The effects of fund commonality in mutual fund families
on fund operating expenses and return correlations:

Evidence from U.S. equity mutual funds

Youngkyun Parka,*
aCollege of Business and Economics, University of Idaho, 875 Perimeter Drive MS 3161, Moscow, ID 83844-

3161, USA

Abstract

This article examines the effects of the commonality of mutual funds within a fund family, which
is measured by common stock holdings and multi-fund management, on fund operating expenses and
return correlations. For U.S. equity funds during the period of 2001–2006, we find that common stock
holdings and multi-fund management are negatively related to fund operating expenses but positively
related to the correlation of fund return residuals, which increases the correlation of fund returns.
Additionally, we find that the fund commonalities can have negative net effects on risk-adjusted
returns of a portfolio with equity funds that have different investment objectives. © 2016 Academy of
Financial Services. All rights reserved.

JEL classification: G23; G11

Keywords: Mutual fund family; Fund commonality; Fund operating expense; Fund return correlation

1. Introduction

Many people in the United States use mutual funds as investment and retirement saving
vehicles. Over 90 million individual investors owned mutual funds and held about 87% of
the total mutual fund assets in 2014 (Investment Company Institute [ICI], 2015). Mutual fund
investors often limit their transactions to one fund family to save search costs (Ciccotello,
Miles, and Walsh, 2007; Sirri and Tufano, 1998), to simplify fund management (Elton,

* Corresponding author. Tel.: �1-208-885-7154; fax: �1-208-885-5347.
E-mail address: youngpark@uidaho.edu

Financial Services Review 25 (2016) 29–49

1057-0810/16/$ – see front matter © 2016 Academy of Financial Services. All rights reserved.



Gruber, and Green, 2007), or as a result of following a fund family’s reputation (Ciccotello,
Greene, and Walsh, 2007; Gerken, Starks, and Yates, 2014). In addition, mutual fund
investors have become sensitive to a fund’s expense ratio.1 Fig. 1 shows the simple average
expense ratios, the asset-weighted average expense ratios, and the difference between the two
expense ratios of equity funds for the period of 1996 to 2014 (ICI, 2015). The figure indicates
that even if the mutual fund industry has recently provided more low-cost equity funds,
equity fund investors have kept their sensitivity to fund expense ratios. Because long-term
returns are significantly affected by fund expenses (e.g., Carhart, 1997; Gil-Bazo and
Ruiz-Verdu, 2009; Haslem, Baker, and Smith, 2008), it is natural that fund investors focus
on a fund’s expense ratio. However, if a lower expense ratio results from increased fund
commonality within a family, seeking only low-cost funds from one family may result in
increasing an investor’s portfolio risk.

In this study, we measure the commonality of funds by how much a fund holds the same
stocks for other funds in the same family and whether a fund is managed by multi-fund
managers who simultaneously manage multiple funds in the family. Common stock holdings
and multi-fund management have been recognized in the recent literature that emphasizes
mutual funds as members of a fund family rather than as stand-alone entities (e.g., Agarwal,
Ma, and Mullally, 2015; Choi, Kahraman, and Mukherjee, 2013; Elton et al., 2007; Yadav,
2010). Common stock holdings and multi-fund management may reduce fund operating
expenses (excluding 12b-1 fee), which are associated with portfolio management and
administrative services. The two fund commonalities, however, may also enhance return
correlations between funds and thereby increase an investor’s portfolio risk.

This study examines the effects of common stock holdings and multi-fund management of
funds on fund operating expenses and return correlations. Common stock holdings and
multi-fund management are assessed only for pairs of funds that have different investment
objectives, by excluding pairs of funds that have the same investment objective. Because of

Fig. 1. Expense ratios in equity mutual funds. Source: Investment Company Institute Fact Book.

30 Y. Park / Financial Services Review 25 (2016) 29–49



this exclusion, the fund commonalities are conservatively evaluated. We analyze 154 ac-
tively managed U.S. equity funds from 46 fund families for the period of 2001 to 2006,
reconciling key data items (such as a fund’s stock holdings and managers) in the Center for
Research in Security Prices (CRSP), the Thomson Financial Mutual Fund Holdings, and the
Morningstar databases. We find that common stock holdings and multi-fund management are
negatively related to fund operating expenses but positively related to the correlation of fund
return residuals, which increases the correlation of fund returns. For a portfolio that is
constructed with funds in the same family, the findings indicate that an increase in a
portfolio’s risk-adjusted return net of expenses (such as a Sharpe ratio using expense-
adjusted returns) because of decreased fund operating expenses can be wiped out by an
increase in the portfolio risk because of increased return correlations. Because of the opposite
effects on a portfolio’s risk-adjusted return, we additionally investigate the effects of the fund
commonalities on risk-adjusted returns of a portfolio that consists of equity funds with
different investment objectives and find that the fund commonalities can have negative net
effects on risk-adjusted returns of the portfolio.

Our findings contribute to the financial services literature in three ways. First, we expand
previous research on mutual fund expenses by showing that a fund’s operating expense can
be reduced by its common stock holdings and managers who simultaneously manage other
funds in the family. Second, we extend extant research on interfamily dynamics by showing
that the fund commonalities within a family can deteriorate an investor’s portfolio diversi-
fication by increasing the correlation of fund returns. Third, our findings provide practical
implications to individual mutual fund investors. When individual investors construct a
portfolio with low-cost equity funds within a family, they should be aware of an investment
risk that fund commonalities that lower fund operating expenses can increase fund return
correlations and thereby can reduce the portfolio’s risk-adjusted return.

The remainder of the article is organized as follows. Section 2 addresses hypotheses on
relations between fund commonalities and operating expense and between fund commonal-
ities and return correlations. Section 3 describes the data and the sample. Section 4 presents
the methodology. Section 5 presents empirical findings. Section 6 concludes with a summary
of findings and implications for individual mutual fund investors.

2. Fund commonalities, operating expense, and return correlation

Funds in a family may have common features even if they have different investment
objectives (Elton et al., 2007). For example, different funds may hold the same stocks as a
result of fund managers’ common view on individual companies from a common security
selection process at a family level (Elton et al., 2007) or of fund managers’ efforts to reduce
the costs of monitoring the performance of the stocks held in their portfolios (Shawky and
Smith, 2005). In addition, different funds may be managed simultaneously by a manager(s)
in particular when the manager(s) exhibits superior past performance (Agarwal, Ma, and
Mullally, 2015). These common features would increase the commonality of funds in the
family. The increased fund commonality, in turn, may influence the characteristics of funds,
such as fund operating expenses and the correlation of fund returns.

31Y. Park / Financial Services Review 25 (2016) 29–49



2.1. Fund commonalities and operating expense

Mutual funds incur ongoing charges for managing fund assets. Fund expenses paid out of
fund assets consist of three broad components: management fees, other expenses including
administrative fees, and 12b-1 fees (Collins, 2003; Khorana & Servaes, 2007; Latzko, 1999).
The first and typically largest is the management fee paid to a fund manager, which
compensates a fund manager for expenses incurred in providing services, including asset
allocation and security selection. The second is other expenses including administrative fees,
which mainly relate to recordkeeping and transactions services to shareholders.2 The third is
the 12b-1 fees spent on advertising, marketing, and distribution services or commissions to
brokers. A fund’s expense ratio may matter to fund families that want to increase assets under
management because mutual fund investors increasingly purchase funds with low expenses
(Khorana & Servaes, 2007). For example, Collins (2007) documents that 90% of net new
cash flow to stock funds went to those funds whose expense ratios were below the average
expense ratio of stock funds offered in the marketplace from 1997 to 2006. Thus, fund
families or managers may increase their efforts to reduce a fund’s expense ratio.

We focus on fund operating expenses, defined by fund expenses net of 12b-1 fees, because
fund operating expenses account for the largest share of a fund’s expense ratio (Collins,
2003) and because 12b-1 fees are mainly related to distribution fees for the services provided
by brokers rather than the services by fund families (ICI, 2015). Fund operating expenses,
consisting of management fees and other expenses including administrative fees (Collins,
2003), may be reduced by an increase in common stock holdings or multi-fund management.
First, common stock holdings may be able to lower a fund’s management fee. Fund managers
in the same family are likely to use the same research analysis produced by either internal
analysts or external research firms (Elton et al., 2007). The sharing of the research analysis
within a family may increase common stock holdings across different funds, but it may also
reduce the costs incurred in the security selection process, which are related to a fund’s
management fees. Thus, we expect that a fund’s common stock holdings are negatively
related to its management fee. Second, the management of multiple funds (i.e., the multi-
fund management) may be able to lower a fund’s other expenses including administrative
fees. Fund families often assign multiple funds to the same portfolio manager. Agarwal et al.
(2015) find that, for the period of 1980 to 2012, 48% of mutual fund managers managed
multiple funds simultaneously. This multi-fund management may enable a fund manager to
reduce administrative fees by efficiently allocating her resources among the funds she
manages. In this respect, we expect that multi-fund management is negatively related to a
fund’s other expenses. Thus, for a fund’s management fee and other expenses, we hypoth-
esize the following:

Hypothesis 1-a: Common stock holdings are negatively related to a fund’s management
fee.

Hypothesis 1-b: Multi-fund management is negatively related to a fund’s other ex-
penses.

32 Y. Park / Financial Services Review 25 (2016) 29–49



2.2. Fund commonalities and return correlation

Except for common exposure to market and systematic factors, return correlations of
funds in a family may be increased by nonmarket or nonsystematic factors that are shared
between funds, such as common stock holdings and the management of multiple funds. To
measure the correlation of fund returns caused by nonmarket or nonsystematic factors, we
use the correlation of return residuals. Following Elton et al. (2007), we decompose the
correlation of fund returns into two parts: (1) the return correlation caused by market and
systematic factors, such as size, book-to-market, and momentum factors, and (2) the corre-
lation of return residuals, which are not captured by market and systematic factors. Then we
focus on the correlation of return residuals and examine whether return residual correlations
are related to common stock holdings and multi-fund management. First, common stock
holdings of funds may enhance the commonality of funds in a family. Elton et al. (2007) find
that return correlations of funds are higher within families than between families and that
about 60% of the increased correlation is because of common stock holdings. Thus, we
expect that common stock holdings are positively related to the correlation of return
residuals, which increases the correlation of fund returns. Second, multi-fund management
may also increase the commonality of funds. Funds within a family tend to pursue a similar
investment strategy (e.g., low- or high-risk strategies) across different investment objectives
(Elton et al., 2007). Similarly, funds managed by a multi-fund manager may adopt a similar
investment strategy even if they have different investment objectives. A multi-fund manager
may pursue her investment strategy for her different funds by selecting stocks with similar
risk characteristics (but not necessarily the same stocks). Thus, funds managed by a multi-
fund manager would be more likely to have similar risk characteristics and, as a result, to
have a higher correlation of return residuals than those managed by different fund managers.
In this respect, for the correlation of return residuals, we hypothesize the following:

Hypothesis 2-a: Common stock holdings are positively related to the correlation of
return residuals.

Hypothesis 2-b: Multi-fund management is positively related to the correlation of return
residuals.

3. Data and sample

The data come from the CRSP, the Thomson Financial Mutual Fund Holdings, and the
Morningstar databases for the period of 2001 and 2006. First, we merge the CRSP and the
Thomson Financial Mutual Fund Holdings and collect the information on funds’ stock
holdings.3 To identify fund investment objectives, we use the Investment Company Data,
Inc.’s (ICDI) classifications that are matched with the Strategic Insight (SI) classifications
because the broad ICDI categories enable us to develop a parsimonious model (Table 1). Our
sample includes only actively managed U.S. equity funds that are classified as Aggressive
Growth (AG), Growth and Income (GI), and Long-Term Growth (LG). Second, we use the
Morningstar to collect the information on fund managers and management fees, instead of

33Y. Park / Financial Services Review 25 (2016) 29–49



the CRSP. The CRSP provides less reliable information on fund managers (Massa, Reuter,
and Zitzewitz, 2010) and reports management fees that include waivers and reimbursements,
which can lead to negative management fees. We merge the Morningstar and the merged
dataset above, using funds’ tickers, CUSIPs, net asset values, and returns that prior studies
use to merge the Morningstar and the CRSP databases (e.g., Berk and van Binsbergen, 2015;
Pástor, Stambaugh, and Taylor, 2015).4 Third, we include only funds that have distinct
portfolios. Mutual funds offer multiple shares, but they usually have the same portfolio
managers, the same pool of securities, and the same returns before expenses and loads (Zhao,
2004). The Morningstar designates the oldest share class as a fund that has the distinct
portfolio.5 Relying on the information provided by the Morningstar, we include only funds
that are identified as ones that have distinct portfolios. Fourth, we include only funds that
existed for the period of 2001 to 2006. Last, we exclude 14 observations where a fund’s
expense ratio is less than its management fee.6 After all of these data-cleaning steps, we have
a sample of 154 U.S. equity funds from 46 fund families.

4. Methodology

4.1. Fund commonalities: Common stock holding and multi-fund management

As proxies to capture the commonality of funds within a family, we use common stock
holdings and multi-fund management (or the management of multiple funds). First, we define
common stock holdings of funds within a family as follows:

MatchHijt � �
k

min�wkit, wkjt� (1)

where wkit and wkjt are the weights of stock k held by funds i and j at year-end t, respectively,
and the sum is taken over all the stocks in both funds. Funds i and j have different investment
objectives. A fund’s stock holding is determined by using the information reported on the
date closest to the end of the year t. MatchHijt ranges from 0 to 1. It is equal to 0 when the
two funds do not have any common stock holdings in their portfolios at year-end t; it is equal
to 1 when they hold exactly the same portfolios at year-end t. In addition, to indicate fund

Table 1 Investment objective classifications for equity mutual funds

ICDI Classification SI Classification

Aggressive Growth (AG) Aggressive Growth (AGG) and Small Company Growth (SCG)
Growth and Income (GI) Growth and Income (GRI) and Income and Growth (ING)
Long-Term Growth (LG) Growth (GRO) and Growth Mid-Cap (GMC)

Notes: This article uses the Investment Company Data, Inc.’s (ICDI) classifications that are matched with the
Strategic Insight (SI) classifications. The matching is conducted on the basis of a two-way table of ICDI and SI
classifications for the period of January 2001 to June 2003. The ICDI classifications are not available from the
CRSP database after July 2003.

34 Y. Park / Financial Services Review 25 (2016) 29–49



i’s common stock holdings for all other investment objective funds in the same family, we
modify Eq. (1), which is defined for a pair of funds, as follows:

ComHit �
�j�1

L MatchHijt

L
(2)

where MatchHijt is the common stock holdings of funds i and j at year-end t and the sum is
taken over L, which indicates the total number of funds in all other investment objectives that
are different from fund i’s investment objective. Thus, ComHit indicates the average weight
of common stock holdings in a portfolio of fund i at year-end t.

