Shopping cart
0 Items | US$ 0.00
Symbols
(Alpha) Outperformance of mutual fund to the benchmark index
(Beta) Mutual fund sensibility of co movement to the benchmark index
,
(Epsilon) Error term of regression
,
Return on a risk free security
,
Return on a mutual fund
,
Return on a benchmark index
,
Weights of asset i in the fund
,
Weights of asset i in the benchmark index
IV
Figures
Illustration 1: Weights of the Morningstar rating. ... 10
Illustration 2: Different types of active and passive management ... 46
V
1
Introduction
1.1
Investments in Mutual Funds
The assets under management (AuM) invested in mutual funds worldwide amount to
$26.8 trillion at the end of 2012. Approximately 50% of these assets are invested in the
United States (US), followed by Europe with 30%.
1
The importance of mutual funds for
individual investors in general has increased in recent decades. This becomes apparent
when looking at the increased share of households owning mutual funds within
individual retirement plans and defined contribution plans. Younger households in the
US tend also to participate stronger in mutual funds,
2
which implies that these retail
investors are holding investments over longer periods and might even have larger
invested stakes in mutual funds. I assume this pattern to apply to the European market,
as well. In Germany, over 50% of assets in mutual funds are hold by individual
investors at the end of 2008. This number has decreased from around 70% in 2005.
3
I
further assume that this decrease is a consequence of the financial crisis, since
individual investors are more likely to follow a general market climate than investing
when markets are down.
4
The main purpose of mutual fund investors is usually receiving a return which is above
or at least close to the mutual fund's benchmark.
5
Consequently investors want identify
and invest in those mutual funds which show these patterns in the future. Some mutual
funds get much on attention since they generate extraordinary high performance;
examples for these mutual funds are the Fidelity Magellan fund or the Schroder Ultra
Fund. The last mentioned fund yields a performance of 107% annually in the time
period of 1999 to 2001.
6
But the question when looking at these mutual funds is: Is it
possible to predict such performance before funds exhibit such outstanding
performance? Recent studies outlined in foot note 7 below showed strong doubts
whether this is possible, since mutual fund performance, at least in the short run, is
rather a random walk than a result of managerial skill.
7
Academics go even further
while stating that active managed mutual funds on average underperform their
1
Cf. ICI Factbook (2013) p. 25.
2
Cf. ICI Factbook (2013) p.116.
3
Cf. Jank (2010) p. 8.
4
Cf. Celati (2004) p. 66.
5
Cf. Barras/Scaillet/Wermers (2010) p. 183.
6
Cf. Kosowski, et al. (2006) p. 2551.
7
Cf. Barras/Scaillet/Wermers (2010) p. 180; Fama/French (2010) p. 1916; Kosowski, et al. (2006)
p. 2554.
1
corresponding benchmark by the costs in expenses.
8
Furthermore, there is some
evidence that the proportion of unskilled managers have increased in recent years while
the proportion of skilled managers have diminished.
9
This finding makes it harder to
"identify" those mutual funds managed by skilled executives.
In the past, mutual fund investors focused extensively on performance or performance
linked patterns, like the Morningstar star rating and thus chased intensively past
performance.
10
This seems somewhat surprising since performance persists only over
short horizons and is also more persistent for weak mutual funds (1 und 2 star rated)
than well performing mutual funds since the distinction between average and top
performing funds is hardly possible.
11
Thus chasing past performance seems to be a
rather inferior strategy.
Consequently investors should have some interest to identify alternative tools, which
show a high correlation to future mutual fund performance. One of these tools could be
the 2004 introduced Morningstar Stewardship rating for mutual funds, which provides
investors with information about some governance related issues of mutual funds.
12
This might be of special interest for investors since some unfavorable patterns of many
mutual funds are partly the result of weak governance, i.e. a weak alignment of
investors' and managers' interests. In many cases such patterns result in an inferior
performance.
13
A working paper published in 2008 suggests that investors seem to react
on the Morningstar Stewardship rating by redeeming or purchasing mutual fund
shares.
14
8
Cf. Fama/French (2010) p. 1941.
9
Cf. Barras/Scaillet/Wermers (2010) p. 199.
10
Cf. Del Guercio/Tkac (2001) p.37; Del Guercio/Tkac (2008) p. 918.
11
Cf. Blake/Morey (2000) p. 467.; Brown/Goetzmann (1995) p. 691 - 693; Carhart (1993) p. 57;
Hendricks/Patel/Zeckhauser (1993) p. 93.
