama (1970) developed a methodology for evaluating investment performance of managed portfolios. He suggested that the overall performance of managed portfolios could be broken down into several components. He argued that the observed return of a fund could be due to ability of fund managers to pick up the best securities at a given level of risk (their selectivity ability). Some portion of this return could also arise due to the prediction of general market price movements (their market timing ability). Fama suggested that return on a portfolio could be subdivided into two parts. The return for security selection and return for bearing risk. Various finer subdivisions of both selectivity and risk were also suggested. The model developed by him combined concepts from modern theories of portfolio selection and capital market equilibrium with those of traditional concepts of what constitute good portfolio management.
Henriksson and Merton (1981) provide a theoretical model and Henriksson (1984) provide an empirical test of the timing ability of fund managers. Motivated by the pioneering work of Henriksson and Merton (1981), Weigel (1991) found managers had reliable, although not perfect, market timing skills. Several investigations, as typified by Kon (1983) and Chang and lewellen (1984), concluded that mutual fund managers generally possess negative market timing skills. Coggin, Fabozzi and Rahman (1993) carried this analysis further by looking at both the timing and selectivity skills of a group of US equity pension fund managers. Using a regression based model with monthly return data for an eight-year period ending in December 1990, they demonstrated that their sample of managers possessed positive, but small, selection skills and negative timing skills.
Most of the performance evaluation methods depend on an asset pricing model. These measures suffer from Roll (1978)'s critique. Papers that provide measures that do not depend on an asset pricing model include Cornell (1979) and Grinblatt and Titman (1993). Besides Chang and Lewellen (1985), and Connor and Korajczyk (1986) use APT frame works to measure performance. These authors also get mixed findings regarding the performance of mutual fund.
Previously numerous studies conducted regarding mutual funds around the world. Our literature review revealed that there is wide open area of research exists in mutual fund industry of Bangladesh. The lacking of research in this sector motivates us to conduct this fundamental study with the objectives given in the next part.
3. Objective:
The present paper attempts to answer two questions relating to mutual fund performance;
Whether the Mutual Fund are earning higher returns than the benchmark index returns in terms of risk.
Whether the Fund Managers' are offering the advantages of Diversification, Market timing and Selectivity of Securities to their investors.
This paper attempts to answer the questions raised, by initially describing some basic concepts and later by employing a methodology which was used by Jenson (1968), Treynor (1965), Sharpe (1966), Fama (1972), Treynor and Mauzy (1966) and finally drawing appropriate conclusions regarding overall mutual fund industry.
Thirteen mutual funds selected for the purpose of this study. We select all mutual funds traded in the DSE those have maturity over 3 years. The study period is 36 months (April, 2007 to March 2010). The data source is monthly Net Asset values (NAVs) published in the DSE news server. DSE 20 is assumed as market index or the benchmark.
4. Basic concept:
Portfolio Returns. Fundamentally, return on a portfolio or a fund is :
Where,
is the return on portfolio
is the Net Asset Value of the fund
't' is the time period; here one month interval.
Monthly returns based on the NAVs of the thirteen funds for a particular period are taken and simple averages of such returns (ARp) are calculated. When the fund issues dividend, dividends are adjusted in T+4 month, where T is the declaration month. We assume the dividend will affect NAV in the 4th month of declaration. We adjust dividend in the following way;
Where,
is the amount dividend declared
Market Return; similarly, returns on the market index (rm) are taken to arrive at the average market return (ARm)
Where,
is the return on market index
Further it is assumed that, the monthly risk free return (rf) is one percent. The reason for this is Post Office quarterly income deposits offer an annual return of 12.5 per cent. Thus, the average monthly risk free return (ARf) should little less than one per cent which is rounded up to one percent.
Risk is the variability of returns. Total risk is measured with the help of standard deviation of returns for both the portfolio and the market
Where,
is total risk of the Portfolio
Excess return of Portfolio over risk free return in a particular period
Average excess return of Portfolio over risk free return
N Number of observations
Where,
is total risk of the Market Index
Excess return of Market Index over risk free return in a particular period
Average excess return of over Market Index risk free return
N Number of observations
There are two components of total risk; systematic risk and unique risk. Systematic risk is measured as follows:
Where
is Beta, systematic risk of the portfolio
t is the time period 1, 2 ………. N
N is Total number of observations
Systematic risk includes broad economic factors which influence all securities traded in the market. Thus, systematic risk of the market is always one; systematic risk of risk free investment is zero, and portfolio or fund systematic risk can be greater than or less than one.
Unique risk is the risk of the portfolio in particular. It is measured with the help of Standard Deviation of Error term (SDEt). Active portfolio managers are always aiming to reduce or remove, if possible, the unique risk component. (Jayadev, 1996) Unique portfolio risk is;
Where,
is Error terms of the portfolio for period 't'
is Average of error terms.
is number of observation
5. The Methodology:
We will use the following methods to answer the questions raised in the objective section little earlier.