Second, we define multi-fund management as follows:

MatchMijt � �1, if there is at least one multi-fund manager between funds i and j
0, otherwise (3)

where funds i and j have different investment objectives. To determine a fund’s managers,
we use a measure of whether or not they manage the fund at the end of the year. MatchMijt

takes a value of 1 if there is at least one multi-fund manager between funds i and j at year-end
t and 0 if there is no multi-fund manager between the funds at year-end t. In addition, to
indicate whether or not fund i is managed by fund managers who simultaneously manage
funds in other investment objectives, we modify MatchMijt, which is defined for a pair of
funds, as follows:

ComMit � �1, if �
j�1

L MatchMijt � 0

0, if �
j�1

L MatchMijt � 0
(4)

where MatchMijt indicates the multi-fund management that is defined for a pair of funds i and
j at year-end t and the sum is taken over L, which indicates the total number of funds in all
other investment objectives that are different from fund i’s investment objective. ComMit

takes a value of 1 if fund i is managed by at least one multi-fund manager at year-end t and
0 if fund i is managed by fund managers who do not manage other investment objective
funds at year-end t.

4.2. Model specification

4.2.1. Fund commonalities and operating expenses
To test the hypotheses on relationships between fund commonalities and fund operating

expenses (H1-a and H1-b), we estimate the following model:

Expenseit � �0 � �1ComHit � �2ComMit � �3Zt � vit (5)

where the dependent variable Expenseit indicates fund i’s operating expense, management
fee, or other expenses in year t. First, we use as the dependent variable a fund’s operating
expense ratio that is computed by subtracting a 12b-1 fee from a fund’s expense ratio

35Y. Park / Financial Services Review 25 (2016) 29–49



(Collins, 2003). To calculate a fund’s operating expense ratio, we use fund expense ratios
from the CRSP database and 12b-1 fees from the Morningstar database. This enables us to
include more valid observations in the sample. For our sample, the CRSP provides fewer
missing values for a fund’s expense ratio than the Morningstar, while the Morningstar
provides fewer missing values for a fund’s 12b-1 fee than the CRSP. For example, if we use
the data on 12b-1 fees from the CRSP, we would lose about 47% of the sample. Second, we
use a fund’s management fee as the dependent variable. For management fees, we use the
Morningstar database because the management fee provided by the CRSP can be offset by
fee waivers and reimbursements, which can lead to negative management fees. These
negative management fees might distort a relationship between fund commonalties and
management fees. Third, we use a fund’s other expense ratio as the dependent variable. A
fund’s other expenses include fees for administrative and business services, such as legal,
accounting, and administrative services (Collins, 2003; Mahoney, 2004). A fund’s other
expense ratio is computed by subtracting a management fee from a fund’s operating expense
ratio.

Eq. (5) also includes a set of control variables (Z) at fund and family levels. Following
prior studies, we first include a fund’s annual turnover ratio that represents trading activity.
A fund with a higher turnover ratio is likely to have a higher expense ratio because of active
trading (Malhotra and McLeod, 1997). The higher expense ratio caused by the active trading,
however, may result from an increase in a fund’s other expenses because they include
registration fees on the fund shares sold each year and custodial fees that cover the costs of
settling trades (Latzko, 2003). Second, we control for a fund’s load, using a dummy variable,
which indicates whether or not a fund has a sales load (either a front-end or back-end load).
The variable takes a value of 1 if a fund has any sales loads and 0 otherwise. Third, we
control for a fund’s investment objective, using two dummy variables with a base category
of AG. Fourth, we control for fund size and family size, using the natural logarithm of one
plus a fund’s year-end total net assets (in billions) and the natural logarithm of one plus the
year-end total net assets (in billions) managed by a family, respectively. Last, we include
family fixed effects, which control for unobserved time-invariant family characteristics such
as different strategic behaviors at a family level (e.g., Clare, O’Sullivan, and Sherman, 2014;
Gaspar, Massa, and Matos, 2006), and year fixed effects.

4.2.2. Fund commonalities and return residual correlation
To test the hypotheses on relationships between fund commonalities and the correlation of

return residuals (H2-a and H2-b), we first decompose the correlation of returns into two parts,
following Elton et al., (2007):

Corr�Ri, Rj� � Corr�Fi, Fj� � Corr�ei, ej� (6)

where Corr(Ri, Rj) is the correlation of returns of funds i and j; Corr(Fi, Fj) is the correlation
of fund returns caused by systematic movements; and Corr(ei, ej) is the correlation of return
residuals. We calculate Corr(ei, ej) with residuals that are obtained from estimating the
Carhart four-factor model (Carhart, 1997):

36 Y. Park / Financial Services Review 25 (2016) 29–49



Rit � RFt � �i � �iEM�EMt� � �iSMB�SMBt� � �iHML�HMLt� � �iMOM�MOMt�

� �it (7)

where Rit is the return of fund i in month t; RFt is the one-month Treasury-bill rate in month
t; EMt is the monthly excess return on the market, which is calculated by the monthly return
of the CRSP value-weighted index less the one-month Treasury-bill rate in month t; SMBt,
HMLt, and MOMt denote the size-factor, the book-to-market factor, and the momentum
factor in month t, respectively; and �it is a random error.

Second, we regress return residual correlations (Corr(ei, ej)) on the average common stock
holding and the multi-fund management between funds as follows:

Corr�ei, ej� � �0 � �1AMatchHij � �2AMatchMij � �3W � uij (8)

where AMatchHij is the average common stock holding between funds i and j for the period
of 2001 to 2006; AMatchMij is a dummy variable to indicate whether there exists any
multi-fund management between funds i and j for the period; and W indicates a set of control
variables at fund and family levels.7 We control for funds’ investment objectives, using two
dummy variables with a base category of a combination of GI and LG. In addition, at a
family level, we include (1) the natural logarithm of the total number of funds having distinct
portfolios that are offered in AG, GI, and LG investment objectives by a family over the
period and (2) family size defined by the natural logarithm of one plus the average total net
assets (in billions) managed by a family over the period. Last, we include family fixed
effects, which control for unobserved time-invariant family characteristics.

5. Results

5.1. Summary statistics

Table 2 presents summary statistics of the sample for the period of 2001 to 2006. The
average expense ratio is 1.17%; the average 12b-1 fee is 0.17%; and the average operating
expense ratio is 1.00%. A fund’s management fee and other expenses, on average, account
for 71% and 29%, respectively, of a fund’s operating expense ratio. Funds hold, on average,
17% of their portfolios with the same stocks across other funds in different investment
objectives. About 39% of the funds in the sample are managed by multi-fund managers over
the period. Most funds in the sample have sales loads. This is because we use funds that have
distinct portfolios defined by the Morningstar and they are likely to be Class A shares that
generally have a front-end load.

Table 2 also provides summary statistics by fund investment objectives: AG, GI, and LG.
Funds seeking AG, on average, have a higher expense ratio, a higher operating expense ratio,
and a higher management fee than those seeking GI or LG. However, 12b-1 fees and other
expenses are similar across the fund investment objectives. In addition, funds seeking AG
have significantly less common stock holding than those seeking GI or LG, but AG funds are
more likely to be managed by multi-fund managers than are LG funds.8 Last, the average
total net asset of AG funds is less than that of GI or LG funds.