12
Cf. Morningstar Inc. (2010b) p. 1; Wellman/Zhou (2008) p.1.
13
Cf. Weak alignment of investors and mutual fund managers may be the result of various reasons. For
example management and board link pattern, but as well patterns linked to the mutual fund itself like
the affiliation to a bank or if the mutual fund is public hold or private hold.
14
Cf. Wellman/Zhou (2008) p.1.
2
1.2
Remarks and Structure of the Study
This study focuses on open end mutual funds and do not look at closed end mutual
funds, since closed end mutual funds are much less common and have some different
characteristics. Additionally I focus mainly on actively managed mutual funds, as these
funds are more common
15
and also show more prosperities than passive mutual funds.
Nevertheless, when analyzing some aspects I will look at passive mutual funds, since it
is easier to infer on investors' behavior. This is due to the fact that some active mutual
fund attributes blur some effects and behavior pattern.
It should be noted that the vast majority of literature in this study are based on works
relating to the US market, other markets are only considered partly. Thus when adopting
these results, those results linked to regulatory constraints and requirements might be
misleading in particular. Nevertheless, the basic statements should be valid for markets
outside the US as well, especially those linked to the behavior pattern and incentives.
The study is structured as follows:
x In chapter two I give a short overview about the economics and both advantages
and disadvantages of mutual funds.
x In chapter three I focus on mutual fund performance and raise the question of the
purpose and usefulness of active investments. Based on that I ask whether
performance is persistent and what the determinants of mutual fund flows are.
x In chapter four I will analyze mutual funds in general and also focus on some
influence factors on its performance. Some points in this chapter are of rather
theoretic nature.
x In chapter five I will introduce some alternative measures and will also explain
which of the aforementioned attributes or methods should be used or rather avoided
when selecting mutual funds.
x In chapter six I will conclude the results and will and briefly address open
questions.
15
The share of closed end funds for the US amounts less than 2% at the end of 2012. See ICI Factbook
(2013) p. 2 (front cover).
3
2
Mutual Funds
2.1
Mutual Fund Basics
2.1.1
Overview and Economies
Open end mutual funds could be distinguished in different ways. Important distinctions
are active or passive managed mutual funds. Passive mutual funds or index mutual
funds are used to be "replicating the return on an index with a strategy of buying and
holding all (or almost all) index stocks in the official index proportions"
16
with relative
low costs.
17
When extending index funds to Exchange traded funds (ETFs), the defini-
tion above is no more valid, since sometimes those funds are swap-based with the result
that the assets in the portfolios of these funds are sometimes totally uncorrelated to the
index mirrored.
18
By assuming that markets are efficient and thus generating outperfor-
mance in active investments is nearly impossible, investments in passive funds might be
considered to be rational. Further reasons are that these funds do not trade actively and
consequently save costs for research and active trading.
19
Passive mutual funds are com-
monly associated with ETFs,
whereas lots of "classic" mutual funds are also tradable on
exchange, though they are no ETFs.
20, 21
An active investment is more complicated to
specify, basically it can be "defined as any deviation from passive management"
22
,
respective benchmark index or strategy. Furthermore, mutual funds differ by the assets
they are invested in (e.g. equity, bonds, mutual funds, etc.).
Alongside the increased importance of mutual funds for individual investors, funds are
also used by institutional investors. Increasing institutional holdings in mutual funds
show some important signals which may be partly supportive for retail investors when
selecting mutual funds.
23
Mutual fund markets around the world show partly different characteristics. The US
market for example is different to the rest of the world in regard of the average fund
size. US funds are on average five times bigger than European funds and 17 times
bigger than Asian funds.
24
16
Cremers/Petajisto (2009) p. 3334.
17
Cf. Fama/French (2010) p. 1915 f., 1922.
18
Cf. Heidorn/Winker/Löw (2010) p. 7.
19
Cf. Malkiel (2010) p. 119.
20
Cf. Heidorn/Winker/Löw (2010) p. 7.
21
ETFs are unique due to a so called "in kind process", which causes that each investor carry the costs of
redeeming or purchasing fund shares. See: Huang/Guedj (2009) p. 2.
22
Cremers/Petajisto (2009) p. 3334.
23
Cf. Chung/Zhang (2011) p. 254; Evans/Fahlenbrach (2012) p. 3530, Matsumura/Shin (2005) p. 107.
24
Cf. Ferreira et al. (2013) p. 488.
4
The existence of economies of scale and economies of scope in mutual funds are
broadly discussed in literature, but the results are diverse.
