To see is to believe: Corporate site visits and mutual fund herding

• Published date: 01 December 2023
• Author: Xiaofeng Quan,Cheng Xiang,Donghui Li,Kelvin Jui Keng Tan
• Date: 01 December 2023
• DOI: http://doi.org/10.1111/fima.12421

DOI: 10.1111/fima.12421

ORIGINAL ARTICLE

To see is to believe: Corporate site visits and

mutual fund herding

Xiaofeng Quan1Cheng Xiang2Donghui Li3Kelvin Jui Keng Tan4

1Business School, Soochow University,

Suzhou, China

2School of Economics and Business

Administration,Chongqing University,

Chongqing, China

3College of Economics, Shenzhen University,

Shenzhen, China

4Business School, The University of

Queensland, Brisbane, Australia

Correspondence

Cheng Xiang, School of Economics and

Business Administration,Chongqing

University,Chongqing, China.

Email: xiangcheng@cqu.edu.cn

Donghui Li, College of Economics, Shenzhen

University,Shenzhen 518060, Guangdong,

China.

Email: lidonghui2019@hotmail.com

Fundinginformation

National NaturalScience Foundation of China,

Grant/AwardNumbers: 71772131,

71973018; National Social Science Fundof

China, Grant/AwardNumber: 22BGL295

Abstract

Using a unique data set of corporate site visits by mutual

funds to Chinese firms listed on the Shenzhen Stock

Exchange from 2013 to 2021, we find that firms with vis-

its (more visits) are associated with lower mutual fund

herding than those with no (fewer) visits. In addition, we

demonstrate that mutual funds’ visits to a firm drive the

change in their herding propensity by verifying hard infor-

mation (e.g., the firm’s technology, innovation, accounting,

and finance information) and obtaining soft information (e.g.,

management’s risk appetite, employee morale, and corpo-

rate culture). Furthermore,corporate site visits are found to

strengthen herding’s price impact without return reversals.

Overall, our results are consistent with information cascade

theory.

KEYWORDS

corporate site visits, institutional herding, mutual funds, soft infor-

mation

1INTRODUCTION

Long-standing literatureshows that mutual funds trade similarly, or “herd” (Brown et al., 2014; Jiang & Verardo, 2018;

Koch, 2017; Lakonishok et al., 1992; Wermers, 1999). Although existingstudies argue that information plays an essen-

tial role in the formation of mutual fund herding (Avery & Zemsky,1998; Banerjee, 1992; Froot et al., 1992), little is

known about how mutual funds’ information-acquisition activities affect their herding behaviors. One possible expla-

nation is that such activities are usually private, without relevant data for testing (Cheng et al., 2016). This study aims

to fill this gap by employing a unique, recently available data set of corporate site visits by mutual funds, which are

© 2023 Financial Management Association International.

Financial Management. 2023;52:711–740. wileyonlinelibrary.com/journal/fima 711

712 QUAN ETAL.

not mandatorily disclosed in most major markets exceptfor China’s Shenzhen Stock Exchange (SZSE). In other words,

this paper investigates the impact of a specific type of newly availableinformation-acquisition activity (corporate site

visits) on mutual fund herding.

Corporate site visits are globally prevalent (Brown et al., 2015; Cheng et al., 2016) and are very important because

they offer visitors private access to management teams (Koch et al., 2013). However, there are three competing

theories of how corporate visits affect mutual fund herding. On the one hand, information cascade theory predicts

a negative relationship between corporate visits and mutual fund herding. On the other hand, both correlated sig-

nal theory and characteristic herding theory suggest a positive relationship. Specifically,information cascade theory

suggests that mutual funds imitate their peers’ earlier decisions when they have what they believe to be unreliable

information (Banerjee, 1992; Bikhchandani et al., 1992;Sias,2004). Meanwhile, previous studies document that cor-

porate site visits facilitate the collection of private information that helps visitors better understand firms in two ways

(Chen et al., 2022; Cheng et al., 2016; Han et al., 2018; Liu et al., 2017). First, site-visiting mutual funds can collect soft

information about firms’ operation management standardization, management risk appetites, employee morale, and

corporatecultures to infer the firms’ operational situations and prospects.1Second, visitors may take the advantage of

question and answer (Q&A) sessions to inquire about firms’ detailed, contextual hard information, such as their pub-

lic announcements, and verify the quality of the hard information. In both cases, they increase the reliability of their

own information. In this way,information cascade theory predicts that corporate site visits will attenuate mutual fund

herding.

Correlated signal theory argues that mutual funds make similar trades simply because they receive similarfunda-

mental information about firms (Choi & Skiba, 2015; Froot et al., 1992). Different mutual funds may receive similar

signals if they successively visit the same firm in a short period. Consequently, they makesimilar trades. If so, more

corporate site visits predict higher mutual fund herding. Characteristic herding theory argues that mutual funds herd

because they tend to investin firms with similar characteristics. Cheng et al. (2019) found that the likelihood of visiting

a site correlates with firm characteristics, such as size, book-to-market ratio, age, and profitability.Hence, corporate

visits further amplify the characteristic investing method that induces mutual fund herding. As the discussion above

shows, whether firms that mutual funds visit more are associated with higher or lower herding remains an empirical

issue.

Using the unique data set of mutual funds’ corporate site visits to SZSE firms from 2013 to 2021, we find that firms

with visits (more visits) are associated with lower herding than those with no (fewer) visits. The herding measure con-

structed using the methodology of Lakonishok et al. (1992) is approximately 10.2% lower for firms visited by mutual

funds than for those not visited, controlling for a set of known herding factors and firm, industry, and quarter fixed

effects. This negative impact of corporate site visits by mutual funds on their herding is robust to alternativesamples

and alternative herding measures.

We then use the instrumental variable approach to establish causality between the negative impacts. Specifically,

we instrument the visit measures with the travel convenience scores of the cities in which the firms are headquar-

tered. Convenienttransportation facilitates mutual funds’ visits to a firm and is exogenous to their herding propensity.

We find that the instrumented visit measures are significantly and negatively related to herding, indicating the causal

impact of visits on herding.

Further analyses show that corporate site visits by mutual funds mitigate their herding but insignificantly impact

that of non-visiting mutual funds. In addition, no evidence shows that non-visiting mutual funds imitate the trades of

their visiting peers. These findings demonstrate that mutual funds’ firm visits directly drivethe change in their herding

propensity.The nature of the site visit involves personal, face-to-face interactions between visitors and managers. The

soft information that site-visiting mutual funds collect and their verification of firms’ hard information are private and

1Because soft information is subjective and textual (Liberti & Petersen,2019), its value and usage are also subjective, depending on visiting mutual funds’

priorbeliefs. The obtained soft information allows visiting mutual funds to verify the reliability of their own information (i.e., updating their priors) but not to

makesimilar decisions if they have different ex ante beliefs.