37Y. Park / Financial Services Review 25 (2016) 29–49



Table 2 Sample summary statistics

Observation Mean Standard
deviation

Min Max

Full sample (154 funds) Expense ratio (%) 903 1.17 0.34 0.40 2.31
12b-1 fee (%) 913 0.17 0.22 0.00 1.00
Operating expense (%) 901 1.00 0.29 0.15 1.90
Management fee (%) 913 0.71 0.22 0.24 1.50
Other expenses (%) 901 0.29 0.21 0.00 1.13
Fraction of common stock holding

in a fund portfolio
915 0.17 0.16 0.00 0.96

Multi-fund management (yes � 1) 915 0.39 0.49 0.00 1.00
Turnover ratio 897 0.77 0.73 0.03 7.76
Load fund (yes � 1) 915 0.98 0.13 0.00 1.00
Fund size ($ billion) 915 3.88 10.61 0.001 120.84
Family size ($ billion) 276 52.11 158.43 0.17 1,149.30

Aggressive Growth
(43 funds)

Monthly gross return (%) 3,096 0.88 5.50 �34.28 25.95
Expense ratio (%) 254 1.33 0.31 0.79 2.31
12b-1 fee (%) 255 0.19 0.27 0.00 1.00
Operating expense (%) 253 1.13 0.22 0.36 1.80
Management fee (%) 255 0.85 0.18 0.46 1.50
Other expenses (%) 253 0.29 0.22 0.00 1.03
Fraction of common stock holding

in a fund portfolio
256 0.13 0.16 0.00 0.96

Multi-fund management (yes � 1) 256 0.45 0.50 0.00 1.00
Turnover ratio 253 0.76 0.63 0.04 3.58
Load fund (yes � 1) 256 0.98 0.15 0.00 1.00
Fund size ($ billion) 256 0.89 1.25 0.004 9.51

Growth and Income
(43 funds)

Monthly gross return (%) 3,079 0.61 3.68 �14.39 18.46
Expense ratio (%) 246 1.04 0.33 0.55 2.27
12b-1 fee (%) 252 0.16 0.19 0.00 1.00
Operating expense (%) 245 0.88 0.30 0.32 1.90
Management fee (%) 252 0.61 0.21 0.24 1.00
Other expenses (%) 245 0.28 0.19 0.00 1.13
Fraction of common stock holding

in a fund portfolio
253 0.17 0.15 0.00 0.81

Multi-fund management (yes � 1) 253 0.38 0.49 0.00 1.00
Turnover ratio 243 0.54 0.39 0.03 2.17
Load fund (yes � 1) 253 1.00 0.06 0.00 1.00
Fund size ($ billion) 253 5.84 12.71 0.014 79.33

Long-Term Growth
(68 funds)

Monthly gross return (%) 4,877 0.46 4.99 �24.85 29.00
Expense ratio (%) 403 1.15 0.33 0.40 2.09
12b-1 fee (%) 406 0.16 0.20 0.00 1.00
Operating expense (%) 403 0.99 0.28 0.15 1.79
Management fee (%) 406 0.69 0.20 0.31 1.00
Other expenses (%) 403 0.30 0.21 0.00 1.13
Fraction of common stock holding

in a fund portfolio
406 0.20 0.15 0.00 0.81

Multi-fund management (yes � 1) 406 0.35 0.48 0.00 1.00
Turnover ratio 401 0.92 0.89 0.03 7.76
Load fund (yes � 1) 406 0.98 0.15 0.00 1.00
Fund size ($ billion) 406 4.53 12.00 0.001 120.84

Notes: The sample consists of 154 actively managed U.S. equity funds from 46 fund families for the period of
2001 and 2006. Summary statistics are obtained from fund-year observations except for family size and monthly
gross return. Summary statistics of family size are obtained from family-year observations and summary statistics
of monthly gross return from fund-month observations.

38 Y. Park / Financial Services Review 25 (2016) 29–49



5.2. Regression results

5.2.1. Effects of fund commonalities on fund operating expenses
Regression results in Table 3 present that fund operating expenses are significantly

negatively related to two fund commonalities: common stock holdings and multi-fund
management.9 First, common stock holdings are significantly negatively related to a fund’s
operating expense ratio. A 10% increase in common stock holding, for example, reduces a
fund’s operating expense ratio by 1.3–1.4%. To further examine how common stock
holdings are related to the components of fund operating expenses, we estimate Eq. (5) for
funds’ management fees (Columns 3 and 4) and other expenses (Columns 5 and 6).
Regression results show that common stock holdings are significantly negatively related to
a fund’s management fee, supporting hypothesis H1-a. However, common stock holdings are
not significantly related to a fund’s other expenses. Thus, the results indicate that a decrease
in a fund’s operating expense ratio caused by an increase in common stock holdings occurs
through a decrease in its management fee, not through its other expenses. Second, multi-fund
management is significantly negatively related to a fund’s operating expense ratio (Columns
1 and 2). For example, a fund’s operating expense ratio is reduced by 4.2–5.1% when a fund

Table 3 Effects of fund commonalities on fund expenses

Operating expense Management fee Other expenses

(1) (2) (3) (4) (5) (6)

Common stock holding
(ComH)

�0.138** �0.134* �0.158** �0.076* 0.020 �0.047
(0.047) (0.063) (0.032) (0.035) (0.042) (0.060)

Multi-fund management
(ComM, yes � 1)

�0.042** �0.051* 0.014 0.018 �0.056** �0.069**
(0.014) (0.025) (0.012) (0.016) (0.013) (0.024)

Turnover ratio 0.022 0.019 �0.018** �0.012 0.041** 0.033*
(0.014) (0.015) (0.007) (0.007) (0.012) (0.015)

Load fund (yes � 1) 0.035 �0.070 0.032 �0.014 0.001 �0.058
(0.087) (0.049) (0.033) (0.014) (0.065) (0.042)

Growth and Income (GI) �0.157** �0.204** �0.180** �0.165** 0.023 �0.042*
(0.019) (0.018) (0.015) (0.010) (0.018) (0.016)

Long-Term Growth (LG) �0.060** �0.068** �0.087** �0.079** 0.025 0.007
(0.018) (0.014) (0.013) (0.008) (0.016) (0.014)

Fund size �0.119** �0.105** �0.043** �0.028** �0.076** �0.076**
(0.010) (0.012) (0.006) (0.006) (0.009) (0.011)

Family size �0.026** �0.083** �0.038** 0.011 0.012* �0.091**
(0.005) (0.017) (0.004) (0.016) (0.005) (0.021)

Year fixed effects Yes Yes Yes Yes Yes Yes
Family fixed effects No Yes No Yes No Yes
Observations 895 895 895 895 895 895
Adjusted R2 0.426 0.641 0.468 0.807 0.148 0.392

Notes: The table reports the results of the regressions of fund expenses on fund commonalities. The dependent
variables are operating expense (Columns 1 and 2), management fee (Columns 3 and 4), and other expense
(Columns 5 and 6) of mutual funds. A fund’s operating expense is divided into a fund’s management fee and other
expenses. Fund size and family size indicate the natural logarithm of one plus a fund’s year-end total net assets
(in billions) and the natural logarithm of one plus the year-end total net assets (in billions) managed by a family,
respectively. Robust standard errors are in parentheses.

*p � 0.05, **p � 0.01.

39Y. Park / Financial Services Review 25 (2016) 29–49



is managed by multi-fund managers who simultaneously manage other funds that have
different investment objectives. In addition, regression results on management fees (Columns
3 and 4) and other expenses (Columns 5 and 6) show that multi-fund management is not
significantly related to a fund’s management fee but significantly negatively related to a
fund’s other expenses, supporting hypothesis H1-b. Thus, the results indicate that a decrease
in a fund’s operating expense ratio caused by multi-fund management occurs through a
decrease in its other expenses, not through its management fee.