25
This can be argued by
different sources of such economies, which are sometimes supportive or obstructive to
these economies and these effects differ over diverse markets and different funds.
Economies or diseconomies of scale are influenced by many sources and thus the over-
all impact is not clear, since the single effects might be weighted out. Looking briefly at
the economies of scale they seem to be common features for the mutual funds AuM
outside the USA and for international invested funds located in the USA. In contrast,
domestic funds in the USA show diseconomies of scale.
The diseconomies of scale in
the USA are explained by liquidity constrains. This finding persists even when control-
ling for the size of US funds, which seems irritating because the US stock market is
more developed in comparison to other markets.
26
These conclusions are especially im-
portant because most of the mutual fund studies are located in the US. Thus interpreting
this finding as common feature for mutual funds worldwide might cause a bias.
Additionally, increasing AuM leads to adverse price movements, when buying or
selling the underlying stocks. This effect especially applies when the mutual fund is
becoming larger or when the fund is investing in smaller markets or less liquid stocks.
The increasing size also causes the problem that mutual funds are getting less flexible
compared to smaller funds when shifting portfolio holdings quickly. In addition, mutual
funds tend to scale their portfolio, meaning that they primarily increase their holdings in
those stocks they are already invested in, instead of searching for alternatives.
Investments in alternatives only take place when the costs of the current investments are
getting too high. This results in the fact that the diversifications of larger mutual funds
are only marginal higher than those of smaller ones.
27
The evidence of economies of scale and economies of scope on the fund family level is
much clearer. It can be explained due to savings in back office and market research
costs, since these costs are divided over a broader base.
28
Fund families usually run
different mutual funds which are invested in the same stocks or at least in the same
market. At the same time the essential research costs do not increase proportionally with
AuM and are not affected by increased trading sizes. The same is true for the back
office costs, because ordinary supervisory costs do not increase to the same extent as
25
See for example: Cf. Yan, X. ((2008); Chen et al. (2004).
26
Cf. Ferreira et al. (2013) p. 483, 488.
27
Cf. Chen et al. (2004) p. 1276ff., Pollet/Wilson (2008) p.2941 f., 2944.
28
Cf. Yan, X. ((2008) p. 760.
5
AuM increases. It is argued further that larger fund families are able to receive better
concessions on trading commissions.
29
Limits in economies of scale are induced due to
hierarchies which are commonly associated with larger funds, because managers have to
explain their decisions and investment strategies they want to implement.
30
2.1.2
Advantages and Disadvantages of Mutual Funds
The main advantages of investments in mutual funds are portfolio diversification, a
professional asset management, liquidity and usually an overall equity structure. The
diversification yields in a decrease of unsystematic risk, which individual investors can
hardly receive by investing their assets in single stocks.
The professional portfolio
management is important, because individuals are usually unable to scan the entire
market and identify "good" investing opportunities.
Liquidity is related to the fact that
investors can buy and redeem fund shares on a daily basis.
31
These advantages count
especially for less wealthy individual investors, who cannot afford a professional
portfolio manager or investing their assets diversified on their own.
However, mutual funds show some drawbacks. For example only a few (if any) active
mutual funds are able to outperform their benchmark after costs. Even before costs are
deducted the picture is only slightly better.
32
This raises the question if investments in
active mutual funds are reasonable, since on the long run such investments seem to be a
zero sum or even a losing game.
Beside this, especially in mutual funds affiliated to banking conglomerates or publicly
hold mutual fund families, the conflicts of interest hypothesis play an important role and
cause or may cause some serious problems. The conflicts of interest hypothesis is not
clearly defined but it can be summarized as a situation in which one party can take a
gain by actions harmful for the counterparty. Such actions require information
asymmetries. Consequently given a situation in which both parties are equally informed,
nobody can take advantages of these actions.
33
In our case the mutual fund manager
respective the financial institution has superior information against the mutual fund
investors. The conflict of interests in mutual funds may result for example in
unfavorable features like a lack of disclosure to investors, tax inefficiency and
29
Cf. Chen et al. (2004) p. 1278.
30
Cf. Chen et al. (2004) p. 1278f.; Ferreira et al. (2013) p. 518.
31
Cf. Ciccotello (2010) p. 5 ff.; Smith, (2010a ) p.38 ff..
32
Cf. Fama/French (2010) p. 1923, 1941; Ferreira et al. (2013) p. 485,493,518, Smith, (2010a) p.40.
33
Cf. Mehran/Stulz (2007) p. 267 - 269.
