See discussions, stats, and author profiles for this publication at: [Link]

net/publication/310457941

Can Mutual Funds Outguess the Market: Evidence from Bangladesh?

Article · November 2016

DOI: 10.12691/jfa-4-1-2

CITATIONS READS

5 2,309

1 author:

Md Bokhtiar Hasan
Islamic University
65 PUBLICATIONS 1,276 CITATIONS

SEE PROFILE

All content following this page was uploaded by Md Bokhtiar Hasan on 17 November 2016.

The user has requested enhancement of the downloaded file.

Journal of Finance and Accounting, 2016, Vol. 4, No. 1, 11-19
Available online at [Link]
©Science and Education Publishing
DOI:10.12691/jfa-4-1-2

Can Mutual Funds Outguess the Market: Evidence from

Bangladesh?
Md. Bokhtiar Hasan1,*, A. F. M. Mainul Ahsan2
1
Department of Finance and Banking, Islamic University, Kushtia, Bangladesh
2
Department of Accounting and Corporate Governance, Macquarie University, Sydney, Australia
*Corresponding author: bokhtiar_bank@[Link]

Abstract This study principally analyzes the fund managers’ ability to outguess the market in Bangladesh. We
perform the investigation on weekly data of 25 mutual funds for the period of May 16, 2010 to April 28, 2016. To
serve our objective, we tested both selection and market timing skills of the fund managers. We have used six
measures; average return, Sharpe ratio, Treynor ratio, Information ratio, Jensen’s alpha and M square; to confirm the
selection skill of fund managers and found no selection skill persistent to most of the fund managers (excluding
Aims 1st M.F, ICB AMCL 2nd NRB M.F. and 6th ICB M.F.). In addition, the negative values of alpha indicate that
fund managers become not only failed to add value to their portfolio, but also pool wrong assets which hurt the
return resulting negative profit. On the other hand, we have employed two popular methodologies; Treynor and
Mazuy [24] and Henriksson and Merton [10]; to test the market timing skill of fund managers and found no market
timing skill persistent to the fund managers. Thus, with a little exception, we can conclude that fund managers have
no ability to outguess the market in Bangladesh.
Keywords: mutual funds, NAV, market timing, selectivity, fund performance
Cite This Article: Md. Bokhtiar Hasan, and A. F. M. Mainul Ahsan, “Can Mutual Funds Outguess the
Market: Evidence from Bangladesh?” Journal of Finance and Accounting, vol. 4, no. 1 (2016): 11-19.
doi: 10.12691/jfa-4-1-2.

investment decisions appropriately while fund managers

do this to evaluate their own performance in order to
1. Introduction realize their flaw. Prior to 1965 the performance of a
mutual fund were measured by comparing with other
Mutual funds have evolved over many years and have funds’ average returns. They just average the funds’
become an imperative tool to investors, especially small returns over a number of periods and then rank them
ones. The concept was first used in Netherlands in 1774, according to their highest returns. In spite of having the
but the modern day mutual funds came into existence in importance of risks, very few investors would consider
1924, thereafter it tended to obtain popularity. risks in their analysis. Although since 1960s investors
Nonetheless, the term ‘mutual fund’ can be defined as one have knowledge about quantifying and measuring risk
variety of investment vehicles that collects funds from with respect to variability of returns, no single tool
various investors and professionally invests the same in essentially considered risk and return simultaneously.
diversified assets; like stocks, bonds, money market Treynor [25] was the first to address this issue and provide
securities or other assets; to facilitate forming portfolio. a way to measure the funds’ risk-adjusted performance.
Formerly, it was assumed that investment diversification Treynor ratio calculates excess returns of fund for per unit
was the foremost attraction of mutual funds, but now the of systematic risk measured by beta. Another approach to
purposes of mutual funds extended widely. Suppose, calculate fund’s return over and above of risk free return
someone has funds, but may not have enough time, for per unit of total risk measured by standard deviation is
expertise and even resources to undertake such large called Sharp ratio developed by William Sharpe [21]. This
diversification while a mutual fund can do. In contrast, ratio is close to Treynor ratio and employs beta as the
someone may necessitate fixed income to make payment measure of risk, whereas the Sharp ratio utilizes standard
of a loan. Now a day, mutual fund managers are offering deviation to quantify risk. However, for completely
investors with much wider customized funds to deal with diversified portfolio, the two measures give the identical
their various investments needs. ranking for an entirely diversified portfolio as the total
Now the question is whether the funds mangers are risk and systematic risk of a fully diversified portfolio are
utilizing the investors’ money efficiently or not. Are same. Thereafter two years, to appraise the fund managers’
mutual funds are generating returns greater than market skills in assets selection, Jensen [12] proposes a model to
average? Here the performance of mutual fund is always calculate alpha (popularly called Jensen’s alpha) which
concerned for both the investors and fund managers alike. evaluates the additional return that the fund produces after
Investors seek this information to facilitate their adjusting for its systematic risk. In an efficient market, the
12 Journal of Finance and Accounting