For the effects of the control variables on fund operating expenses, Table 3 shows similar
results reported in prior studies. First, a fund’s turnover ratio, indicating trading activity, is
positively related to a fund’s other expenses (Columns 5 and 6), which include administrative
fees such as registration fees on the fund shares sold and the costs of settling trades (Latzko,
2003). Second, a fund’s load is not significantly related to a fund’s operating expenses. This
result is consistent with Khorana and Servaes (2007) claiming that a positive relationship
between loads and fund expenses is driven by 12b-1 fees rather than by fund operating
expenses. Third, funds seeking GI or LG are likely to have a lower operating expense ratio
than those seeking AG. The lower operating expense ratios of GI and LG funds result mainly
from their lower management fees. Last, fund size and family size are significantly nega-
tively related to fund operating expenses, the negative relation that indicate economies of
scale at fund and family levels, respectively (e.g., Dowen and Mann, 2004; Latzko, 1999).
In particular, consistent with Elton, Gruber, and Blake (2012), we find that a negative
relationship of other expenses and fund size is stronger than a negative relationship of
management fee and fund size.

In summary, the regression results in Table 3 indicate that a fund’s operating expense ratio
can be reduced by an increase in common stock holdings or by multi-fund management.10

The influences of the two commonalities on a fund’s operating expense ratio, however, occur
through different channels: common stock holdings reduce a fund’s management fee while
multi-fund management lowers a fund’s other expenses.

5.2.2. Effects of fund commonalities on return residual correlations
Next, we examine whether common stock holdings and multi-fund management affect the

correlation of fund returns within families. For doing this, we focus on the correlation of
return residuals, which is a component of the correlation of fund returns (Eq. 6).

Panel A of Table 4 presents summary statistics for pairs of funds that have different
investment objectives, such as AG-GI, AG-LG, and GI-LG. In particular, GI-LG funds have,
on average, a higher return correlation but a lower return residual correlation than the other
pairs of funds. The higher return correlations between GI and LG funds may result from their
similar investment objectives (GI vs. LG), while the lower return residual correlations may
be caused by a lower level of nonsystematic commonality between the funds.

Panel B of Table 4 presents regression results of return residual correlations, which are
obtained after estimating the Carhart four-factor model (Eq. 7). First, results in Columns (1)
and (2) show that average common stock holding (AMatchH) and multi-fund management
(AMatchM) are significantly positively related to the correlation of return residuals, support-
ing hypotheses H2-a and H2-b, respectively. For example, a 10% increase in common stock

40 Y. Park / Financial Services Review 25 (2016) 29–49



Table 4 Effects of fund commonalities on return residual correlation
Panel A: Summary statistics of pairs of funds within families

Observation Mean Standard deviation Min Max

Return correlation: 256 0.82 0.18 �0.15 0.98
between AG and GI funds (AG-GI) 38 0.76 0.23 �0.15 0.91
between AG and LG funds (AG-LG) 100 0.80 0.16 �0.03 0.98
between GI and LG funds (GI-LG) 118 0.84 0.17 �0.07 0.98

Return residual correlation: 256 0.16 0.25 �0.66 0.87
between AG and GI funds (AG-GI) 38 0.17 0.25 �0.31 0.69
between AG and LG funds (AG-LG) 100 0.21 0.24 �0.39 0.87
between GI and LG funds (GI-LG) 118 0.11 0.24 �0.66 0.58

Average common stock holding 256 0.18 0.16 0.00 0.79
Multi-fund management for the period (yes � 1) 256 0.17 0.37 0.00 1.00
Family size ($ billion) 46 52.11 153.54 0.33 872.43
Total number of funds offered by a family (TNFO) 46 3.83 3.95 2.00 27.00

Notes: AG, GI, and LG indicate fund investment objectives of Aggressive Growth, Growth and Income, and
Long-Term Growth, respectively. Average common stock holding indicates the average of the common stock
holdings between funds for the period of 2001 to 2006. TNFO indicates the total number of funds having distinct
portfolios that are offered in AG, GI, and LG investment objectives by a family over the period.

Panel B: Regression results of return residual correlation

(1) (2) (3) (4) (5) (6)

Average common stock holding
(AMatchH)

0.695** 0.767** 0.778** 0.789** 0.706** 0.775**
(0.090) (0.118) (0.114) (0.141) (0.091) (0.117)

Multi-fund management for the
period (AMatchM, yes � 1)

0.191** 0.195** 0.249** 0.223**
(0.038) (0.057) (0.054) (0.086)

AMatchH � AMatchM �0.278 �0.100
(0.163) (0.222)

AMatchM1 [A] 0.164** 0.133*
(0.044) (0.060)

AMatchM2 [B] 0.281** 0.395**
(0.042) (0.089)

Fund objectives: AG-GI 0.153** 0.136** 0.159** 0.138** 0.154** 0.130*
(0.044) (0.052) (0.045) (0.053) (0.044) (0.051)

Fund objectives: AG-LG 0.145** 0.111** 0.157** 0.114** 0.149** 0.121**
(0.028) (0.034) (0.030) (0.036) (0.028) (0.035)

log(TNFO) 0.078* 0.077* 0.065*
(0.032) (0.033) (0.033)

Family size �0.050** �0.050** �0.044**
(0.014) (0.014) (0.014)

Family fixed effects No Yes No Yes No Yes
Observations 256 256 256 256 256 256
Adjusted R2 0.387 0.451 0.390 0.449 0.391 0.466
p-value for F-test: [A] � [B] 0.025* 0.007**

Notes: AG, GI, and LG indicate fund investment objectives of Aggressive Growth, Growth and Income, and
Long-Term Growth, respectively. TNFO indicates the total number of funds having distinct portfolios that are
offered in AG, GI, and LG investment objectives by a family over the period of 2001 to 2006. Family size
indicates the natural logarithm of one plus the average total net assets (in billions) managed by a family for the
period. Observations indicate the number of different pairs of funds. Robust standard errors are in parentheses.

*p � 0.05, **p � 0.01.

41Y. Park / Financial Services Review 25 (2016) 29–49



holdings increases the correlation of return residuals by 7%, while the multi-fund manage-
ment between funds increases the correlation of return residuals by 19%.

Second, we include an interaction term of average common stock holdings (AMatchH) and
multi-fund management (AMatchM) in a model to see if they interact with each other on
return residual correlations (Columns 3 and 4). The coefficient of the interaction term
(MatchH � MatchM), however, is not statistically significant while the main effects of the
two commonalities remain significant. The results, thus, indicate that common stock holdings
and multi-fund management independently influence the correlation of return residuals.

Third, we further examine whether or not different multi-fund management structures are
related to return residual correlations. We divide multi-fund management into two categories
(Columns 5 and 6). One is that a pair of funds is managed by at least one multi-fund manager,
but not managed by exactly the same managers (MatchM1). The other is that a pair of funds
is managed by exactly the same fund manager(s) (MatchM2). Because the management by
exactly the same managers is more likely to increase similarity between funds, we expect that
the coefficient of MatchM2 is higher than that of MatchM1. MatchM1 and MatchM2 are
dummy variables that have the same base category, which is that a pair of funds is managed
by two completely different groups of fund managers. Results in Columns (5) and (6) show
that both variables, MatchM1 and MatchM2, are significantly positively related to return
residual correlations and that the coefficient of MatchM2 is significantly greater than that of
MatchM1 (p-values for F-tests � 0.05), as expected.

In summary, the regression results in Table 4 show that common stock holdings and
multi-fund management are positively related to the correlation of return residuals. The
results indicate that, when an investor constructs a portfolio with equity funds that have
different investment objectives within a family, the fund commonalities can deteriorate the
investor’s portfolio diversification by increasing the correlation of return residuals, which is
a component of the correlation of fund returns.

5.3. Fund commonalities and risk-adjusted returns

Regression results presented in Tables 3 and 4, taken together, indicate that common stock
holdings and multi-fund management decrease funds’ operating expense ratios, but increase
their return correlations. These results imply that an increase in a portfolio’s risk-adjusted
return net of expenses caused by a decrease in fund operating expenses can be wiped out by
an increase in the portfolio risk resulting from increased return correlations. In this section,
we examine net effects of the fund commonalities on a portfolio’s Sharpe ratio net of
expenses.

For the examination, we employ a hypothetical investor who constructs a diversified
portfolio with equity funds that have AG, GI, and LG investment objectives, as suggested by
Moreno and Rodriguez (2013). Suppose that a portfolio consists of two different investment
objective funds with equal weights and that the two funds are not managed by multi-fund
managers. For each portfolio, we use a set of assumptions presented in Panel A of Table 5.
First, we compute a portfolio’s annual gross return (before expenses) and expense ratio,
using the average monthly gross returns and the average expense ratios presented by fund
investment objectives in Table 2. Second, we calculate a portfolio’s risk, using annualized

42 Y. Park / Financial Services Review 25 (2016) 29–49



standard deviations of monthly gross returns in Table 2 and the average return correlations
in Panel A of Table 4. Third, we use a risk-free rate. Last, we calculate a portfolio’s Sharpe
ratio net of expenses, which is defined by Sp � �Rp � ERp � RF�/�p where Rp is a
portfolio’s annual gross return; ERp is a portfolio’s weighted expense ratio with equal
weights; RF is a risk-free rate of 2.57%, which is the average rate of one month Treasury bills
for 2001–2006; and �p is a portfolio’s risk.

Panels B and C present changes in a portfolio’s expense ratio, risk, and Sharpe ratio
because of an increase in fund commonality such as a 10% increase in common stock

Table 5 Increased fund commonalities and risk-adjusted returns
Panel A: Annual gross returns, expense ratios, risks, and Sharpe ratios of portfolios constructed with different
investment objective funds

Portfolio Annual gross
return (Rp)

Portfolio expense
ratio (ERp)

Portfolio risk (�p) Portfolio Sharpe ratio (Sp)

AG-GI 9.33% 1.19% 14.96% 0.372
AG-LG 8.38% 1.24% 17.25% 0.265
GI-LG 6.65% 1.10% 14.44% 0.207

Notes: The panel presents a set of assumptions on three different portfolios of AG-GI, AG-LG, and GI-LG
funds. AG, GI, and LG indicate fund investment objectives of Aggressive Growth, Growth and Income, and
Long-Term Growth, respectively. Each portfolio consists of two different investment objective funds with equal
weights.

Panel B: Common stock holding increased by 10%

Portfolio: AG-GI AG-LG GI-LG

(1) (2) (1) (2) (1) (2)

�Portfolio expense ratio �1.14% �1.14% �1.15% �1.15% �1.14% �1.14%
�Portfolio risk 0.36% 0.45% 0.22%
�Sharpe ratio 0.24% �0.12% 0.31% �0.14% 0.42% 0.20%
Additional gross return (basis points) 2.0 2.1 0.7

Notes: The panel reports changes in a portfolio’s expense ratio, risk, and Sharpe ratio due to a 10% increase
in common stock holding between funds in a portfolio. Additional gross return indicates how much additional
gross return is needed to have the same increase in a Sharpe ratio (reported in Column 1 for each portfolio) caused
by a decrease in a portfolio’s expense ratio.

Panel C: Multi-fund management by exactly the same fund managers

Portfolio: AG-GI AG-LG GI-LG

(1) (2) (1) (2) (1) (2)

�Portfolio expense ratio �4.33% �4.33% �4.36% �4.36% �4.35% �4.35%
�Portfolio risk 1.73% 2.30% 1.27%
�Sharpe ratio 0.92% �0.80% 1.18% �1.09% 1.60% 0.32%
Additional gross return (basis points) 9.7 10.6 3.9

Notes: The panel reports changes in a portfolio’s expense ratio, risk, and Sharpe ratio when funds in a portfolio
are managed by exactly the same fund managers. Additional gross return indicates how much additional gross
return is needed to have the same increase in a Sharpe ratio (reported in Column 1 for each portfolio) caused by
a decrease in a portfolio’s expense ratio.

43Y. Park / Financial Services Review 25 (2016) 29–49



holdings (Panel B) or multi-fund management by exactly the same fund managers (Panel C).
The results of portfolios of AG-GI, AG-LG, and GI-LG funds are obtained by combining the
assumptions in Panel A and the regression results in Tables 3 and 4. We compute a change
in a portfolio’s expense ratio, using the results on fund operating expenses (reported with
year and family fixed effects) in Column (2) of Table 3 and assuming that a 12b-1 fee is not
affected by a change in fund commonality.11 Next, we calculate a change in a portfolio’s risk,
using the results on return residual correlations (reported with family fixed effects) in
Column (6) of Panel B, Table 4. To investigate a change in a Sharpe ratio net of expenses
(Panels B and C, Table 5), we examine two cases: one considers only a decrease in a
portfolio’s expense ratio (Column 1), and the other considers both a decrease in a portfolio’s
expense ratio and an increase in a portfolio’s risk (Column 2). The first case is to reflect that
an investor disregards an enhanced portfolio risk resulting from an increase in the fund
commonalities, while the second case is to reflect that an investor incorporates the enhanced
risk into her portfolio return.

Panel B of Table 5 presents changes in a portfolio’s expense ratio, risk, and Sharpe ratio
when common stock holdings between two funds in a portfolio are increased by 10%. First,
for a portfolio of AG-GI funds, a 10% increase in common stock holdings has a negative net
effect on the portfolio’s risk-adjusted return. The increased fund commonality decreases the
portfolio’s expense ratio by 1.14%, a decrease which raises the portfolio’s Sharpe ratio by
0.24% (Column 1). However, the increased common stock holding also increases the
portfolio risk by 0.36%; as a result, the portfolio’s Sharpe ratio is eventually decreased by
0.12% (Column 2). Thus, the increased common stock holding results in a negative net effect
on the portfolio’s risk-adjusted return. For an increase in the Sharpe ratio caused by the
decreased portfolio expense ratio, we further investigate how much additional gross return is
needed to have the same increase in the risk-adjusted return. We find that the portfolio should
additionally earn a 2.0 basis point return to have the same increase of 0.24% in the Sharpe
ratio (reported in Column 1). Second, for a portfolio of AG-LG funds, a 10% increase in
common stock holdings results in a negative net effect on the portfolio’s risk-adjusted return,
a result which is similar to that of AG-GI funds. The increased fund commonality reduces
the portfolio’s expense ratio by 1.15%. The lower portfolio expense ratio increases the
portfolio’s Sharpe ratio by 0.31% (Column 1). However, the increased fund commonality
increases the portfolio risk by 0.45% (Column 2). As a result, the portfolio’s Sharpe ratio is
decreased by 0.14% (Column 2), indicating a negative net effect on the portfolio’s risk-
adjusted return. For an increase of 0.31% in the Sharpe ratio because of the decreased
portfolio expense ratio, the portfolio needs an additional gross return of 2.1 basis points.
Third, in contrast to portfolios of AG-GI and AG-LG funds, a 10% increase in common stock
holding results in a positive net effect on the risk-adjusted return of a portfolio of GI-LG
funds. The increased fund commonality reduces the portfolio’s expense ratio by 1.14%,
which increases the Sharpe ratio by 0.42% (Column 1). However, the increased fund
commonality raises the portfolio risk by 0.22% (Column 2). As a result, the Sharpe ratio is
increased only by 0.20% (Column 2). For an increase of 0.42% in the Sharpe ratio because
of the decreased portfolio expense ratio, the portfolio needs an additional gross return of 0.7
basis points. In summary, an increase in common stock holdings of a portfolio decreases the
portfolio’s expense ratio, but increases its risk. The increased portfolio risk can cause a

44 Y. Park / Financial Services Review 25 (2016) 29–49



negative net effect on the portfolio’s risk-adjusted return, as shown in the portfolios of
AG-GI and AG-LG. In addition, because of the increased portfolio risk, the portfolio needs
an additional return to have the same Sharpe ratio that an investor would have from the
decreased portfolio expense ratio.

Panel C of Table 5 presents changes in a portfolio’s expense ratio, risk, and Sharpe ratio
when funds in a portfolio are managed by exactly the same fund managers. Results in Panel
C are similar to those reported in Panel B, though the magnitudes of the changes are greater
in Panel C than in Panel B. The multi-fund management reduces a portfolio’s expense ratio
by 4.43% for AG-GI, 4.36% for AG-LG, and 4.35% for GI-LG funds, and the decreased
expense ratios raise the portfolios’ Sharpe ratios by 0.92%, 1.18%, and 1.60%, respectively
(Column 1). However, the multi-fund management also increases the portfolio risk by 1.73%
for AG-GI, 2.30% for AG-LG funds, and 1.27% for GI-LG funds. As a result, the Sharpe
ratios are decreased by 0.80% for AG-GI funds and 1.09% for AG-LG funds, while the
Sharpe ratio of GI-LG funds is increased by 0.32%. The results indicate that the multi-fund
management has negative net effects on risk-adjusted returns for portfolios of AG-GI and
AG-LG funds, while it has a positive net effect on risk-adjusted returns for a portfolio of
GI-LG funds. Finally, each portfolio needs an additional return to have the same Sharpe ratio
that the portfolio would have from the decreased portfolio expense ratio. The AG-GI
portfolio needs an additional gross return of 9.7 basis points; the AG-LG portfolio needs an
additional gross return of 10.6 basis points; and the GI-LG portfolio needs an additional
gross return of 3.9 basis points. In summary, similar to an increase in common stock
holdings, the multi-fund management by exactly the same managers decreases a portfolio’s
expense ratio, but increases its risk. The increased risk can bring a negative net effect on the
portfolio’s risk-adjusted return, as shown in the portfolios of AG-GI and AG-LG.

6. Conclusion

In this study, we provide extensive evidence that the commonality of funds within a family
can influence funds’ operating expenses and return correlations. Using common stock
holdings and multi-fund management as proxies for the fund commonality in a family, we
have analyzed a sample of 154 actively managed U.S. equity funds from 46 fund families for
the period of 2001 to 2006. We first find that common stock holdings and multi-fund
management can decrease funds’ operating expenses and that the influences on fund oper-
ating expenses occur through different channels: common stock holdings reduce a fund’s
management fee while multi-fund management lowers a fund’s other expenses that include
administrative fees. Second, we find that common stock holdings and multi-fund manage-
ment can increase the correlation of return residuals, which raises the correlation of fund
returns. Third, because the results indicate opposite effects of the fund commonalities on
risk-adjusted returns, we have investigated net effects on risk-adjusted returns of portfolios
with equity funds that have different investment objectives and found that the fund com-
monalities can have negative net effects on the portfolios’ risk-adjusted returns.

This study enhances an understanding of the commonality of funds within a family. We
expand previous research by showing that fund commonalities within a family can influence

45Y. Park / Financial Services Review 25 (2016) 29–49



fund operating expenses and return correlations. This study also provides practical implica-
tions to individual mutual fund investors. When investors construct a portfolio with different
low-cost equity funds from a single family, they should be aware of an investment risk that
fund commonalities that lower fund operating expenses can increase the correlation of fund
returns and, as a result, can reduce the portfolio’s risk-adjusted return. In other words, when
expense-conscious investors invest in equity funds in the same family, they need to pay
additional attention to fund commonalities such as funds’ common stock holdings and
management structure.

Notes

1 Mutual fund investors have also invested largely in low-cost funds. For example, at
year-end 2014, equity funds with expense ratios in the lowest quartile held 74% of the
funds’ total net assets, whereas those with expense ratios in the upper three quartiles
held only 26% (ICI, 2015).

2 The transaction services include providing statements and reports, disbursing divi-
dends, and paying state and local taxes and custodial, legal, audit, registration, and
directors’ fees (Collins, 2003; Latzko, 1999).

3 To merge the CRSP and the Thomson Financial Mutual Fund Holdings, we use the
MFLINKS files from the Wharton Research Data Services (WRDS).

4 When using fund returns to merge the two databases, we focus on monthly returns in
December and compare fund returns reported from the two databases. When the
difference in returns is less than or equal to 10 basis points in any year, we include the
funds in the sample. However, when the return difference is greater than 10 basis
points, we drop the funds from the merging process.

5 The Morningstar adds a note that funds having distinct portfolios are likely to have the
A share class.

6 An inclusion of the 14 dropped observations does not make regression results signif-
icantly different from the results reported in this article.

7 Year-end stock holdings that are used to calculate the average common stock holding
between funds are not matched with monthly fund returns that are used to compute the
correlations of return residuals. However, using the annual holdings data would not
significantly affect results reported in this study because it focuses on common stock
holdings of funds in the same family rather than funds’ trading behavior in which a
more precise measurement of the timing of trades, such as monthly holdings data, is
needed (Elton et al., 2010).

8 For common stock holdings by fund investment objectives, pairwise t-test results show
that common stock holdings are significantly different among the three investment
objectives: the difference between AG and LG funds is significant at 1% level
(t-statistics�5.03) and the differences between AG and GI and between GI and LG
funds are significant at 5% level (t-statistics � 2.41 for AG-GI and t-statistics � 2.45
for GI-LG). For the multi-fund management, pairwise t-test results show that the
difference between AG and LG funds are significant at 5% level (t-statistics � 2.31)

46 Y. Park / Financial Services Review 25 (2016) 29–49



but the differences between AG and GI funds and between GI and LG funds are not
statistically significant.

9 Common stock holding and multi-fund management are weakly positively correlated
with a correlation coefficient of 0.137 (p � 0.001). The positive correlation, however,
does not cause a serious multi-collinearity in estimating the regression models of Eqs.
(5) and (8). For example, the values of the variance inflation factor (VIF) for both
variables are lower than 10, which is usually used to indicate whether a serious
multi-collinearity problem exists or not (e.g., Hair et al., 1995; Kennedy, 1992; Neter,
Wasserman, and Kutner, 1989).

10 We check the robustness of the results presented in Table 3, using the data on fund
expense ratios from the Morningstar, instead of the CRSP. Appendix A reports
regression results. The results are not significantly different from those reported in
Table 3. Thus, our results are robust to the use of different data sources for fund
expense ratios.

11 The assumption of no change in 12b-1 fees would be valid because a 12b-1 fee that
relates to distribution services is less likely to be affected by a change in fund holdings
or fund management structure.