6
inadequate high fees and loads, window dressing and other questionable actions.
34
Some
of these actions became apparent in the scandals in the early 2000s. During those
scandals mutual funds allowed some investors to use abusive and partly illegal market
timing or late trading. The "costs" of these actions were mainly borne by buy-and-hold
investors,
35
such as retail investors holding mutual funds in their individual retirement
plans. Interestingly the participants in this scandal are mainly big players with nearly
20% or $1.2 trillion out of $7.2 trillion in long term assets und management.
36
Such a
situation does not necessary cause drawbacks for investors. In some situations investors
might even take profits from this superior information.
37
2.2
Mutual Fund Functionality
This chapter will present a brief overview about the functionality of mutual funds.
These are commonly organized as corporation or trust. They are supervised by a board
of inside and outside directors. Outside directors might have diverse backgrounds and
their number should amount for least 75% of the board members. Outside directors are
at least theoretically independent and should strengthen the power of the board and
should protect investors' rights as "independent watchdogs".
38
The board of directors
appoints the investment advisors respective the managers running the operative
business, who are commonly employed by an advisory firm. In addition, the board also
negotiates the fee with the advisory firm to be charged to the investors in the fund.
39,40
Mutual funds are created by mutual fund companies. After funds were initialized,
mutual funds provide seed money and operate outside the market for instance. After a
period of approximately 6 months the fund is opened for sale to investors. The return
history of this period is usually reported only, if the fund has performed well.
41
In the
day to day business, sell and purchase orders for fund shares are collected by the fund
and executed at the end of the trading day.
42
This characteristic is advantageous for the
34
Cf. Adams/Mansi/Nishikawa (2012) p. 2245; Kacperczyk/Sialm/Zheng (2006) p. 2379; Smith, (2010a)
p.40 ff..
35
Cf. McCabe (2009) p. 1.
36
Cf. Bogle, J. C. (2010) p. 283.
37
Mutual funds affiliated to credit bank usually profit from this relationship, see Massa/Rehman (2008)
p.288 ff. and chapter 4.5.3.
38
Cf. Chen/Huang (2011) p. 312; Ciccotello (2010) p. 4 ff.; SEC (1999) p. 2.
39
Cf. Ding/Wermers (2012) p. 6; Mehran/Stulz (2007) p. 271.
40
When a mutual fund is initiated the investment advisory firm is only shareholder, consequently he is
able to chose the "first" set of board members, see Kuhnen (2009) p. 2186. The conflict of interest
resulting on this structure is discussed in chapter 4.3.2.
41
Cf. Fama/French (2010) p. 1923f..
42
Cf. Huang/Guedj (2009) p. 2.
7
mutual fund. By collecting shares, the fund trades less because purchasing and selling
orders might be weighting out each other on a daily basis. Mutual funds must be able to
redeem and sell their share from investors at the net asset value every day. This
structure is called "self-liquidating". The feature requires the fund to hold some cash, or
at least highly liquid assets, which causes the so called cash-drag and can erode
performance.
43
The above described lagged structure brings some drawbacks for
investors because the price of the mutual fund might have dropped within that day. This
characteristic has arisen with the increasing importance of online banking and the broad
availability of ETFs, which are virtually tradable at any second at market prices. On the
other hand "classic" mutual funds are also tradable at exchanges, without meeting the
characteristics' of ETFs and therefore do not cause this process.
Mutual funds usually charge fees and expenses to their investors. Fees can be
distinguished between front-end load fee, back-end load fee
44
and operating expenses.
45
These costs vary across the globe and the different funds
46
. The front-end load fee is a
one-time upfront fee which is charged from investors when buying new shares. The fee
represents the costs for the fund associated with buying new stocks and ensures that the
existing investors are unharmed by the newly arrived money. The charge also contains
remunerations for financial intermediaries to sale shares to private investors. The back-
end load fees are charged when selling mutual funds. They cover the costs of selling the
underlying assets and should ensure that the remaining investors are unaffected of these
trades. Sometimes the back-end load fees are related to the invested amount and the
holding period and thus giving incentives for increased holding periods because the cost
of redeeming funds becomes lower when the holding period increases.
47
Problems with
this structure are that investors in the fund have to bear the price impact of asset inflows
and outflows. Some mutual funds introduce back-end fees to lower this impact.
48
The
ongoing expenses are covering the operational costs of mutual funds, such as the
management fee and other operative costs which include marketing expenses, employee
salaries and so on. They are charged on a daily basis as expense ratio based on net asset
value.