expected Jensen’s alpha should be zero. A positive alpha (AIMS) took the initiative to float the first private sector
implies that the fund manager adds value which is mutual fund named AIMS First Guaranteed Mutual Fund
attributable to their skill in pooling superior assets into which truly expanded the orbit of the stock market in
their portfolio. A comparatively new method to calculate Bangladesh. The fund was listed at Dhaka Stock
risk-adjusted performance is M2 which is derived by Exchanges in May 2000 making a milestone in the history
Franco Modigliani and Leo Modigliani [18]. In estimating of the capital market in Bangladesh. However, only closed
M2, a scheme’s portfolio can be levered or de-levered to end mutual funds have been offered to investors till 2010.
reflect a standard deviation that is identical with that of the Prime Finance Asset Management Company Limited
market so that M2 can be directly compared with the return (PFAMCL) decided to go one step ahead and launched
in the market. Another risk-adjusted measure is thecountry’s first ever open-end mutual fund. As of April,
information ratio (IR), the ratio of average excess return to 2016, there are total 36 mutual funds are listed on the
tracking error (risk). Like the sharp ratio, the information Dhaka stock exchange. Even though lately banks and non-
ratio also calculate excess return per unit of risk with the bank financial institutions are also getting heavily
exception that IR measures excess return and risk relative involved in mutual fund initiation, this sector accounts
to a specific benchmark index instead of risk free return. less than one percent of the total market capitalization in
The aforesaid five techniques are popularly used to Bangladesh (Financial Express, 2015).
measure risk-adjusted performance of mutual funds. These Since the stock market size and depth were broadening,
techniques are generally used to inspect the selection skill mutual fund initiators hurdled to meet investors’ demand
of fund managers. by providing funds with different flavors. High volatility
In addition to risk-adjusted performance, mutual fund in stock market in Bangladesh always has been a concern
managers’ market timing skill is also tested. Market for policy makers. Even though investing in mutual fund
timing ability implies the competence of fund managers to is generally assumed as safe investment, paradoxically,
predict the market movement correctly, that is, fund mutual funds did attract much attention from investors.
managers will effectively “time the market by increasing
(decreasing) portfolio exposure prior to market bullish
(bearish)” [4]. Two methods are commonly employed to 3. Literature Review
examine the market timing skill of mutual fund managers:
the first one was developed by Treynor and Mazuy [24] Although the history of mutual funds in Bangladesh is
whereas the other was developed by Henriksson and very long around 33 years, it has not been flourished or
Merton [10]. Hence, from the above discussion it is developed yet. Mutual funds did not get proper attention
apparent that two types of skills are usually inspected to from our investors. The performance of mutual funds
measure the performance of mutual funds: one is security managers in Bangladesh is now questionable. Are they
selection skill and another is market timing skill. The efficient in producing abnormal return consistently? Do
abovementioned methods are employed in our study, fund managers possess market timing skill? For the
which are also detailed in the methodology section. greater interest of the investors we need to evaluate the
Many researchers and academicians are consistently performance of funds managers repeatedly. Extensive
having researches on mutual funds performance in the researches have been performed regarding the ability of
globe while a few worked on the same in Bangladesh. mutual fund managers to beat the market comparing to
Furthermore, researchers in Bangladesh used monthly data benchmark index in the developed countries. And a very
as well as small sample in their studies whereas we use few researchers in Bangladesh attempted to estimate the
weekly data of 25 mutual funds for the period of May 16, performance of mutual funds managers. However, a few
2010 to April 28, 2016. We tested both the selection and number of previous works documented fund managers’
market timing skills and found neither skill is pervasive to market timing skill and to produce abnormal return
the fund managers in Bangladesh. persistently. Now, we will discuss some of the important
The reminder of the paper is therefore organized as works related to our study.
follows. Section 2 describes the history of mutual funds in Treynor and Mazuy [24] offered an approach to
Bangladesh. Section 3 discusses the previous literature. examine fund managers’ market timing capability. They
Section 4 explains the data and methodology applied in concluded that the Using 57 open-ended mutual funds
the paper. Section 5 presents the result analysis of this from the US from 1953 to 1962, they found no convincing
research, and Part 6 concludes the study. support that the fund managers of any of the sample funds
had outperformed the market.
Jensen [12] assessed the talent of the fund managers in
2. Mutual Funds in Bangladesh picking undervalued stocks. He found that fund managers
from his 115 sample mutual funds failed so badly in
Investment Corporation of Bangladesh (ICB), a statutory forecasting security prices that result in unsuccessful
company of Bangladesh government that specialized recover research expenses and fees.
capital market intermediary, was launched on October 1, Using all mutual funds in the U.S. survived between
1976. With an aim to enhance market depth and product January 1, 1975 and January 1, 1995, analyzing
diversity, ICB issued the first ever mutual fund, the First book-to-market, size, momentum and turnover of stocks
ICB Mutual Fund, in Bangladesh on April 25th, 1980. It held in deciles, Chen, Jegadeesh, and Wermers [5] noticed
initiated some eight closed end mutual funds till 1996. It that growth funds have distinct ability in selecting
took about two decades for private firms to launch any undervalued large-growth stocks. They also observed that
mutual funds in Bangladesh. In March 2000, Asset and high-turnover funds are better in picking stocks than
Investment Management Services of Bangladesh Limited
 Journal of Finance and Accounting 13

low-turnover funds which have been confirmed by chosen. The report found no evidence that mutual fund
Wermers [26]. managers in India are able to time the market and instead
Kon and Jen [15] studied fund managers’ capability to rely only on the stock selection skill to maximize returns.
time the market and selectivity. They concluded that even Charleset et al. (2012) evaluated fund managers’ skill to
though fund managers did not possess superior market time market liquidity. Employing mutual fund data from
timing skill, little evidence is observed regarding CRSP database over the period of 1974 to 2009, they
selectivity. found evidence that fund managers exhibit the talent to
Jiang [12] applied a nonparametric test to a sample of time market wide liquidity not only at portfolio level but
U.S. funds that used diverse benchmark indices to also at the individual fund level. However, funds
investigate the market timing skill of mutual fund demonstrating liquidity-timing skill are likely to have
managers. However, the report failed to provide evidence extended histories, greater expense ratios, and larger
of superior market timing skill among fund managers turnover rates.
during 1980–1999. However, she agreed that it is Rahman et al. [19] examined the performance of
problematic to foretell funds’ timing ability from their sixteen growth oriented mutual funds listed on the DSE.
visible features. To this end, the study used monthly returns, the DSE
Kosowskiet et al. [16] employed bootstrap method to General Index (DGEN) as the benchmark and risk
net returns data of the universe of U.S. equity funds adjusted performance measures proposed by Jenson [12],
during the 1975 to 2002 period. They noted robust Treynor [25] and Sharpe [21] Their result suggests that
indication of superior performance which is persistence even though some funds have performed better according
mostly among growth oriented funds. However, no to Jenson and Treynor measure but did not perform well
support of market timing skill of income-oriented fund based on Sharpe ratio.
managers is observed. Hasan and Akhter [9] test the risk adjusted performance
Using a sample of 485 arbitrarily selected mutual funds of thirteen Bangladeshi mutual funds analyzing monthly
from the DataStream database for the period from 1997 to data for the period of March, 2007 to April, 2011 and
2002, Jeffrey [11] observed that mutual funds overall found evidence that most of the mutual funds performed
exhibit some evidence of negative sector timing abilities. better than DSE benchmark index (DS20). They also
However, managers in specific groups of funds, for tested fund managers’ various skills, including
example, aggressive growth funds, seem to hold better diversification, market timing and selectivity skill and
sector timing abilities compared to other types of funds, found evidence that neither skill of mutual funds was
and this is more obvious after controlling for market statistically significant in Bangladesh.
downturns.
Cuthbertson et al. [6] used 935 actual equity funds
performance data from UK between 1975 and 2002 4. Data and Methodology
combined with cross-section bootstrap approach to
examine whether superior fund performance is because of 4.1. Data
fund managers’ “skill” or “luck.” They observed stock
Weekly NAV of 25 mutual funds listed on the Dhaka
picking talent for about 5% to 10% of top performing
Stock Exchange (DSE) has been collected. The data
equity funds in UK. They also found that some of the top
period covers between May 16, 2010 and April 28, 2016.
performing equity-income funds show stock picking skills,
DSE General Index is taken as the benchmark or market
whereas such ability is generally not found among small
index while 91-day Treasury bill of Bangladesh is
stock funds and ‘all company’ funds. This is consistent
considered as the risk free rate applicable for Bangladesh.
with the recent empirical evidence found for the U.S.
Data on NAV and DGEN is collected from the DSE
[5,16].
library. Treasury bill data is gathered from the Bangladesh
Scaillet et al. [20] applied the “False Discovery Rate”
Bank library.
(FDR) technique to the returns of 2,076 U.S. equity funds
that subsisted between 1975 and 2006, and noted that the
ratio of “skilled” fund managers has declined speedily, 4.2. Methodology
while the percentage of “unskilled” fund managers has The Sharpe ratio, also called reward-to-variability ratio,
amplified considerably. They also noticed that the ratio of is developed by William Sharpe [21] basically describes
“skilled” fund managers declined from 14.4% in early how well the return of fund 𝑖𝑖 compensates investors for
1990 to 0.6% in late 2006, whereas the proportion of the per unit risk taken. It is computed by dividing the
“unskilled” funds managers improved from 9.2% to excess return of fund 𝑖𝑖 by its volatility measured by
24.0%. standard deviation. Higher the Sharpe ratio of a fund is
Białkowski and Otten [1] used a sample of 140 Polish preferred. The formula is given below:
mutual funds for a period of 2000-2008 combined with
Carhart’s four factor model to investigate the Polish R p,i − R f
S p,i = .
mutual fund industry. They documented that mutual funds σ p,i
in Poland on average failed to add value, as shown by
their negative net alphas. Here, R p,i implies the average daily return for the fund
Miglani [17] examined fund managers’ market timing
skill during 1999 to 2004 in India. To this end, he 𝑖𝑖 and R f is the average daily risk-free rate, found by
employed Treynor and Mazuy [24] and Henriksson and dividing the return of the 91-day treasury bill of
Merton [10] approaches. With a sample of 98 mutual Bangladesh by 91, and 𝜎𝜎𝑝𝑝,𝑖𝑖 is the standard deviation of
funds with diverse investment objectives have been fund 𝑖𝑖’s return used to measure of the fund’s volatility.
 14 Journal of Finance and Accounting

However, an investor must take into account the length However, a general consensus in the investment arena is
of the time period used in the calculation of Sharpe ratio; that IR of 0.20 or 0.30 is superior [14].
longer time period tend to result in lower volatility To evaluate fund managers’ talent in choosing stocks
measures. Spurgin [23] showed that the annualized that could outperform the index on a risk-adjusted basis,
standard deviation of returns tends to be higher for shorter the Jensen’s alpha which derived from the following
periods: Daily returns have higher standard deviations regression model is used:
than weekly returns, which have higher deviations than
monthly returns. He concluded that stretching the ( )
R p,i − R f = α p,i + β p,i Rm − R f + ε p ,i .
estimation period may overstate the Sharpe ratio. Sharpe
recommended using short periods, such as, monthly, to Here, the coefficient 𝛼𝛼𝑝𝑝,𝑖𝑖 , also known as Jensen’s alpha,
estimate risks and returns and then annualizing the data measures the additional return that the fund 𝑖𝑖 earns after
[22]. He believed using multi-period returns complicates adjusting for its systematic risk. In an efficient market, the
the ratio because of compounding or potential serial expected Jensen’s alpha should be zero. A positive alpha
correlation. implies that the fund manager adds value. However, if a
Next we estimated the Treynor ratio, introduced by fund has a negative Alpha, it is placed below the security
Jack Treynor [25], to rate mutual funds based on managers’ market line (SML), and is underperforming what the
risk adjusted performance. CAPM would expect its performance to be. Funds with
positive (negative) alphas should be positioned at the top
R p,i − R f (bottom) of funds’ ranking. 𝑅𝑅𝑝𝑝,𝑖𝑖 means the daily
T p,i = .
β p,i portfolio’s return for the sample’s equity mutual fund 𝑖𝑖,
𝑅𝑅𝑚𝑚 is the return on the market portfolio and 𝑅𝑅𝑓𝑓 is the risk-
Here, 𝑅𝑅�𝑝𝑝,𝑖𝑖 and 𝑅𝑅�𝑓𝑓 are defined as above and 𝛽𝛽𝑝𝑝,𝑖𝑖 is the free rate, estimated by dividing the return of the 91-day
systematic risk of fund 𝑖𝑖. Like the Sharpe ratio, the higher treasury bill of Bangladesh by 91. The coefficient 𝛽𝛽𝑖𝑖
the Treynor ratio, the better is the performance of the fund. denotes the systematic risk of fund 𝑖𝑖, that is, the degree of
Modigliani and Modigliani [18] risk-adjusted performance fund’s sensitivity with respect to the movements of market
measure ( M 2 ) is relatively a new technique which is portfolio, and 𝜀𝜀𝑖𝑖 signifies the residuals of regression
closely linked to the Sharpe ratio. In calculating M 2 , a equation.
portfolio is levered or de-levered so that its standard After ranking all the 25 mutual funds on the basis of
deviation is equal to that of the market portfolio. And their performance, the consistency of ratings among the
the M 2 of a portfolio is the return that this adjusted four individual criterions have been tested, through the
portfolio earns. The reasoning behind this adjustment following single cross-sectional regression model:
practice is to compare portfolio’s adjusted return directly α + βRating j + u
Ratingi =
to the market return for the period. The traditional form of
the M 2 is given in the below: Where, the terms 𝑖𝑖 and 𝑗𝑗 denote the couple of ratings that
are regressed. Statistically significant estimations for beta
R p,i − RA
=M2 σ=
m + Rf (Sharpe's Ratio) ( σ m ) + R f . coefficient that reaches unity indicate high consistency
σ p,i among the regressed performance ratings.
To test the market timing ability of mutual fund
Thus, M 2 value of a fund refers to the additional return managers in Bangladesh, two different models have been
an investor can earn from holding a portfolio instead of employed. The first one was developed by Treynor and
the market index. Mazuy [24] and is expressed as below:

( )
The information ratio (IR), the ratio of average excess
return to tracking error, aims to capture the mean-variance R p − R f =α + β p,i Rm − R f + γ p,i ( Rm − R f )2 + ε p ,i
attributes of an active portfolio in a single number [8]. The
Where, 𝑅𝑅𝑝𝑝 = the return on the fund;
information ratio basically quantifies the magnitude of
excess return which is generated from the size of excess 𝑅𝑅𝑚𝑚 = the return on the market portfolio;
risk taken comparing to the benchmark. In this case the 𝑅𝑅𝑓𝑓 = the risk free term;
benchmark doesn’t have to be the risk-free rate. IR is 𝜀𝜀𝑖𝑖 = the random error term; and
calculated as below: 𝛼𝛼, and 𝛽𝛽are parameters of the model
In the above equation, 𝛾𝛾p,i measures mutual fund
Rp,i − RA manager’s market timing ability. If a fund manager can
IR p,i =
SP − A forecast market movement accurately, he will increase
(decrease) his fund’s exposure to market portfolio
� p,i = Return of the portfolio
R preceding to market up (decline), and 𝛾𝛾p,i will be positive
� A = Return of the index or benchmark
R caused by the convex function of fund’s return with regard
SP−A = Tracking error which is standard deviation of the to market return.
differences between returns of the mutual fund and the Another model used to test market timing ability of the
returns of the index. mutual fund managers was introduced by Henriksson and
A portfolio with information ratio of 1.0 is graded as Merton [10].
“exceptional,” whereas 0.75 and 0.50 is considered “very
good,” and “good” respectively [7]. After observing IRs ( )
R p − R f =α + β Rm − R f + γ [ D( Rm − R f )2 ] + ε p,i
over a 10-year horizon, Goodwin [8] noticed that even
among constantly outshining long-only managers, not Where, D is a dummy variable that equals 0 for (𝑅𝑅𝑚𝑚 >𝑅𝑅𝑓𝑓 )
many are able to maintain an IR of 0.50 or higher. and –1 for (𝑅𝑅𝑚𝑚 <𝑅𝑅𝑓𝑓 ).
 Journal of Finance and Accounting 15

The Henriksson-Merton market-timing measure allows ����

Where, 𝑅𝑅𝑚𝑚 = average weekly market return 𝑅𝑅𝑚𝑚1 , 𝑅𝑅𝑚𝑚2 ,
for the beta risk to be different in ex-post up and down 𝑅𝑅𝑚𝑚3 …… 𝑅𝑅𝑚𝑚𝑚𝑚 are returns for the 1st, 2nd, 3rd, and nth
markets. According to this model in each period fund week.
managers will try to predict whether or not the market will
have positive or negative excess returns (rm,t+1>0 or ( R p1 + R p 2 + R p3 +…+ R pn )
Rp =
rm,t+1<0). In favor of an optimistic view towards the N
market, a manager will most likely take more systematic
risk, i.e., beta, comparing to a pessimistic forecast about Where, 𝑅𝑅𝑝𝑝1 , 𝑅𝑅𝑝𝑝2 , 𝑅𝑅𝑝𝑝3 … 𝑅𝑅𝑝𝑝𝑝𝑝 are return for 1st, 2nd, 3rd and
the market. In other words, the mutual fund beta is lower nth week based on NAVs
in the case of a bearish market prediction, and the market N = total number of weeks of a year.
beta will be higher in the case of a bullish market
prediction.
Measurement of Return on Mutual Fund Schemes:
5. Empirical Results
MPt − MPt −1 5.1. Descriptive Statistics
Rm =
MPt −1 Table 1 represents the descriptive statistics of weekly
Where, 𝑀𝑀𝑀𝑀𝑡𝑡 = market price of a scheme for current week; returns of 25 funds and a comparable index, DGEN (a
𝑀𝑀𝑀𝑀𝑡𝑡−1 = market price of a scheme for the preceding week benchmark index of DSE). It is surprising to observe that
all the funds (except 6th ICB M.F., Aims 1st M.F. and
NAVt − NAVt −1 ICB AMCL 2nd NRB M.F.) including index have
R pt =
NAVt −1 negative mean returns during the sample period. If we
compare the funds’ average mean returns (-0.0014) with
Where, 𝑅𝑅𝑝𝑝,𝑖𝑖 = Return on fund 𝑖𝑖; that of index (-0.0011), the funds’ performance is inferior
𝑁𝑁𝑁𝑁𝑁𝑁𝑡𝑡 = Net asset value at time 𝑡𝑡; to index. Most of the funds are more risky than index as
𝑁𝑁𝑁𝑁𝑁𝑁𝑡𝑡−1 = The corresponding value at time 𝑡𝑡 − 1. their standard deviations are larger than that of index. The
funds’ average value of standard deviation is 10.9% while
R + Rm 2 + Rm3 +…+ Rmn
Rm = m1 the index’s standard deviation is 5.70% which is much
N lower than that of the funds’ average.

Table 1. Descriptive Statistics of weekly returns

Name of Funds Mean Median Min Max Standard Deviation Kurtosis Skewness
DGEN Index Return -0.0011 0.009 -0.265 0.297 0.057 2.047 -0.072
1st ICB M.F. -0.007 0.000 -0.899 0.117 0.082 101.773 -9.247
2nd ICB M.F. -0.007 0.002 -0.904 0.204 0.090 71.415 -7.065
3rd ICB M.F. -0.006 0.000 -0.903 0.158 0.087 78.756 -7.652
4th ICB M.F. -0.007 0.000 -0.905 0.168 0.087 80.846 -7.773
5th ICB M.F. -0.007 0.000 -0.905 0.165 0.088 77.658 -7.537
6th ICB M.F. 0.001 -0.001 -0.908 1.726 0.178 68.135 5.228
7th ICB M.F. -0.008 0.003 -0.903 0.235 0.090 70.554 -6.913
8th ICB M.F. -0.007 0.000 -0.903 0.168 0.089 74.231 -7.293
1st BSRS -0.008 0.002 -0.907 0.193 0.088 75.743 -7.428
Aims 1st M.F 0.058 0.003 -0.514 8.399 0.721 135.582 11.511
ICB AMCL 1st M.F. -0.010 -0.001 -0.880 0.221 0.091 59.391 -6.294
ICB AMCL Islamic M.F. -0.008 -0.005 -0.900 0.355 0.095 56.385 -5.598
Grameen M.F. one -0.004 0.003 -0.373 0.149 0.050 21.062 -2.757
ICB AMCL 1st NRB M.F. -0.008 -0.001 -0.894 0.227 0.089 68.287 -6.809
ICB AMCL 2nd NRB M.F. 0.008 -0.001 -0.898 3.429 0.308 109.889 9.512
Grameen One: Sch. Two -0.001 0.001 -0.150 0.123 0.034 4.624 -0.478
Prime Finance First M.F. -0.001 0.000 -0.365 0.598 0.078 27.382 2.474
EBL First Mutual Fund -0.001 -0.001 -0.185 0.214 0.039 13.367 -0.311
ICB AMCL Second M.F. -0.010 -0.003 -0.886 0.217 0.089 67.482 -6.755
ICB Employees Provident MF One: Sch. One -0.001 -0.004 -0.315 0.204 0.068 3.983 -0.157
Trust Bank First M. F. -0.001 0.001 -0.136 0.180 0.032 9.952 0.542
Prime Bank 1st ICB AMCL M.F. -0.002 0.000 -0.115 0.202 0.046 2.429 0.424
DBH First Mutual Fund -0.001 0.002 -0.152 0.133 0.033 5.343 -0.504
IFIC Bank 1st Mutual Fund -0.001 0.000 -0.120 0.087 0.024 6.456 -0.667
Phoenix Finance 1st MF -0.002 0.000 -0.120 0.220 0.049 2.834 0.575
Funds’ Average performance -0.0014 0.000 -0.606 0.724 0.109 51.742 -2.439
Another observation is that among all the funds, Aims (5.80%), then ICB AMCL 2nd NRB M.F. generates the
1st mutual fund generates the highest mean returns second highest (0.8%) and the third highest is 6th ICB
16 Journal of Finance and Accounting

M.F. (0.1%). These three funds have generated their mean total average return, funds’ average return (– 0.13%) is
returns significantly greater than that of index. Although less than index’s return (-0.10%). The index is ranked on
the mean return of IFIC Bank 1st Mutual Fund is negative, eighth position, implying that most of the funds generating
it has the lowest risk with the standard deviation of 2.4% average return less than that of index. The funds’ average
whereas the Aims 1st mutual fund’s returns has the highest Sharpe ratio is -0.144 while the corresponding value for
risk with highest mean return. Based on the descriptive the index is -0.132 which is better than funds’ average
statistics, we can conclude that most of the funds’ ratio. According to Sharpe ratio, index is ranked on sixth
managers have failed to perform better than the market place revealing that most of the funds are performing
portfolio; i.e. index. worse than index. Akin to average return and Sharpe ratio,
the remained measures; Treynor, information ratio,
5.2. Risk-adjusted Performance Analysis Jensen’s alpha and M square also reveal the same results.
The average values of Treynor ratio, information ratio,
Table 2 reports the results of the funds’ as well as Jensen’s alpha and M square for the funds are -0.041,
index’s risk-adjusted performance as measured by total -0.031, -0.001 and -0.002 respectively whereas the
average return, Sharpe ratio, Treynor ratio, Information corresponding values for index are -.007, 0, 0 and -0.001
ratio, Jensen’s alpha and Modigliani and Modigliani (M respectively. The index is ranked on third, eighth, fourth
square). Funds are ranked based on the value of these six and sixth according to Treynor ratio, information ratio;
measures and compare with index return. According to Jensen’s alpha and M square correspondingly.

Table 2. Risk-adjusted performance measures (Rating Analysis)

Total Average Modigliani–Modigliani
Name of Mutual Sharpe Ratio Treynor Ratio Information Ratio Jensen’s Alpha
Return (M square)
Funds
Value Ranking Ratio Ranking Ratio Ranking Ratio Ranking Alpha Ranking M square Ranking
DGEN Index
-0.0010 8 -0.132 6 -0.007 3 0 8 0.000 4 -0.001 6
Return
1st ICB M.F. -0.006 15 -0.148 12 -0.052 18 -0.058 15 -0.006 20 -0.002 12
2nd ICB M.F. -0.007 19 -0.143 9 -0.037 12 -0.063 19 -0.006 16 -0.002 9
3rd ICB M.F. -0.007 20 -0.147 11 -0.038 13 -0.064 20 -0.006 18 -0.002 11
4th ICB M.F. -0.007 21 -0.150 13 -0.046 17 -0.065 21 -0.007 22 -0.002 13
5th ICB M.F. -0.007 18 -0.145 10 -0.045 16 -0.061 18 -0.007 21 -0.002 10
6th ICB M.F. 0.000 3 -0.032 3 -0.014 4 0.007 6 0.001 3 0.004 3
7th ICB M.F. -0.007 16 -0.140 7 -0.042 15 -0.059 16 -0.006 17 -0.001 7
8th ICB M.F. -0.007 17 -0.143 8 -0.041 14 -0.060 17 -0.006 19 -0.002 8
1st BSRS -0.008 22 -0.153 15 -0.018 5 -0.082 24 -0.004 15 -0.002 15
Aims 1st M.F 0.063 1 0.081 1 0.062 1 0.091 1 0.068 1 0.010 1
ICB AMCL 1st
-0.010 25 -0.170 18 -0.080 25 -0.085 25 -0.010 25 -0.003 18
M.F.
ICB AMCL
-0.008 24 -0.151 14 -0.071 22 -0.072 22 -0.009 24 -0.002 14
Islamic M.F.
Grameen M.F. one -0.004 14 -0.193 21 -0.021 9 -0.052 14 -0.002 12 -0.004 21
ICB AMCL 1st
-0.008 23 -0.160 16 -0.063 20 -0.076 23 -0.009 23 -0.003 16
NRB M.F.
ICB AMCL 2nd
0.009 2 0.011 2 0.003 2 0.034 2 0.014 2 0.006 2
NRB M.F.
Grameen One:
-0.001 10 -0.221 25 -0.019 6 -0.008 10 -0.001 7 -0.006 25
Sch. Two
Prime Finance
-0.001 7 -0.088 4 -0.055 19 0.001 7 -0.002 10 0.001 4
First M.F.
EBL First Mutual
-0.001 11 -0.194 22 -0.030 11 -0.008 11 -0.002 9 -0.004 22
Fund
ICB AMCL
-0.010 26 -0.177 19 -0.072 23 -0.090 26 -0.010 26 -0.003 19
Second M.F.
ICB Employees
Provident MF -0.001 9 -0.104 5 -0.075 24 -0.002 9 -0.002 11 0.000 5
One: Sch. One
Trust Bank First
-0.001 4 -0.199 23 -0.020 7 0.012 3 0.000 5 -0.005 23
M. F.
Prime Bank 1st
-0.002 13 -0.177 20 -0.130 26 -0.019 13 -0.004 14 -0.003 20
ICB AMCL M.F.
DBH First Mutual
0.000 6 -0.208 24 -0.020 8 0.007 5 -0.000 6 -0.005 24
Fund
IFIC Bank 1st
-0.001 5 -0.288 26 -0.025 10 0.010 4 -0.001 8 -0.009 26
Mutual Fund
Phoenix Finance
-0.002 12 -0.166 17 -0.069 21 -0.018 12 -0.003 13 -0.003 17
1st MF
Funds' Average
-0.0013 -0.144 -0.041 -0.031 -0.001 -0.002
performance
 Journal of Finance and Accounting 17