Methods of risk-adjusted performance evaluation using mean-variance criteria will be use to answer the question one. Basically this evaluation technique is based on Capital Asset Pricing Model (CAPM). Jack Treynor (1966), William Sharpe (1966) and Michael Jensen (1968) recognized immediately the implications of the CAPM for rating the performance of managers. We use these three classic model to evaluate the risk adjusted performance of the mutual funds.
Jensen Measure: Jensen's measure is the average return on the portfolio over and above that predicted by the CAPM i.e. equilibrium returns, given the portfolio's beta and the average market return. Jensen measures the portfolios alpha value.
Where,
is the Jensen Alpha
is the average portfolio return
= Equilibrium Average Return
If the alpha is positive, the portfolio has performed better and if alpha is negative it has not shown performance up to the benchmark i.e. the market index.
Reward to volatility ratio: This ratio introduced by Treynor (1965). Here, additional returns of the portfolio over the risk free return is expressed in relation to portfolio's systematic risk;
The numerator of this ratio is the risk premium and the denominator is a measure of risk, the total expression indicates the portfolio's risk premium return per unit of risk. All risk averse investor would prefer to maximize this value. This model implicitly assumes a completely diversified portfolio, which means that systematic risk is the relevant risk measure. Reward to volatility of the market is:
Here, an additional return of market over risk free return is the benchmark. Greater value of the portfolio over the market indicates a superior performance of the fund.
The analysis on the basis of above two measures may lead to the same conclusion. This is so because both the measures are based on only systematic risk and exclude unique risk of the portfolio. Hence it is necessary to evaluate the performance of the fund in terms of its total risk. The following measure is used for the purpose.
Reward to variability: It was developed by William F. Sharpe (1966). This measure divides average portfolio excess return over the sample period by the standard deviation of returns over that period. It measures the reward to total variability trade off.
The bench mark is additional return of market over risk free return related with market portfolio's total risk.
A fund which performed better according to first two measures namely Jensen and Treynor measures and not according to the third measures indicates amount of unsystematic risk is excessive and few diversification of risk in compare to reward. (Jayadev, 1996)
The second question rises to check the diversification power, Market timing ability and Selectivity of the active fund management. The following three techniques will check these three parameters.
Diversification: The fundamental motive of portfolio management is to remove the unsystematic risk of individual stock through diversification. As small investor does not have the capacity and capability to make well diversified portfolio, they primarily looking towards fund managers to get this benefit. Diversification can be measured with the help of coefficient of determination (). This can be obtained by regressing the portfolio's excess return () against the market excess returns (). A high value indicates greater diversification of fund and vice-versa.
Market timing: It means actively manage the portfolio's composition based on the market condition. A fund manager who would like to prefer market timing, structures the portfolio to have a relatively high beta during a market rise and relatively low beta during market decline (Treynor and Mazuy, 1966). To do this manager will adjust the portfolio β according to the return on the market portfolio as and substituting this relationship into excess return equation, we get:
Which gives us the quadratic Treynor and Mazuy equation. Compared to the standard security market line this equation includes a new term, which is the excess return of the market squared. If is positive and significantly different from zero, we identify selection skills, as in the market line model, and if is positive and significant, the mutual fund manager possesses timing ability. (Christensen, 2005)
Selectivity: selectivity is the skill of the fund manager to select undervalued securities (priced lower than their intrinsic value at a particular point of time) in order to earn higher returns. It can be known with the help of Fama's (1972) decomposition measure.
A positive high value indicates that the fund has achieved superior returns and investor's are benefited out of the selectivity exercised by the Fund Manager.
6. Limitations:
There are some limitation exist in this article. Those are given below:
A very long time interval is needed in order to be able to obtain a measure of risk adjusted performance that can distinguish skill from luck on the part of the investment manager. Number of observations used in this article is 36 months return due to lack of data availability. Prior to this period there was no obligation to declare NAV by the fund manager.
The study period is three years starting from April, 2007 to evaluate the performance of the selected mutual Fund's but not from their inception.
DSE 20 is taken as the benchmark index for the market which is consist 20 stocks from 443 listed securities.
The models used in this article do not consider expenses and fees involving active management. This expenses and fees will lower the return of the funds but extent of reduction is not gauged.
7. Results and Analysis:
Mutual Fund's performance in contrast to benchmark index:
Table 2: exhibits risk and return of the selected portfolios and benchmark index. All of the mutual funds outperform the market index strongly. In terms of total risk (σ) there are mix outcome exist. Some funds possesses higher than market risk, some are lower than market risk. Among all mutual funds 5th ICB mutual fund depicts highest total risk (σ) as well as systematic risk (β). But in case of return 2nd ICB mutual fund illustrates the highest average return. On the other hand ICB Islamic mutual fund shows lowest total risk (σ) and lowest average return.
Table 3: exhibits risk adjusted performance measures. These measures are calculated on the basis of rules discussed in the methodology section. All of the selected mutual funds are holding positive alpha (α), indicating superior performance. 2nd ICB mutual fund has the highest alpha value among all mutual funds. Fund reward to volatility ratio for 2nd ICB is also highest due to underlying computation principle. Whereas ICB Islamic mutual fund presents lowest alpha and reward to volatility ratio. Moreover every mutual fund depicts higher than market in case of reward to volatility as well as reward to variability ratio.
Reward to variability measures the amount of excess return for per unit of total risk i.e. standard deviation of the portfolio. It is based on ex post Capital Market line (CML). 3rd ICB mutual fund has the highest reward to variability ratio though 2nd ICB mutual fund performs better according to earlier two measures. This is because out of total risk 3rd ICB have more systematic risk in compare to 2nd ICB mutual fund. In case of ICB Islamic mutual fund the result is consistent with the previous two measures indicates this fund's risk adjusted performance is poor according to all measures.
Diversification, market timing and selectivity:
Table 4: Risk and Diversification. The lowest R2 of 2nd ICB mutual fund indicates that this fund have the smallest amount of diversification i.e. low amount of systematic risk in total risk. As a result this fund has the highest amount of unsystematic risk. AIMS 1st mutual fund has the highest value of R2 indicates highest level of diversification. That means 62% of AIMS 1st mutual fund return explain by the market return. This fund will be performing superior in the good times of the market, vice versa. Mutual fund of DSE collectively does not offering high level of diversification to the investors. Nine out of thirteen mutual funds posses less than .50 R2 value with statistical significance.
Table 5: Treynor and Mauzy quadratic equation. From this table market timing ability of the fund manager can discover. Out of thirteen funds seven funds have positive Bp1, meaning that those funds consisted of high beta securities when the market return was high and low beta securities when the market return was low. All remaining 6 funds have negative Bp1, indicate the reverse scenario. But not a single co-efficient found as statistically significant at 1%, 5%, 10% level of significance. Moreover Treynor and Mauzy quadratic equation facilitate to observe the selectivity with statistical significance. In this table αi row provides positive value for every mutual fund indicate each and every fund posses the selectivity. Out of these funds three funds' selectivity is statistically significant at 1%, eight funds at 5% and ten funds at 10% level of significance. There is no significance found in selectivity for other three mutual funds.
Table 6: Components of Return. This table presents break up of portfolio returns with the help of Fama's decomposition measures. All of the selected funds have the positive values in the row five indicates that funds have earned superior returns due to selectivity by the fund managers. This result is similar with the Treynor and Mauzy (1966) method. Moreover most of the mutual fund evident that net selectivity is the single dominant component in total return. That means funds have offered professional expertise to the investors.
8. Summary and Conclusion:
From the result analysis it is clearly evident that overall performance of the mutual funds are superior to market return. Funds have higher than market average return, strong excess return over equilibrium return, high level of reward for bearing risk. Moreover most of the sample funds have the significant selectivity and some level of diversification. That means fund managers providing professional expertise in their portfolio management except timing skill.
There are differences exist in the level of performance among the funds of this industry. Though this issue is not the main focus of this paper still there are some facts which affected the performance measures. In our selection there are nine funds managed by state-owned investment agencies and other four managed by ICB subsidiaries and private Asset Management Company. It is commonly argued Securities and Exchange Commission's (SEC) mutual fund rules are creating uneven playing field between this two types of manager. Because state owned mutual funds have unlimited borrowing power, not bound by any quantitative restrictions on single company or industry, access to the insider information by be the member of the board of public limited company etc. These privileges in turn damage the managerial capacity of these agencies and give a misleading performance scenario of the funds. (AIMS 2007) This augment weaken the quantitative findings regarding the comparative performance of selected funds.
In Bangladesh capital market has failed to attract the entrepreneurs as the key source of capital. That is why Bangladesh capital market has very low market capital as percentage of GDP (7.5%) while neighboring country India has very high market capital to GDP (108.0%) ratio in 2007. For rapid industrial development capacity of the capital market should be stronger. This market can be flourished by attracting more institutional investor who has the capacity to accumulate fund and managerial skill. That's why level playing field among the managers should be ensured.
Despite regulatory drawbacks and limitations mentioned earlier overall mutual fund market doing well in terms of NAV. But in recent times high demand in market, puzzle of bonus issue, directive regarding close down of older mutual funds, volatile margin rule create a very rough time for this industry. Still market capitalization of mutual funds is 1.46 times over its NAV. This high market price will reduce the yield regardless of good return on the basis of NAV for the investors. Moreover current average P/E ratio of all traded mutual funds is 32.73 times indicating high price compare to earnings. Lacking of skill investor, weak institutional framework, asymmetrical regulation, small amount of academic research is moving this prudential sector to a vague way. All related parties should come forward to direct this sector in right way. performance analysis with larger sample size, persistence of the returns, expense and fee adjusted performance evaluation, validity of Capital Asset Pricing Model can be very good future research in the area of mutual funds of Bangladesh.