49
43
Cf. Ciccotello (2010) p. 5; Heidorn/Winkler/Löw (2010) p.10.
44
Synonyms for example are: redemption fee, redemption load, rear load, deferred load.
45
The cost compoinents may separated differently.
46
Cf. Khorana/Servaes/Tufano (2007) p. 1279.
47
Cf. Bechmann/Rangvid (2007) p. 667 f., Smith (2010b) p.52 f..
48
Cf. Bollen (2007) p. 687.
49
Cf. Iannotta/Navone (2012) p. 847.
8
2.3
Mutual Fund Ratings
2.3.1
Basics of Mutual Fund Ratings
Two common types of ratings are fund ratings and credit ratings. These two types show
very important distinctions. While a credit rating is forward looking and measures an
absolute default risk, fund ratings are commonly only backward looking and measure
the past performance relative to funds in a peer group and do not incorporate qualitative
facts.
50
In recent times some agencies extended or provided ratings with non-
performance factors.
51
There are a few independent rating agencies rating mutual funds
around the globe, the most influential one in the USA is Morningstar offering the "star
rating".
52
However, there are many more agencies rating mutual funds like Standard &
Poors, Feri, Lipper, Euro Fondsnote and Stiftung Warentest.
53
Conceptually (mutual
fund) ratings should provide unbiased, new and useful information to market
participants to help investors to allocate their assets.
54
Most mutual fund rating agencies
represent the results in a simplified score. Morningstar for example provide stars.
Especially naive investors tend to rely strongly on such heavily condensed ratings, since
retail investors are usually not able to scan the wide range of investment opportunities
mutual funds are offering. Rating agencies like Morningstar claim that investors should
use the rating only as a starting point and thus not rely solely on the star rating.
55
But the
empirical evidence, above all the US market, shows a different picture. The data shows
that changes in ratings cause immediate statistically significant market reactions.
56
The
European market (e.g. Germany and Finland) presents a different picture. The overall
market reaction is lagged and is much weaker.
57
The effect of changes in ratings
depends on various reasons like affiliation of investment funds and banks, the usage of
ratings for advertising purposes and maybe a less developed mutual fund market.
58
2.3.2
The Star Rating
Since the Morningstar "star rating" is the most influential mutual fund rating the most
literature refers to, it is outlined briefly below. This rating analyses funds and assigns
50
Cf. Hereil et al. (2010) p. 6 - 7.
51
Cf. Morningstar Inc. (2010b).
52
Cf. Morningstar Inc. (2010a); Carhart (1997) p. 66.
53
Cf. Meinhardt (2011) p. 1; Füss et al. (2010) p. 76.
54
Cf. Del Guercio/Tkac (2008) p. 909, 912; Terraza/Toque (2009) p. 149.
55
Cf. Füss et al. (2010) p. 76, Morningstar Inc. (2010a) p.1; Russel (2006) p. 85.
56
Cf. Del Guercio/Tkac (2008) p. 907, 918.
57
Cf. Füss et al. (2010) p. 83 ff..
58
Cf. Bessler/Drobetz/Zimmermann (2009) p. 287; Del Guercio/Tkac (2008) p.925; Gerrans, P. (2006)
p. 605 f.; Jank (2010) p. 3,6,13 f.,14,18; Russel (2006) p. 85.
9
one to five stars based on past performance. A five star rating represents the highest
score and one star the lowest. The rating scale builds on a normal distribution. The top
10% of funds receive five stars, the next 22.5% receive four stars, and the following
35% of funds from the medium segment and are rated with three stars. The next 22.5%
receives two stars and for the bottom 10% one star is assigned. The costs associated
with buying and selling mutual funds are deducted. The ongoing expenses are only
indirectly incorporated, since they are deducted from the returns. Furthermore the star
rating bases on three different time horizons. Time horizons and the weights for each
horizon are defined as follows:
Weights of the Morningstar rating
Age of fund
Overall rating
At least three years, but less than five
100% three-year rating
At least five years, but less than 10
60% five-year rating
40% three-year rating
At least 10 years
50% 10-year rating
30% five-year rating
20% three-year rating
Illustration 1: Weights of the Morningstar rating.
59
Funds younger than 3 years are not included in the rating. The fund's rating bases on the
Morningstar's categories defined for different markets and assets where funds are
invested in. 2002 the star rating has been changed, since the rating was biased by funds
invested in extraordinary well performing industries, Morningstar introduces new and
more detailed categories to solve this (potential) drawback. The Morningstar risk
adjusted returns now base on an extended expected utility theory.
