Mutual Fund Governance and Performance: A Quantile Regression Analysis of Morningstar's Stewardship Grade

• Published date: 01 July 2011
• Author: Carl R. Chen,Ying Huang
• DOI: http://doi.org/10.1111/j.1467-8683.2011.00858.x
• Date: 01 July 2011

Mutual Fund Governance and Performance: A

Quantile Regression Analysis of Morningstar’s

Stewardship Grade

Carl R. Chen and Ying Huang*

ABSTRACT

Manuscript Tytpe: Empirical

Research Question/Issue: We study the relation between mutual fund performance and Morningstar’s f‌iduciary grades

using quantile regression models. This research is important because we shed new light on the principal-agent issue in the

mutual fund industry, and investors and professional managers often use Morningstar’s rating system to guide their

investment decisions. Quantile regressions allow us to examine the differential impact of mutual fund governance variables

such as manager incentives and board quality on fund performance across the entire performance distribution.

Research Findings/Insights: Quantile regressions f‌ind a strong contemporaneous association between Stewardship Grade

and Sharpe Ratio. Manager incentives, though not signif‌icant in the OLS regression, are positively related to the fund

performance for better performing funds, and negatively correlated with portfolio turnover. However, manager incentives

are shown to have little predictive power for funds’ future performance. In contrast, board quality, which is in the center

stage of the proposed SEC mutual fund regulation, bears little relationship with contemporaneous fund performance, yet

is strongly related to funds’ future performance.

Theoretical/Academic Implications: Our analyses and results have two important implications for the principal-agent

theory. Our f‌inding that board quality has a more profound impact on fund performance than manager incentives suggests

that incentive alignment through compensation contracts is less effective in mitigating agency problems than quality board

monitoring. Our quantile regression results also provide new insights to the analysis of corporate governance research. The

major advantage of using quantile regression is that it reveals the whole spectrum of heterogeneous mutual fund perfor-

mance responses to their governance efforts, especially the behavior of funds at the tails of the distribution; hence better

corporate governance policies can be drawn up and instituted. This methodology is useful for other governance research as

well and entails meaningful regulatory implications when f‌irms/industries are heterogeneous and the conditional means

of traditional regression analysis does not provide much information and guidance for corporate decision making.

Practitioner/Policy Implications: This study shows that fund performance is associated with manager incentives contem-

poraneously, while board quality predicts fund performance in the long haul. This f‌inding has important regulatory

implications, as the eff‌icacy of SEC’s proposal on board and chairman independence is under great debate. Our results also

provide useful information for investors who use Morningstar’s Stewardship grades to guide their investments.

Keywords: Corporate Governance, Board of Director Mechanisms, Stewardship, CSR Mechanisms, Executive Compen-

sation, Board Policy Issues, Corporate Governance Rating/Index, Firm-Level Governance Outcomes

INTRODUCTION

The objective of this study is to investigate the relation

between mutual fund governance and fund perfor-

mance. The conf‌lict of interest between mutual fund share-

holders and fund managers is obvious–shareholders try to

maximize their realized risk-adjusted returns, while fund

managers have a strong incentive in expanding the fund

size because management fees are determined by the asset

size (See Mahoney, 2004). Managerial incentive alignment

and/or board monitoring offer viable solutions to this

agency problem. Pay-performance relationship docu-

mented in Murphy (1999), Core, Guay, & Larcker (2003),

and Hall & Liebman (1998) supports the view that incen-

tive pays, such as bonuses, options, and stock grants align

managerial interest with that of the shareholders. In

*Address for correspondence: Academyof Financial Research and College of Econom-

ics, Zhejiang University,38 Zheda Road, Hangzhou, 310027, China. Tel:+86-571-8795-

1721; Fax: †86-571-8795-3937; E-mail: yhuang616@gmail.com

311

Corporate Governance: An International Review, 2011, 19(4): 311–333

© 2011 Blackwell Publishing Ltd

doi:10.1111/j.1467-8683.2011.00858.x

addition to the managerial incentive alignment, a second

strand of the literature emphasizes the effectiveness of

board monitoring. Some f‌ind a positive relation between

f‌irm performance and board independence (e.g., Rosen-

stein & Wyatt, 1990), yet others do not (e.g., Hermalin &

Weisbach, 1991, and Yermack, 1996).

Differing from previous studies, we examine Morningstar

Stewardship Grade and compare the effectiveness of its two

major components, namely, manager incentives and board

quality, as proxies for governance effectiveness in a single

study. Morningstar’s grading system provides us with an

integrated grading for mutual funds, and this makes our

study differ from most previous ones that concentrate on

individual aspects of corporate governance, e.g., blockhold-

ings (Demsetz & Lehn, 1985; Demsetz & Villalonga, 2001),

board size (Eisenberg, Sundgren, & Wells, 1998; Yermack,

1996), and board composition (Hermalin & Weisbach, 1991).

