Mandatory Portfolio Disclosure, Stock Liquidity, and Mutual Fund Performance

• DOI: http://doi.org/10.1111/jofi.12245
• Published date: 01 December 2015
• Date: 01 December 2015

THE JOURNAL OF FINANCE •VOL. LXX, NO. 6 •DECEMBER 2015

Mandatory Portfolio Disclosure, Stock Liquidity,

and Mutual Fund Performance

VIKAS AGARWAL, KEVIN A. MULLALLY, YUEHUA TANG,

and BAOZHONG YANG∗

ABSTRACT

Weexamine the impact of mandatory portfolio disclosure by mutual funds on stock liq-

uidity and fund performance. We develop a model of informed trading with disclosure

and test its predictions using the May 2004 SEC regulation requiring more frequent

disclosure. Stocks with higher fund ownership, especially those held by more informed

funds or subject to greater information asymmetry,experience larger increases in liq-

uidity after the regulation change. More informed funds, especially those holding

stocks with greater information asymmetry, experience greater performance deteri-

oration after the regulation change. Overall, mandatory disclosure improves stock

liquidity but imposes costs on informed investors.

MANDATORY DISCLOSURE OF PORTFOLIO holdings by institutional money man-

agers is a vital component of securities market regulation. Mandated by the

Securities Exchange Act of 1934 and the Investment Company Act of 1940,

∗Agarwal is with the J. Mack Robinson College of Business at Georgia State University and

the Center for Financial Research at University of Cologne. Mullally and yang are with the J.

Mack Robinson College at Georgia State University. Tang is with the Lee Kong Chian School

of Business at Singapore Management University. This paper has benefited from comments and

suggestions from two anonymous referees and Ken Singleton (the Editor), Robert Bartlett, Con-

rad Ciccotello, Chris Clifford, Sevinc Cukurova, Gerry Gay, Thomas George, Edith Ginglinger,

Itay Goldstein, Christian Gourieroux, Carole Gresse, Alan Huang, Jiekun Huang, Wulf Kaal,

Jayant Kale, Madhu Kalimpalli, Matti Keloharju, Kate Litvak, Pedro Matos, Michelle Paul, Blake

Phillips, Sugata Ray, Adam Reed, Christopher Schwarz, Clemens Sialm, Laura Starks, Avanid-

har Subrahmanyam, Craig Tyle, Mathijs van Dijk, Jin Wang, Kelsey Wei, Russ Wermers, and

seminar participants at the 2013 Conference for Empirical Legal Studies, the 2013 FMA Annual

Meetings, the ICI/CFP Academic/Practitioner Conference at the University of Maryland, the 6th

Professional Asset Management Conference, Aalto University,Georgia State University, Louisiana

State University,Universitas Negeri Jakarta, University of Paris-Dauphine, University of Sydney,

University of Technology Sydney, Universitas Tarumanagara, University of Waterloo, and Wilfred

Laurier University. We are especially grateful to Zhi Da, Pengjie Gao, and Ravi Jagannathan for

data on liquidity- and characteristics-adjusted mutual fund performance; Jianfeng Hu for TAQ

order imbalance data; Lei Jiang, Mahendrarajah (Nimal) Nimalendran, and Sugata Ray for data

on TAQ liquidity measures; and Christopher Schwarz for data on mutual fund holdings. Wewould

also like to thank Muneem Ahad and Ashutosh Tyagi for excellent research assistance. Kevin Mul-

lally thanks the Center for Economic Analysis of Risk (CEAR) at Georgia State University for its

financial support. YuehuaTang acknowledges the D.S. Lee Foundation Fellowship from Singapore

Management University.None of the authors of this article have conflicts of interest as defined by

the policy of The Journal of Finance.

DOI: 10.1111/jofi.12245

2733

2734 The Journal of Finance R

portfolio disclosure provides the public with information about the holdings

and investment activities of institutional investors. Among the mandatory dis-

closure requirements on institutional investors, those on mutual funds result

in perhaps the most detailed information about portfolio holdings (see Sec-

tion I for more details). Such disclosure requirements have broad implications.

On the one hand, mandatory portfolio disclosure can help improve the trans-

parency of capital markets. On the other hand, it can reduce fund managers’

incentives to collect and process information. To shed light on the costs and

benefits of mandatory portfolio disclosure by mutual funds, we examine how

disclosure affects (i) the liquidity of disclosed stocks and (ii) fund performance.

Identifying the causal effects of portfolio disclosure on stock liquidity and

fund performance presents a challenge. We overcome this challenge by using a

May 2004 Securities and Exchange Commission (SEC) regulation change that

required mutual funds to increase their disclosure frequency from a semiannual

basis to a quarterly basis. We use this regulation change as a quasi-natural

experiment to identify the effects of funds’ portfolio disclosure on stock liquidity

and fund performance.