60
In earlier periods
Morningstar has defined risk by a bilinear utility function.
61
Risk was defined as
underperformance to the 90 day treasury-bill. When the treasury bill was outperformed
the mutual fund has been classified as "riskless". The expected utility theory states that
investors are more averse against a potential loss than a potential positive outcome and
that investors are willing to give up some assets for a certain yield.
62
Additionally, since 2004 Morningstar provides a Stewardship Rating. It was introduced
after the scandals in the early 2000s and should allow investors identifying well and
poorly managed mutual funds.
63
59
Cf. Morningstar Inc. (2010a).
60
Cf. Del Guercio/Tkac (2008) p.927; Morningstar Inc. (2010a).
61
Cf. Sharpe (1998) p. 21.
62
Cf. Kahneman/Tversky (1979) p. 279; Lisi/Caporin (2009) p. 5; Morningstar Inc. (2006); Morningstar
Inc. (2010a).
63
Cf. Morningstar Inc. (2010b) p. 1; Phillips/Kaplan (2010) p. 171.
10
3
Mutual Fund Performance
3.1
What is Important about Mutual Fund Performance
Ratings mainly based on performance and fund performance itself are the dominant
investment criteria when looking at mutual funds. Especially the influence of the star
rating becomes apparent when observing the market reactions following changes in
Morningstar fund ratings in the USA.
64, 65
When looking at the mutual fund literatures
regarding the performance of mutual funds, three main scopes are important:
x Are mutual fund managers able to outperform their benchmark by generating a true
alpha? In other words: Are mutual fund managers really skilled? The empirical
evidence shows that skill seems to be a scare source and that the short run
performance is mainly derived by luck, good luck as well as bad luck.
66
x The next scope is predictability and persistence in mutual fund performance. This is
obviously difficult since performance is used to predict performance and
performance is at least partly a random walk.
67
Furthermore, performance
persistence is rather a short run phenomenon and is rather a result compared with
worse performing mutual funds.
68
x Mutual fund flows are of special interest, since both the fund's performance
influences the asset flows in and out of funds and the flows influence the
performance and vice versa. The performance directly influences the fund flows, as
investors chase past performance.
69
Inflows on the other hand lead to increasing
assets under management and the funds face decreasing economies of scale.
70
3.2
Active Investments
3.2.1
Why Using Active Strategies?
Investors use active investment strategies to outperforming a corresponding passive
alternative. Otherwise it would be rational to invest the assets in passive strategies like
an ETF, which yields the benchmark return minus marginal costs for "sure", without the
64
Cf. Del Guercio/Tkac (2001) p. 37; Del Guercio/Tkac (2008) p. 918; Knuutila/Puttonen/Smythe (2007)
p. 89.
65
See chapter 3.4 for a more detailed discussion
66
Barras/Scaillet/Wermers (2010) p. 179 ff.; Fama/French (2010) p. 1919 f.; Kosowski, et al. (2006)
p. 2551 - 2555.
67
Cf. Barras/Scaillet/Wermers (2010) p. 180; Fama/French (2010) p. 1916; Kosowski, et al. (2006)
p. 2554.
68
Cf. Hendricks/Patel/Zeckhauser (1993) p. 93.
69
Cf. Del Guercio/Tkac (2001) p.37; Del Guercio/Tkac (2008) p.918.
70
Cf. Blake/Morey (2000) p. 467.
11
risk of underperforming the benchmark substantially.
71
Due to the fact that stocks in
active mutual funds are traded actively and thus have other weights than the benchmark,
they may show higher risks than their passive counterparts. Furthermore, active mutual
funds charge higher fees on the expense of their investors to remunerate managerial
skill, research costs and higher trading costs.
72
Unfortunately studies show that at least
on average outperforming passive strategies is hardly possible
73
or even impossible as
on average a negative alpha indicates.
74, 75
Investors are going to anticipate the inability
to receive excess returns (in the long run) and now only want to collect at least the
return of the passive benchmark (alpha equals zero).
76
3.2.2
Skill in the Mutual Fund Industry
Assuming fairly efficient markets it seems apparent that mutual fund managers need
skill or superior information to outperform the market. Otherwise no rational investor
would put any money in an active mutual fund, since it would be a zero sum or even a
losing game.
77
Skill can be defined either as the managers' ability to generate a true
positive alpha after costs or a true positive alpha before costs. The second definition
might be justifiable since the expenses charged are not essentially linked to the
management itself.