It is interesting to observe that among all the funds, basically indicates the fund managers’ asset selection skill
Aims 1st M.F is ranked top according to all the measures. into their portfolios. A positive alpha implies that the fund
ICB AMCL 2nd NRB M.F. is ranked second, 6th ICB managers add value which is attributable to their skill in
M.F. is ranked three while ICB AMCL Second M.F. is pooling superior assets into their portfolios. Here it is
ranked lowest by most of the measures. Top rank divulges observed that most of the funds’ alpha is negative
the securities selection skill of the funds’ managers. Thus, illuminating that the securities selection skill of most of
the fund managers of Aims 1st M.F, ICB AMCL 2nd the fund managers is very poor. They pooled wrong assets
NRB M.F. and 6th ICB M.F. have evidently better into their portfolios which resulting negative alpha (other
selection skill than other funds’ managers. Hence, skills are penalized by this negative value). Only Aims 1st
considering all the six measures it is evidenced that most M.F, ICB AMCL 2nd NRB M.F., 6th ICB M.F., DBH
of the funds are ranked after the index, meaning that most First Mutual Fund and Trust Bank First M. F. have
of the funds’ managers happen to failed to perform positive alpha signifying that their fund managers become
superiorly than the index as a proxy of market portfolio. able to add value to the portfolio. Our result is consistent
with Jensen [12], Kon and Jen [15], Białkowski and Otten
5.3. Regression Analysis Using Jensen’s [1], Rahman et al. [19] and Hasan and Akhter [9] who also
Alpha Model confirm the inability of fund managers to show assets
selection skill.
Table 3 represents the results of the regression of the
Jensen’s alpha model for all the funds. Jensen’s alpha

Table 3. Regression results of Jensen’s Alpha (1969) Model

Name of Mutual Funds Alpha T-test Beta T-test R-Square Observation
1st ICB M.F. -0.006 -0.942 0.236 1.822* 0.023 291
2nd ICB M.F. -0.006 -0.855 0.349 2.495** 0.042 291
3rd ICB M.F. -0.006 -0.885 0.342 2.510** 0.043 291
4th ICB M.F. -0.007 -0.964 0.289 2.114** 0.031 291
5th ICB M.F. -0.007 -0.908 0.286 2.074** 0.030 291
6th ICB M.F. 0.001 0.091 0.426 1.511 0.016 291
7th ICB M.F. -0.006 -0.854 0.300 2.130** 0.031 291
8th ICB M.F. -0.006 -0.870 0.308 2.216** 0.034 291
1st BSRS -0.004 -0.647 0.737 5.810*** 0.193 291
Aims 1st M.F 0.068 1.127 0.900 0.795 0.004 291
ICB AMCL 1st M.F. -0.010 -1.309 0.201 1.391 0.014 291
ICB AMCL Islamic M.F. -0.009 -1.111 0.202 1.343 0.013 291
Grameen M.F. one -0.002 -0.648 0.457 6.554*** 0.234 291
ICB AMCL 1st NRB M.F. -0.009 -1.155 0.231 1.636 0.019 291
ICB AMCL 2nd NRB M.F. 0.014 0.554 0.994 2.054** 0.029 291
Grameen One: Sch. Two -0.001 -0.268 0.374 8.944*** 0.362 291
Prime Finance First M.F. -0.002 -0.286 0.128 1.038 0.008 291
EBL First Mutual Fund -0.002 -0.523 0.246 4.315*** 0.117 291
ICB AMCL Second M.F. -0.010 -1.354 0.221 1.576 0.017 291
ICB Employees Provident MF One: Sch. One -0.002 -0.406 0.096 0.889 0.006 291
Trust Bank First M. F. 0.000 -0.032 0.328 7.549*** 0.288 291
Prime Bank 1st ICB AMCL M.F. -0.004 -0.943 0.065 0.885 0.006 291
DBH First Mutual Fund 0.000 -0.103 0.333 7.890*** 0.306 291
IFIC Bank 1st Mutual Fund -0.001 -0.406 0.263 9.246*** 0.417 291
Phoenix Finance 1st MF -0.003 -0.790 0.120 1.547 0.097 291
Note. *** Significance at 1% level; ** 5% level; and * 10% level.
the model is the gauge of testing market timing skill of
5.4. Regression Analysis for Testing Market fund managers. The positive value of gamma indicates
Timing Skill timing in right direction while the negative value means
Market timing skill implies the fund manager’s ability wrong direction. In Table 4, nearly all the funds’ gamma
to estimate the market movement correctly. As we estimates are negative and none of the t-value of gamma
referred earlier that we applied two models; one was estimate is statistically significant at 10% level. In
developed by Treynor and Mazuy [24] and the other was particular, out of 25 funds, only 3 funds’ gamma values
developed by Henriksson and Merton [10] to facilitate are positive, but having no statistical significance.
checking the market timing skill of fund managers in Therefore, applying Treynor and Mazuy [24] method it is
Bangladesh. Table 4 reports the regression results of the evidenced that none of the funds’ managers happen to able
Treynor and Mazuy [24] model. The value of gamma of timing the market correctly.
18 Journal of Finance and Accounting

Table 4. Regression results of Treynor & Mazuy (1966) Model

Name of the Mutual Funds Alpha Beta Gamma t(Alpha) t(Beta) t(Gamma) R2
1st ICB M.F. -0.033 0.224 -1.234 -0.409 1.732 -0.987 0.030
2nd ICB M.F. -0.013 0.337 -1.259 -0.360 2.316** -0.920 0.048
3rd ICB M.F. -0.009 0.330 -1.274 -0.371 2.315** -0.956 0.049
4th ICB M.F. -0.002 0.258 -1.174 -0.473 1.936* -0.886 0.036
5th ICB M.F. -0.005 0.272 -1.403 -0.352 1.872* -1.047 0.037
6th ICB M.F. 0.007 0.389 -1.576 0.334 1.392 -0.572 0.018
7th ICB M.F. -0.004 0.267 -1.345 -0.332 1.936* -0.983 0.038
8th ICB M.F. -0.003 0.297 -1.214 -0.382 2.034** -0.900 0.039
1st BSRS -0.005 0.729 -0.430 -0.424 5.653*** -0.349 0.194
Aims 1st M.F 0.083 0.779 -6.982 1.289 0.677 -0.634 0.007
ICB AMCL 1st M.F. -0.006 0.182 -1.081 -0.831 1.241 -0.772 0.018
ICB AMCL Islamic M.F. -0.006 0.183 -1.087 -0.667 1.198 -0.743 0.017
Grameen M.F. one -0.003 0.470 0.163 -0.685 6.481*** 0.240 0.234
ICB AMCL 1st NRB M.F. -0.005 0.221 -1.184 -0.654 1.468 -0.863 0.024
ICB AMCL 2nd NRB M.F. -0.002 1.089 5.484 -0.019 2.223** 1.170 0.038
Grameen One: Sch. Two -0.001 0.387 0.170 -0.445 8.863*** 0.419 0.363
Prime Finance First M.F. 0.003 0.099 -1.668 0.359 0.795 -1.394 0.021
EBL First Mutual Fund -0.001 0.244 -0.123 -0.370 4.204*** -0.222 0.117
ICB AMCL Second M.F. -0.007 0.201 -1.172 -0.844 1.409 -0.860 0.022
ICB Employees Provident MF One: Sch. One 0.001 0.072 -1.411 0.239 0.655 -1.345 0.018
Trust Bank First M. F. 0.000 0.325 -0.121 0.098 7.372*** -0.286 0.288
Prime Bank 1st ICB AMCL M.F. -0.001 0.046 -1.099 -0.168 0.620 -1.543 0.022
DBH First Mutual Fund 0.000 0.330 -0.154 0.074 7.693*** -0.374 0.307
IFIC Bank 1st Mutual Fund 0.000 0.259 -0.220 -0.022 8.973*** -0.796 0.380
Phoenix Finance 1st MF -0.001 0.104 -0.953 -0.160 1.318 -1.266 0.028
Note. *** Significance at 1% level; ** 5% level; * 10% level;’t’ indicates t-statistics & Total observation =143.

Table 5. Regression results of Henriksson & Merton (1981) Model

Name of the Mutual Funds Alpha Beta Gamma t(Alpha) t(Beta) t(Gamma) R2
1st ICB M.F. 0.004 0.454 -0.509 0.308 2.164** -1.305 0.021
2nd ICB M.F. 0.003 0.544 -0.454 0.202 2.409** -1.077 0.037
3rd ICB M.F. 0.004 0.554 -0.494 0.273 2.518** -1.204 0.039
4th ICB M.F. 0.003 0.478 -0.464 0.163 2.208** -1.126 0.026
5th ICB M.F. 0.005 0.536 -0.588 0.411 2.411** -1.415 0.030
6th ICB M.F. 0.011 0.615 -0.439 0.438 1.350 -0.515 0.004
7th ICB M.F. 0.006 0.541 -0.590 0.429 2.426** -1.390 0.031
8th ICB M.F. 0.003 0.512 -0.501 0.278 2.328** -1.196 0.030
1st BSRS 0.002 0.864 -0.310 0.150 4.286*** -0.810 0.185
Aims 1st M.F 0.127 2.195 -3.173 1.447 1.221 -0.928 -0.004
ICB AMCL 1st M.F. -0.001 0.398 -0.483 -0.085 1.738* -1.109 0.008
ICB AMCL Islamic M.F. 0.002 0.438 -0.578 0.170 1.838* -1.275 0.010
Grameen M.F. one -0.004 0.428 0.071 -0.685 3.858*** 0.334 0.223
ICB AMCL 1st NRB M.F. 0.001 0.440 -0.512 0.087 1.966** -1.202 0.015
ICB AMCL 2nd NRB M.F. -0.029 0.069 2.266 -0.758 0.091 1.559 0.032
Grameen One: Sch. Two -0.002 0.346 0.070 -0.585 5.195*** 0.551 0.354
Prime Finance First M.F. 0.009 0.368 -0.587 0.958 1.885* -1.581 0.011
EBL First Mutual Fund 0.000 0.285 -0.096 0.051 3.146*** -0.559 0.106
ICB AMCL Second M.F. -0.001 0.414 -0.472 -0.112 1.860* -1.116 0.012
ICB Employees Provident MF One: Sch. One 0.005 0.251 -0.378 0.566 1.459 -1.156 0.001
Trust Bank First M. F. 0.001 0.354 -0.064 0.335 5.123*** -0.488 0.279
Prime Bank 1st ICB AMCL M.F. 0.002 0.193 -0.313 0.383 1.654* -1.411 0.005
DBH First Mutual Fund 0.002 0.382 -0.120 0.617 5.701*** -0.941 0.301
IFIC Bank 1st Mutual Fund 0.001 0.307 -0.108 0.646 6.821*** -1.265 0.376
Phoenix Finance 1st MF 0.002 0.230 -0.268 0.294 1.862* -1.142 0.012
Note. *** Significance at 1% level; ** 5% level; * 10% level;’t’ indicates t-statistics & Total observation =143.
 Journal of Finance and Accounting 19

