The Lure of the Brand: Evidence from the European Mutual Fund Industry

Published date01 November 2015
Date01 November 2015
The Lure of the Brand: Evidence from
the European Mutual Fund Industry
Jan Jaap Hazenberg
NN Investment Partners, Schenkkade 65, 2595 AS Den Haag, The Netherlands
Fabian Irek
Luxembourg School of Finance, University of Luxembourg, 2B Rue Albert Borschette,
L1246 Luxembourg
Willem van der Scheer
NN Investment Partners, Schenkkade 65, 2595 AS Den Haag, The Netherlands
Mariela Stefanova
SPF Beheer, Arthur van Schendelstraat 850, 3511 ML Utrecht, The Netherlands
We investigate the effect of the fund familys brand on mutual fund ows by using a
unique data set that represents a direct assessment of the brand image of European
fund providers. A superior brand image increases the sensitivity of ows to good
past performance, while it protects against outows when there is under-
performance. Flows of funds of independent providers have a higher sensitivity to
past performance and brand image than ows of funds of providers owned by banks
or insurers. These ndings highlight the importance of marketing and brand in
generating growth for the fund families.
Keywords: mutual funds, fund flows, fund performance, marketing, brand image
JEL classification: G11, G23, M31
We would like to thank David Rakowski, Rafael Zambrana, Lars Kaiser, two anonymous
European Financial Management referees, the editor John Doukas, the participants at the
OxfordMan Institute internal seminar series and at the University of Bath, School of
Management seminar, for their comments. Any remaining errors are our own. We are
grateful to Fund Buyer Focus for making their brand data available to us. Mariela Stefanova
conducted this research while at NN Investment Partners. Fabian Irek is supported by the
National Research Fund, Luxembourg. Correspondence: Fabian Irek.
European Financial Management, Vol. 21, No. 5, 2015, 867904
doi: 10.1111/eufm.12046
© 2014 John Wiley & Sons Ltd
1. Introduction
For fund providers,
net ows into their funds, as a result of investor purchasing and selling of
mutual fund units, translate directly into incremental assets under management and revenues.
Based on empirical evidence, it is well documented that investors tend to chase past
performance when selecting funds.
This nding is consistent with the notion that investors
deduct information about contemporaneous management ability from past performance.
There is extensive marketing literature that suggests that the brand of a product or service
can affect its sales (e.g., Keller, 1993; Shankar et al., 2008). If these insights from the
marketing literature apply to the fund management industry, one would expect investors to
direct their ows based not only on objective measures, such as past performance, but also in
response to marketing efforts and after considering the fund brand image.
There is ample anecdotal evidence that fund advertising can be effective and that the
fund provider brand image has an inuence on fund ows. Fidelity, for example, is said to
have spent EUR 107.5 million on its panEuropean marketing and advertising effort in
the veyear period up to 2000 in order to break into the European mutual fund market.
From its market entry in continental Europe in 1990, Fidelity grew to become the number
11 fund provider in Europe in terms of assets under management at the end of 2001. There
are also examples of how a drop in brand image can result in fund outows. For example,
when Putnams reputation became tarnished in 2003 because of its involvement in the late
trading and market timing scandal in the USA, investors in that market withdrew USD 54
billion in assets from Putnam funds just in the fourth quarter of 2003 (Markham, 2006).
There are several empirical academic studies investigating fund ows that analyse
choice factors related to marketing and distribution and nd that these choice factors
inuence fund ows. These studies can be divided into two categories: one using indirect
proxies for brand or marketing effort (Gruber, 1996; Sirri and Tufano, 1998; Frye, 2001;
Huang et al., 2007) and one using advertising data as a direct proxy (Jain and Wu, 2000;
Peng et al., 2011; Wei et al., 2011; Yankow et al., 2011). Both Frye (2001) and Yankow
et al. (2011) also take into account the providers fund distribution model. Our main
contribution is that we use a direct, surveybased assessment of brand image, rather than
indirect proxies. The data was obtained from a European research company, Fund Buyer
Focus, which has conducted annual surveys that identify how professional fund buyers
assess the brand of European mutual fund families. The data was gathered from 826908
fund buyers in 11 countries over the period 20042011. We are the rst to use this new
dataset for an empirical study. When merged with a sample of equity funds domiciled in
Luxembourg within 30 different Morningstar categories (i.e., fund objectives) that are
In this study, we use the terms, fund management company and fund provider, or simply
provider, interchangeably. The term, fund family, is used for the range of funds offered by a
provider. All funds in the same fund family have the same brand.