Our empirical analyses build on the theoretical literature on mandatory dis-

closure and informed trading. Huddart, Hughes, and Levine (2001) extend the

Kyle (1985) model to study mandatory disclosure of trades by informed traders.

Here, we further extend this framework to allow for different mandatory dis-

closure frequencies.

Our model yields several testable predictions. First, our model predicts that

more frequent disclosure by informed traders improves market liquidity as

measured by market depth, that is, the inverse of the Kyle (1985) lambda.

The intuition is that, with mandatory disclosure, the market maker can infer

information from the disclosed positions of informed traders as well as from

aggregate order flows, which reduces the impact of informed trades on prices.

Second, the liquidity improvement should be greater for stocks subject to higher

information asymmetry. Third, informed traders’ profits are negatively related

to disclosure frequency because the disclosure of trades limits a trader’s ability

to reap all of the benefits of his information. Finally, an informed trader’s profit

decrease should be positively related to both the level of information asymmetry

in the stocks the trader holds and the time it takes the trader to complete his

trades.

We begin our analysis by examining the impact of an increase in portfo-

lio disclosure frequency on the liquidity of disclosed stocks. A large body of

literature shows that mutual funds’ portfolio disclosures contain valuable in-

formation (see Section II for more details). Accordingly, stocks with higher

fund ownership should experience greater increases in liquidity with more fre-

quent disclosure. To test this prediction, we employ a difference-in-differences

approach to examine the change in stock liquidity during the two-year pe-

riod around the SEC rule change in May 2004. In particular, we examine

how changes in stock liquidity (first difference) vary with the ownership of

actively managed domestic equity mutual funds (second difference). Ge and

Zheng (2006) document that some funds voluntarily disclose their portfolios.

Portfolio Disclosure, Liquidity, and Fund Performance 2735

We carefully identify funds that disclose to sources such as Morningstar and

Thomson Reuters in addition to the SEC EDGAR (Electronic Data Gathering,

Analysis, and Retrieval) database (see Section IV for more details). We exclude

these voluntarily disclosing funds to construct a sample of funds that increase

their disclosure frequency due to the 2004 SEC rule change. Focusing on this

sample of funds allows us to isolate the effect of the regulation change from the

voluntary disclosure behavior of certain funds.

We find that stocks with higher fund ownership experience greater improve-

ments in liquidity after funds are subject to more frequent mandatory disclo-

sure. Moreover, the increase in liquidity is economically large. For instance, a

one standard deviation increase in the ownership of funds forced to increase

their disclosure frequency due to the regulation change is associated with a

0.13 and 0.29 standard deviation decrease in the Amihud (2002) illiquidity

measure and Trade and Quote (TAQ) relative bid-ask spread, respectively.

To corroborate this finding, we conduct several sets of placebo tests. First,

we use two types of institutional investors, nonmutual-fund 13F institutions

and hedge funds, as control groups for our cross-sectional placebo tests. The

underlying argument is that the 2004 regulation change only applies to mutual

funds, but not to other institutional investors. In addition, we use domestic

equity index funds as a control group. Unlike the treatment group of actively

managed funds, index funds are passive and thus their disclosed portfolios

should not contain private information. We find that the ownership of actively

managed funds has a larger impact on the change in stock liquidity than does

the ownership of nonmutual funds, hedge funds, or index funds. Second, we

conduct a time-series placebo test using a two-year period around November

2006 as our placebo period. We choose this period to avoid any overlap with

major market events (e.g., 2008 financial crisis) that may affect stock liquidity.

We do not find similar effects of mutual fund ownership on stock liquidity

during the placebo period. Together, cross-sectional and time-series placebo

tests suggest that our finding of improvement in stock liquidity is not driven

by institutional ownership or a time trend in liquidity.

As mentioned earlier, some funds voluntarily disclosed on a quarterly basis

prior to the regulation change. For these funds, the effect of the increase in

mandatory disclosure frequency on stock liquidity should be weaker because

the frequency with which they disclose remains unchanged.1Ge and Zheng

(2006) argue that the decision to voluntarily disclose is strategic. Thus, follow-

ing their study we use the propensity score from a logistic model to construct

a control sample of voluntarily disclosing funds. We find that, compared to the

ownership of voluntarily disclosing funds, the ownership of funds that increase

their disclosure frequency due to the regulation change has a larger effect on

stock liquidity.

1For example, consider a fund that mandatorily discloses twice to the SEC and voluntarily

discloses twice to a data vendor (Morningstar or Thomson Reuters) in the year prior to the rule

change. Subsequent to the rule change, this fund will mandatorily disclose four times per year to

the SEC.