78
But nevertheless investors only receive net returns, thus the second
definition is rather a theoretical one. Skill is derived either due to stock picking ability
or market timing ability, whereas stock picking ability may be simply linked to superior
information, provided by managers of firms the fund is invested in.
79
Based on the
assumption above, skill would be expressed in two different ways, increasing the return
of a portfolio or minimizing the "insurable risk" and so increase risk adjusted returns.
80
This is important since investors usually show some risk aversion and thus prefer less
risky investments instead of risky ones when yielding the same expected return.
81
The
71
Some ETFs have a total expense ratio of 0.15% p.a. or below. The bid-ask spread, expressed by the
Xetra liquidity measure of some ETFs is around 10 basis points when buying and selling 100,000 at
the same time. (Deutsche Börse AG (2013) p.4).
72
Carhart (1997) p.66.
73
Cf. Barras/Scaillet/Wermers (2010) p. 179 ff.; Fama/French (2010) p. 1919 f.; Kosowski, et al. (2006)
p.2551 - 2555.
74
Cf. Barras/Scaillet/Wermers (2010) p. 199.
75
Alpha is constantly negative since 1996 until the end of the study in 2006, where alpha amounts minus
one. See Barras/Scaillet/Wermers (2010) p. 199.
76
Cf. Berk/Green (2004) p. 1271.
77
Cf. Antypas, et al. (2009) p. 2; Ferson/Schadt (1996) p. 426.
78
Cf. Barras/Scaillet/Wermers (2010) p.209.
79
Cf. Bodson/Cavenaile/Sougné (2013) p. 96; Cohen/Frazzini/Malloy (2008) p. 951 f., 954.
80
Cf. Jensen, M. C. (1968) p. 389.
81
Cf. Jensen (1968) p. 389; Tversky/Kahneman (1992) p. 297f..
12
risk aversion applies as long as the probability of the outcome is not extremely unlikely,
which gives investors incentives for gambling.
82
Skill would be expressed in the
managers' ability to "predict" the return of the market or the return of individual stocks.
In addition to the "pure" skill the manager should be able to enforce her strategy within
the fund and should also be willing to invest in the stocks of whom she believes
(knows) that they will outperform the market.
83
Otherwise a managerial skill would be
useless.
Consequently, investors have some interest in knowing whether the manager respective
the management team is skilled and courageous to follow the strategy described in the
prospectus, resulting ideally in outperforming the market. In this context it is also
important that the management is aware of some behavior pattern like the disposition
effect. Mutual funds which are prone to disposition and do not implement suitable
actions might unintended change the investment style.
84
Identifying true alpha funds sounds simple but the main challenge is to differentiate
between skilled funds and those who are just lucky. The Fama and French paper "Luck
versus Skill" has received some attention
85
and documents like many others that the vast
majority of outperformance and underperformance, at least in the short run, is
determined by luck.
86
All these studies state in a more or less pessimistic manner that
mutual funds on average are unable to outperform the market after costs. These studies
do not conclude that managers in general are unable to outperform the market, but skill
is blurred away by less or even unskilled managers.
87
The most important finding is that
it seems nearly impossible to distinguish between funds run by skilled but unlucky
managers and funds run by lucky but unskilled managers, at least when using
performance patterns only.
88
Although this conclusion is not new
89
investors are further
going on and investing in active managed portfolios, preferable those performing well
in the past.
82
Cf. Kahneman/Tversky (1979) p. 265 - 268, 275.
83
Cf. Jensen (1968) p. 389 f., 395.
84
Cf. Cici (2012) p. 799.
85
See for example: Mamudi (2009)
86
Cf. Barras/Scaillet/Wermers (2010) p.179, 181, 184; Fama/French (2010) p.1915; Kosowski, et al.
(2006) p. 2553.
87
Cf. Kosowski, et al. (2006) p. 2562, 2573.
88
Cf. Fama/French (2010) p.1933.
89
Cf. Jensen (1968) p. 389 f..
13
The fraction of mutual fund managers with stock picking ability is higher in
(aggressive) growth oriented funds whereas balanced or income funds tend to show
higher fractions of managers with lower or even no stock picking ability. At the same
time managers in growth oriented funds do not show persistence in skill which might be
linked to the fact that their talent is influenced by decreasing returns to scale.
90
This
finding would, at least theoretically, result in a kind of equilibrium state where all
mutual funds are endowed with zero alphas, as predicted by Berk and Green in their
2004 published paper.
91
Looking closer at the recent published studies, three types of mutual funds and mutual
fund managers are distinguished, respective.
92
x Unskilled managers are unable to generate a return covering the expenses and the
trading costs, resulting in a true negative alpha.
x Average skilled managers are able to cover the costs they cause which results in
zero alpha.
x Skilled Managers are able to cover trading costs and expenses and achieve a
surplus resulting in a true a positive alpha.
93
The percentage of skilled managers or more precisely the percentage of mutual funds
that are able to beat the benchmark by skill is becoming smaller over the last decades.
94
In detail the share of skilled managers has declined from roughly 15% in the early
1990s to less than 1% in the mid-2000s. Additionally the share of unskilled managers
has risen from roughly 10% to 24% in the same period of time. The share of true zero
alpha funds have remained fairly constant at around 75% which reflects the long run
predictions of Berk and Green as equilibrium funds.
95
It should be noted that different
studies show different numbers in detail since they are based on different methods, time
horizons or markets. But the conclusions in these studies regarding the general
development are quite similar.
96
There are some reasons in order to explain this development. The first reason is linked
to the efficient market hypothesis and states that markets are getting more efficient over
time, whereas the managers' ability stays relatively constant. Consequently the share of
90
Cf. Barras/Scaillet/Wermers (2010) p.181,202;.Kosowski, et al. (2006) p.2555 f..
91
Cf. Berk/Green (2004) p. 1271.
92
Some studies assign the skill on the manager level other rather on the mutual fund.
93
Cf. Barras/Scaillet/Wermers (2010) p.183.
94
Cf. Barras/Scaillet/Wermers (2010) p.181; Kosowski, et al. (2006) p.2554.
95
Cf. Barras/Scaillet/Wermers (2010) p. 199.
96
Cf. Barras/Scaillet/Wermers (2010) p.199; Kosowski, et al. (2006) p.2575 f..
14
managers who are potentially able to manage outperforming funds is getting smaller.
The second explanation states that these highly skilled managers are more willing to
work for hedge funds, since mutual funds might limit their managerial abilities by
regulatory constraints. Additionally, hedge funds might offer these managers an even
higher remuneration.
97
It seems also rational to assume that the growth of the mutual
fund market attracts many unskilled managers.
98
A more intuitive reason for the mentioned development is linked to the fact that all
investors on average only can receive the return of the market (passive strategy), if
mutual funds on average would outperform the market, other market participants would
lose. With an increasing number of mutual funds and a bigger market share invested in
mutual funds, outperformance is getting less likely. Moreover, in a situation where
investing in mutual funds is a winning game for "sure" no rational investor would invest
her assets elsewhere, thus mutual funds would reflect the market, consequently an
average outperformance would be impossible. This can even be assumed when more
assets are invested in mutual funds.
In addition to that it can be argued that the proportion of skilled managers is higher than
the common literature suggests. This notion is based on the fact that fund managers are
forced to trade due to liquidity demands of their investors which erodes performance.
99
Assuming an asset turnover of 100%, liquidity induced trading erodes performance by
1.5 to 2% annually.
100
Nevertheless, from the investors' point of view it is basically
meaningless where the performance comes from.
101
3.3
Persistence in Mutual Fund Performance and Mutual Fund Ratings
Performance persistence in mutual funds is debated extensively in the literature.
102
Persistence is linked to managerial characteristics and skill, thus a superior manager
should be able to generate a persistent positive alpha. Due to the fact that ratings are
built strongly on fund performance, conclusions regarding the fund's performance can
be derived when looking at ratings. If mutual fund performance persists, mutual funds
97
The number of incentive fee mutual funds amounts around 1.5 to 3 % of total mutual fund markets in
the US. See: Elton/Gruber/Blake (2003) p. 781; Madhur (2005) p. 5.
98
Cf. Fama/French (2010) p.1941; Kosowski, et al. (2006) p.2575.
99
At this point the disposition effect might play a major role since managers are quite likely to sell
winners instead of losers, when they are facing liquidity demand. See: Cici (2012) p. 799.
100
Edelen (1999) p. 440 f..
101
A potential solution to this would be, closing the fund for new investors. But this is contradicts with a
fee structure based on the assets under management.
102
See for example: Blake/Morey (2000); Brown/Goetzmann (1995); Carhart (1993); Hendricks/Patel/
Zeckhauser (1993).
15
Excerpt out of 88 pages
- scroll top
Details
- Quote paper
- , 2016, Investment Criteria for Mutual Fund Selection, Hamburg, Bedey Media GmbH, https://www.anchor-publishing.com/document/342381
Comments