Moreover, the comparison of the effect of manager incen-

tives with that of the board quality on governance is impor-

tant, as the SEC proposed a governance mechanism that

requires an independent chairman and a board consisting of

at least 75 per cent independent directors. We intend to

address whether managerial incentive or board quality is

more effective in aligning managerial and shareholder inter-

ests. In addition to examining the contemporaneous relation

between fund performance and governance, we also study

the ability of these governance measures to predict future

fund performance as evidence from this analysis can

provide more information to both regulators and retail

investors.

Using Morningstar’s data serves three purposes. First,

Morningstar’s mutual fund rating system is very popular

among retail investors. Retail investors and some profes-

sional managers regularly use Morningstar’s rating system

to guide their investment decisions; hence it is important to

study the effectiveness of such a rating system. Del Guercio

& Tkac (2008) f‌ind signif‌icant positive (negative) fund f‌lows

following Morningstar rating upgrades (downgrades).

Second, Morningstar’s data provide us with a single source

of data to compare the relative importance of various mea-

sures of fund governance in relation to mutual fund perfor-

mance. Third,we examine the effectiveness of Morningstar’s

Stewardship Grade using a different and more pertinent

methodology than previous research in this regard.

Specif‌ically, the employment of quantile regressions

allows us to examine the differential behavior of fund per-

formance across the entire performance distribution. Such

analysis becomes very useful when the relation between

variables varies substantially across distributions. On the

other hand, traditional analysis based on the OLS method is

inappropriatein this setting because the OLS regression only

estimates the conditional means, and therefore the OLS

method may fail to capture certain non-negligible relations

between Morningstar’s governance grades and fund perfor-

mance at the tails of fund performance distribution.

As previously indicated, fund managers have a strong

incentive in expanding the fund size because management

fees are determined by the asset size (See Mahoney,2004). To

alleviate this agency problem, mutual funds design incen-

tive contracts that are commensurate with fund manager

performance, and their governance tends to heavily rely on

their board. However, unlike other f‌inancial institutions,

mutual fund industry’s governance structure is lightly regu-

lated – guided only by the Investment Company Act of 1940.

For example, the 2003 mutual fund scandal involves late

trading and market timing. On September 3, 2003, New York

Attorney General Eliot Spitzer issued a complaint against

Canary Capital, accusing Canary of engaging in late trading

in collusion with Bank of America’s Nations Funds. Late

trading enabled Canary to purchase mutual fund shares at

the closing price after the market had closed. Such trading

impairs fund shareholders’ interest because Canary unfairly

used information about after-hours market developments in

foreign markets. Canary settled the complaints for $40 mil-

lion.1Spitzer and the SEC also charged some other mutual

fund groups, e.g., Strong Capital, Janus, Bank One’s One

Group, and Invesco of market timing, which allowed their

favored clients to trade frequently to take advantage of

market volatility. Market timing conducted by these funds

increases fund cost at the expense of other shareholders.

Indeed, Radin & Stevenson (2006) f‌ind that the governance

model of the mutual fund industry has signif‌icant structural

difference from its corporate counterparts, hence dilutes the

authority of its directors.

These widespread fund f‌laws have prompted the SEC to

propose a stricter fund governance mechanism that requires

an independent chairman and a board consisting of at least

75 per cent independent directors. Such controversial pro-

posal, however, encountered intensive debates in both the

academia and the industry, and was challenged by the Com-

merce Department. Although SEC proposed new guidelines

for the governance of mutual funds, existing researches f‌ind

that the effectiveness of such a requirement is disputable.

For example, Ferris & Yan (2007) conclude that neither the

probability of a fund scandal nor the overall performance is

related to the independence of the chairman or board direc-

tors. In fact, the SEC proposal was ruled by the US Court for

violating the Administrative Procedure Act.

This papercontributes to this avenue of research by exam-

ining the relation between various fund governance mecha-

nisms and fund performance. Our study differs from prior

literature in two aspects. First, we employ the most up-to-

date Morningstar’s Stewardship Grade and its components

to examine the relation between fund performance and

various measures of governance; hence we are able to

compare the effectiveness of these governance measure-

ments in a single study with a sample encompassingperiods

of both economic boom and bust. Therefore, our results

provide important insights for retail investors. Second, we

adopt a quantile regression model to study the relation

between fund performance and fund governance over the

entire distribution of fund performance. This type of mod-

eling is more informative and powerful in analyzing the

diverse mutual fund universe, where funds are heteroge-

neous with signif‌icantly different trading strategies and

investment objectives. It is recognized that the traditional

OLS regression is less informative as it estimates the condi-

tional mean, which essentially treats mutual funds as a

group of investment advisors with homogeneous trading

strategies.

Our empirical f‌indings reveal that OLS results are less

informative than those generated by quantile regressions.

312 CORPORATE GOVERNANCE

Volume 19 Number 4 July 2011 © 2011 Blackwell Publishing Ltd