Appendix A Robustness check for the effects of fund commonalities on fund expenses

Operating expense Management fee Other expenses

(1) (2) (3) (4) (5) (6)

Common stock holding �0.248** �0.275** �0.158** �0.076* �0.037 �0.128**
(0.047) (0.064) (0.032) (0.035) (0.036) (0.043)

Multi-fund management (yes � 1) �0.067** �0.054* 0.014 0.018 �0.074** �0.054**
(0.016) (0.025) (0.012) (0.016) (0.012) (0.020)

Turnover ratio �0.018 �0.001 �0.018** �0.012 �0.002 0.016*
(0.012) (0.010) (0.007) (0.007) (0.009) (0.008)

Load fund (yes � 1) 0.044 �0.093 0.032 �0.014 0.012 �0.065
(0.103) (0.066) (0.033) (0.014) (0.081) (0.059)

Growth and Income (GI) �0.200** �0.186** �0.180** �0.165** �0.020 �0.038**
(0.019) (0.017) (0.015) (0.010) (0.016) (0.013)

Long-Term Growth (LG) �0.087** �0.092** �0.087** �0.079** �0.001 �0.030**
(0.019) (0.013) (0.013) (0.008) (0.016) (0.011)

Fund size �0.088** �0.068** �0.043** �0.028** �0.044** �0.042**
(0.010) (0.012) (0.006) (0.006) (0.007) (0.008)

Family size �0.033** 0.004 �0.038** 0.011 0.004 0.008
(0.005) (0.015) (0.004) (0.016) (0.004) (0.012)

Year fixed effects Yes Yes Yes Yes Yes Yes
Family fixed effects No Yes No Yes No Yes
Observations 817 817 895 895 817 817
Adjusted R2 0.432 0.701 0.468 0.807 0.103 0.546

Notes: This table is to check the robustness of the results presented in Table 3, using the data on fund expense
ratios from the Morningstar, instead of the CRSP. The dependent variables are operating expense (Columns 1 and
2), management fee (Columns 3 and 4), and other expense (Columns 5 and 6) of mutual funds. Fund size and
family size indicate the natural logarithm of one plus a fund’s year-end total net assets (in billions) and the natural
logarithm of one plus the year-end total net assets (in billions) managed by a family, respectively. Robust standard
errors are in parentheses.

*p � 0.05, **p � 0.01.

47Y. Park / Financial Services Review 25 (2016) 29–49



Acknowledgments

The author is grateful to Sojung Park and Jesse Thomas and would like to thank the editor
and anonymous referees for constructive comments.

References

Agarwal, V., Ma, L., & Mullally, K. (2015). Managerial Multitasking in the Mutual Fund Industry. Working
Paper (available at http://ssrn.com/abstract�1741282).

Berk, J. B., & van Binsbergen, J. H. (2015). Measuring skill in the mutual fund industry. Journal of Financial
Economics, 118, 1–20. http://dx.doi.org/10.1016/j.jfineco.2015.05.002.

Carhart, M. M. (1997). On persistence in mutual fund performance. Journal of Finance, 52, 57–82.
Choi, D., Kahraman, C. B., & Mukherjee, A. (2013). Performance-Chasing Behavior and Mutual Funds: New

Evidence From Multi-Fund Managers. Working paper (available at http://ssrn.com/abstract�2219846).
Ciccotello, C. S., Miles, J. A., & Walsh, L. S. (2006). Should investors choose funds from focused families?

Financial Services Review, 15, 247–264.
Ciccotello, C. S., Greene, J. T., & Walsh, L. S. (2007). Supermarket distribution and brand recognition of

open-end mutual funds. Financial Services Review, 16, 309–326.
Clare, A., O’Sullivan, N., & Sherman, M. (2014). Family status and mutual fund performance. J of Asset

Management, 15, 163–175.
Collins, S. (2003). The expenses of defined benefit pension plans and mutual funds fees. Investment Company

Institute Perspective, 9.
Collins, S. (2007). Fees and expenses of mutual funds 2006, Investment Company Institute Research Funda-

mentals, 16.
Dowen, R. J., & Mann, T. (2004). Mutual fund performance, management behavior, and investor costs. Financial

Services Review, 13, 79–91.
Elton, E., Gruber, M. J., & Green, T. C. (2007). The impact of mutual fund family membership on investor risk.

Journal of Financial and Quantitative Analysis, 42, 257–278.
Elton, E. J., Gruber, M. J., Blake, C. R., Krasny, Y., & Ozelge, S. O. (2010). The effect of holdings data frequency

on conclusions about mutual fund behavior. Journal of Banking & Finance, 34, 912–922.
Elton, E. J., Gruber, M. J., & Blake, C. R. (2012). Does mutual fund size matter? The relationship between size

and performance. Review of Asset Pricing Studies, 2, 31–55.
Gaspar, J.-M., Massa, M., & Matos, P. (2006). Favoritism in mutual fund families? Evidence on strategic

cross-fund subsidization. Journal of Finance, 61, 73–104.
Gil-Bazo, J., & Ruiz-Verdu, P. (2009). The relation between price and performance in the mutual fund industry.

Journal of Finance, 64, 2153–2183.
Gerken, W. C., Starks, L. T., & Yates, M. C. (2014). The Importance of Family: The Role of Mutual Fund Family

Reputation in Investment Decisions. Working paper (available at SSRN: http://ssrn.com/abstract�2022503).
Hair, J. F., Jr., Anderson, R. E., Tatham, R. L., & Black, W. C. (1995). Multivariate Data Analysis (3rd ed.). New

York, NY: Macmillan.
Haslem, J. A., Baker, H. K., & Smith, D. M. (2008). Performance and characteristics of actively managed retail

equity mutual funds with diverse expense ratios. Financial Services Review, 17, 49–68.
Investment Company Institute. (2015). 2015 Investment Company Fact Book (55th edition). Washington, DC.
Kennedy, P. (1992). A Guide to Econometrics. Oxford: Blackwell.
Khorana, A., & Servaes, H. (2007). Competition and Conflicts of Interest in the U.S. Mutual Fund Industry.

Working paper (available at http://faculty.london.edu/hservaes/competition.pdf).
Latzko, D. A. (1999). Economies of scale in mutual fund administration. Journal of Financial Research, 22,

331–339.
Latzko, D. A. (2003). Mutual Fund Expenses: An Econometric Investigation. Working paper, Pennsylvania State

University, York Campus.

48 Y. Park / Financial Services Review 25 (2016) 29–49



Mahoney, P. G. (2004). Manager-investor conflicts in mutual funds. Journal of Economic Perspective, 18,
161–182.

Malhotra, D. K., & McLeod, R. W. (1997). An empirical analysis of mutual fund expenses. Journal of Financial
Research, 20, 175–190.

Massa, M., Reuter, J., & Zitzewitz, E. (2010). When should firms share credit with employees? Evidence from
anonymously managed mutual funds. Journal of Financial Economics, 95, 400–424.

Moreno, D., & Rodríguez, R. (2013). Optimal diversification across mutual funds. Applied Financial Economics,
23, 119–122.

Neter, J., Wasserman, W., & Kutner, M. H. (1989). Applied Linear Regression Models. Homewood, IL: Irwin.
Pástor, L’uboš, Stambaugh, R. F. & Taylor, L. A. (2015). Scale and skill in active management. Journal of

Financial Economics, 116, 23–45.
Shawky, H. A., & Smith, D. M. (2005). Optimal number of stock holdings in mutual fund portfolios based on

market performance. Financial Review, 40, 481–495.
Sirri, E. R., & Tufano, P. (1998). Costly search and mutual fund flows. Journal of Finance, 53, 1589–1622.
Yadav, V. (2010). Portfolio Matching by Multi-Fund Managers: Effects of Fund Performance and Flows.

INSEAD University working paper.
Zhao, X. (2004). Why are some mutual funds closed to new investors? Journal of Banking and Finance, 28,

1867–1887.

49Y. Park / Financial Services Review 25 (2016) 29–49