Table 5 presents the regression results of the [2] Bollen, N.P., &Busse, J.A. (2001). On the timing ability of mutual
Henriksson and Merton [10] model. In accordance with fund Managers. Journal of Finance, 56(3), 1075-1094.
[3] Bollen, N.P., &Busse, J.A. (2004). Short-term persistence in
the Treynor and Mazuy [24] model, the gamma of this mutual fund performance. Review of Financial Studies, 18(2),
model is also the measure of testing market timing skill 569-597.
and the same decision rule will also be applicable here. [4] Cao, C., Simin, T. T., & Wang, Y. (2013). Do mutual fund
The Henriksson and Merton [10] model produces almost managers time market liquidity? Journal of Financial Markets,
the same results as the Treynor and Mazuy [24] model did 16(2), 279-307.
[5] Chen, H. L., Jegadeesh, N. &Wermers, R. (2000). The value of
earlier. Out of 25 funds, only 3 funds’ gamma values are active mutual fund management: an examination of the
positive, but having no statistical significance at 10% level. stockholdings and trades of fund managers. Journal of Financial
Thus, both models evidences that no market timing ability and Quantitative Analysis, 35(3), 343-368.
for fund’s managers is persistent in Bangladesh. The [6] Cuthbertson, K., Nitzsche, D., & O’Sullivan, N. (2008). UK
results are consistent with Treynor and Mazuy [24], Kon mutual fund performance: Skill or luck? Journal of Empirical
Finance, 15(4), 613-634.
and Jen [15], Jiang [13] and Miglani [17] who also found [7] Grinold, R. C., & Kahn, R. N. (1995). Active portfolio
no evidence for market timing ability of fund managers. management. Chicago, IL: Richard D. Irwin.
Indeed, very few researchers in the world found the [8] Goodwin, T. H. (2009). The Information ratio. Financial Analysts
persistence of market timing skill of fund managers, but Journal, 54(4), 34-43.
selection skill is commonly evidenced. [9] Hasan, M. H., and Akhter, T. (2011). Risk Adjusted Performance
of Mutual Fund: Evidence from Bangladesh. Journal of Finance
and Banking, 9(1), 1-24.
[10] Henriksson R.D., & Merton, R.C. (1981). On the market timing
6. Conclusions and investment performance of managed portfolios II. Statistical
procedures for evaluating forecasting skills. Journal of Business,
In this study, we inspected the ability of fund managers 54(4), 513-533.
to outguess the market in Bangladesh. We investigated [11] Jeffrey Junhua Lu (2007). Can Mutual Fund Managers Outguess
Sectors? Social Science Research Network, Working Paper Series,
both selection and market timing skills to check the Retrieved from [Link]
aptitude of fund managers to outguess the market. We [12] Jensen, M.C. (1969). Risk, the pricing of capital assets, and the
found evidence that unlike Aims 1st M.F, ICB AMCL 2nd evaluation of investment portfolios. Journal of Finance, 24(5),
NRB M.F. and 6th ICB M.F., most of the fund managers 959-960.
[13] Jiang, W. (2003). A nonparametric test of market timing. Journal
can show neither skill. The three funds’ managers; Aims of Empirical Finance, 10, 399-425.
1st M.F, ICB AMCL 2nd NRB M.F. and 6th ICB M.F; [14] Kidd, D. (2011). The Sharpe ratio and the Information ratio.
become able to show their assets selection skill, but fail to Retrieved from
demonstrate market timing skill. Out of the six risk- [Link]
adjusted measures most of the measures ranked these three [15] Kon, Stanley J. and Jen, C. Frank (1979). The investment
performance of mutual funds: An empirical investigation of timing,
funds top and even ahead of benchmark index. In addition, selectivity, and market efficiency. Journal of Business, 52(2),
unlike these three funds, all have negative value of alpha 263-289.
evidencing that fund managers have no selection skill and [16] Kosowski, R., Timmermann, A. Wermers, R. & White, H. (2006).
even they pool wrong assets into their portfolios. Even Can mutual fund “stars” really pick stocks? New evidence from a
bootstrap analysis. Journal of Finance, 61(6), 2551-2595.
though, three funds’ gamma values are positive in both
[17] Miglani, S. K. (2011). Market Timing Ability of Indian Mutual
models, but having no statistical significance. Thus, with a Funds. VSRD International Journal of Business and Management
few exceptions, we can conclude that neither skill is Research, 1(7), 416-427.
persistent to the funds’ managers in Bangladesh. This [18] Modigliani, F., & Modigliani, L. (1997). Risk-Adjusted
result is not surprising as the market is straggling since the Performance. The Journal of Portfolio Management, 23(2), 45-54.
early 2011 due to market turmoil. Most of the stocks’ [19] Rahman, A.B.M. M., Qiang, Prof. F. and Barua, S. (2012). Mutual
Fund Performance: An Analysis of Monthly Returns of an
price goes down very sharply just after 2010, which is still Emerging Market, Research Journal of Finance and Accounting,
continuing. But, this reason cannot give comfort to the 3(4), 34-46.
fund managers to conceal their ineptness in assets [20] Scaillet, O., Barras, L., & Wermers, R. R. (2010). False
management of their portfolio. More researches are discoveries in mutual fund performance: Measuring luck in
estimated alpha. Journal of Finance, 65(1), 179-216.
welcome with updated data and methodology to unfold the [21] Sharpe, W. F. (1966). Mutual fund performance. Journal of
causes of poor performance of fund managers. Business, 39(1), 119-138.
Researchers are also suggested to include more variables [22] Sharpe, W. F. (1994). The Sharpe ratio. Journal of Portfolio
in their researches in order to get more vibrant results in Management, 21(1), 49-58.
this regard. [23] Spurgin, R. (2001). How to game your Sharpe ratio. Journal of
Alternative Investments, 4(3), 38-46.
[24] Treynor, J., & Mazuy, K. (1966). Can mutual funds outguess the
market? Harvard Business Review, 44(4), 131-136.
References [25] Treynor, J. (1965). How to rate management of investment funds.
Harvard Business Review, 43(1), 63-75.
[1] Białkowski, J. & Otten, R. (2011). Emerging market mutual fund [26] Wermers, R. (2000). Mutual fund performance: An empirical
performance: Evidence for Poland. The North American Journal decomposition into stock-picking talent, style, transaction costs,
of Economics and Finance, 22(2), 118-130. and expenses. Journal of Finance, 55(4), 1655-1703.

View publication stats