See, for example, Gruber (1996), Chevalier and Ellison (1997), Sirri and Tufano (1998) for
the USA, Keswani and Stolin (2008) for the UK, and Ferreira et al. (2012) for an international
comparison. Using market share changes as ow measure, Spiegel and Zhang (2013) conrm
the positive relationship between past performance and ows. However, according to their
results, as opposed to studies using relative ows, the relationship between market share
changes and past performance is not convex.
See Wall Street Journal Europe, 8 February 2001.
© 2014 John Wiley & Sons Ltd
868 J. J. Hazenberg, F. Irek, W. van der Scheer and M. Stefanova
sold across Europe, the nal sample contains 2,179 funds from 215 different brands. For
these funds, we collect data on, among others, performance, ows and business model.
The research adds to the existing marketing literature by establishing a direct relationship
between the products brand image (in this case, a mutual fund) and subsequent sales (in
this case, the net new money invested by consumers).
Our research adds to the existing nance literature by rst comparing and contrasting
our survey brand data to the brand proxies used in earlier studies. Gruber (1996) suggests
that, in addition to past performance, marketing effort and general reputation also play a
role in fund ows. With previous period ows as a proxy for these marketingrelated
elements of the fund industry, both Gruber (1996) and Frye (2001) nd that these ows
have signicant explanatory power for subsequent ows. Although we conrm that fund
ows are related to lagged ows, we show that, at the fund family level, neither lagged
ows nor lagged performance can serve as a proxy for brand. Huang et al. (2007) show
that performance sensitivity depends on several proxies for brand, such as family assets
under management, the number of categories the family offers and whether the family has
produced star funds. We run a comparison between several brand proxies at the fund
family level, where we nd that the number of categories or funds in a fund family seems
to be the best proxy of brand. Finally, we also compare the effect of each of the brand
proxies on the owperformance relationship at the individual fund level and nd again
that the number of categories changes the owperformance relationship in a manner
most similar to our brand variable.
A second contribution by our study is that in using the new brand data set and nding
that brand has a substantial effect on fund ows, we conrm the results found by studies
using advertising data that marketing aspects of the fund management business
signicantly inuence ows. Jain and Wu (2000) nd that in the oneyear period prior to
the advertisement, funds that advertise perform signicantly better than the benchmarks,
but show inferior performance in the year after the advertisement. Compared to a control
group of funds that are similar in terms of past performance, past ows and fund size,
ows into funds that advertise are larger than ows into funds that do not advertise.
Yankow et al. (2011), Wei et al. (2011), and Peng et al. (2011) also nd that fund
investors are sensitive to advertising. Our results indicate that funds with a good brand
receive more inows. When dividing funds into ve groups, the superior brand funds get
3.4% more inows annually than the inferior brand funds. At the same time, the ows into
superior brand funds are three times (when using relative ows) to four times (when using
the market share changes of Spiegel and Zhang (2013)) more persistent than the ows into
their inferior brand peers. Although we nd that brand also has an inuential effect, past
performance is the stronger and more consistent driver of ows than brand because the
difference in annual relative ows for top and bottom quintile performing funds is 49%.
It should be noted, however, that performance is measured at the individual fund level,
whereas brand is at the fund family level. Therefore, we proceed to compare fund family
performance with brand. We nd that brand is not simply a proxy for past fund family
performance. However, brand has a strong inuence on the convex owperformance
relationship at the individual fund level, which is also of high economical signicance. In
particular, brand has a positive impact on ows for highperforming funds, which obtain
27% more relative inows per year when they have a superior brand than when their brand
is inferior. When using the market share change measure, the superior brand funds with
high past performance have market share changes which are about three and a half times
larger than those of inferior brand funds with high past performance. We also nd these
© 2014 John Wiley & Sons Ltd
The Lure of the Brand 869

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT