Has local informational advantage disappeared?

• Author: Qin Wang,Alok Kumar,Johan Sulaeman,Gennaro Bernile
• Date: 01 January 2019
• Published date: 01 January 2019
• DOI: http://doi.org/10.1002/rfe.1046

38

|

wileyonlinelibrary.com/journal/rfe Rev Financ Econ. 2019;37:38–60.

© 2019 University of New Orleans

1

|

INTRODUCTION

Existing studies document that geographic proximity affects the behavior and performance of local agents even when there is

no institutional barrier to entry. Coval and Moskowitz (1999, 2001) find that U.S. institutional investors overweight their local

stocks—relative to other U.S. stocks—and earn higher returns from their investments in local stocks. In a similar vein, Malloy

(2005) finds that analysts located near firm headquarters have an informational advantage. The advantage associated with geo-

graphic proximity may arise because local agents are able to directly inspect local firms and acquire information about these

firms. They may also be able to collect information from other local sources (e.g., customers and suppliers, local media) at a

lower cost. The informational advantage of local agents, however, should depend on the firms’ information environment. A

more opaque and less competitive environment should provide more profitable opportunities for local agents, if they can obtain

information at a lower cost. In contrast, the lifespan of such advantage may be shortened when the playing field is more level.

One important determinant of the quality and competitiveness of firms’ information environ ment is regulation. Rules that

curb uneven access to private information and improve the quality of publicly available information should reduce the potential

advantage of local agents. At the turn of the millennium, the U.S. securities regulation moved further in this direction with

the enactment of the Regulation Fair Disclosure (Reg FD) and the Sarbanes–Oxley Act (SOX). The new rules were designed

to affect the flow and timing of information dissemination, as well as influence the quality and reliability of information that

firms disclose publicly.1

During the same period, there was a breathtaking increase in the availability of and speed of access to corporate information

due to the Internet. Internet users increased from 35.8% to 61.7% of the U.S. population between the end of 1999 and 2003.2

Although broadband access had begun to grow, the vast majority of households continued to rely on dial- up connections at the

turn of the millennium. Between the end of 1999 and 2003, however, availability of broadband access rose from about 1% to

over 20% of the U.S. population.3 This epochal technological progress allowed wider and more timely access to firm disclosures

(e.g., via the SEC’s EDGAR system of corporate filings), as well as the widespread electronic dissemination of corporate news

(e.g., Bloomberg).

If local bias and superior local performance depend on the information environment, we expect that they would decline fol-

lowing these exogenous changes to the quality and competitiveness of the environment. We first examine this conjecture using

Received: 17 July 2018

|

Accepted: 6 September 2018

DOI: 10.1002/rfe.1046

SPECIAL ISSUE ARTICLE

Has local informational advantage disappeared?

Gennaro Bernile1

|

Alok Kumar1

|

Johan Sulaeman2

|

Qin Wang3

1Finance Department,University of Miami,

Miami, Florida

2Department of Finance,NUS Business

School,National University of Singapore,

Singapore City, Singapore

3Department of Finance,Spears School

of Business,Oklahoma State University -

Tulsa, Tulsa, Oklahoma

Correspondence

Johan Sulaeman, Department of Finance,

NUS Business School, National University

of Singapore, Singapore City, Singapore.

Email: sulaeman@nus.edu.sg

Abstract

This study examines how changes in the information environment affect the infor-

mational advantage of geographically proximate agents. The long- term advantage of

local agents disappeared at the turn of the millennium. This is accompanied by the

reduction in local bias of institutional investors and equity analysts. However, insti-

tutional investors continue to trade local stocks disproportionately more often than

non- local stocks; moreover, their local trades outperform non- local trades in the

short term—even for large and liquid stocks. Our results are consistent with im-

provements in the information environment shortening the horizon of proximity-

based informational advantage.

KEYWORDS

geography of information, information environment, informational advantage, local agents, local bias

|

39

BERNILE EtaL.

quarterly portfolio holdings of 13(f) insti tutions during the 1996–2008 period. We observe a sharp decline in the excess local

holdings of institutional investors in the more competitive information environment. In particular, we observe a discrete 50%

drop in the average excess local institutional ownership around the turn of the millennium. The average excess local ownership

is around 8% in the old regime (1996–1999), but only around 4% in the new regime after the turn of the millennium.4

We next investigate whether the information content of institutional investors’ local holdings also changes around these

exogenous shifts in information environment. A corollary of our main argument is that the performance of institutional inves-

tors’ local holdings would decline in the new regime. We find evidence consistent with this prediction. However, institutional

investors’ local holdings obtain higher returns relative to their non- local holdings in the earlier regime (consistent with previous

literature, such as Baik, Kang, & Kim, 2010), the superior performance of local holdings disappears in the later period. This

finding poses a difficult question for the literature on the geography of stock ownership, and particularly for studies that rely on

arguments based on disproportionate local holdings.5

We find similar patterns when examining equity analysts. Malloy (2005) documents that sell- side equity analysts have

relatively superior information about local firms using a sample that largely consists of pre- 2000 years. Consistent with the

evidence for institutional investors, we find that there is a significant decline in the (excess) local coverage of sell- side equity

analysts in the new regime after the turn of the millenium. Moreover, the reduction in analysts’ local bias is accompanied by

a significant reduction—in fact, a disappearance—of the local information advantage that sell- side equity analysts seemed to

enjoy in the old regime.

To shed more light on the time- series dynamics of the local information advantage and its horizon, we tap into the recent

studies that document the superior short- term performance of institutional investors and examine their actual intraquarter

trades. In particular, using trading data provided by ANcerno, Puckett and Yan (2011) document that institutional investors

earn superior intraquarter returns on their trades and that this performance is persistent in the cross- section of institutions.

However, the information content of institutional trades is incorporated quickly into prices as these trades do not predict re-

turns at longer horizons. Motivated by this evidence, we hypothesize that local investors continue to enjoy an informational

advantage in the short term—although this advantage may not exist at longer horizons as our earlier evidence based on hold-

ings data suggests.

We begin this analysis by examining institutional investors’ relative trading propensity among local and non- local stocks.

One unique advantage of the ANcerno trading data relative to 13(f) quarterly portfolio snapshots is that it allows a direct cal-

culation of excess trading levels within each quarter, which can be used to detect local trading bias. Our evidence indicates that

existing studies (as well as our earlier analysis) relying on low- frequency (i.e., quarterly) holdings likely underestimate the de-

gree of institutional investors’ local bias. In particular, institutional investors buy and sell local stocks disproportionately more

often than they trade non- local stocks—even when they do not hold more local stocks. We find that the abnormal local trading

propensity averages between 1% and 2.5% of the total institutional trading volume.

Institutional investors’ higher propensity to trade local stocks provides indirect evidence for the conjecture that these inves-

tors are likely to have access to short-term private information about local stocks. We next directly examine whether there is

a corresponding pattern in relative trading performance. To measure short- term trading performance, we follow the approach

used by Puckett and Yan (2011) and analyze trading profits in the quarter in which the trades are executed Extending Puckett

and Yan (2011), we examine institutional trades in local and non- local stocks separately, which allows us to decompose the total

intraquarter performance of institutional trading into its local and non- local components.6

The geography- based per formance decomposition indicates that the superior average intraquarter trading profits of insti-

tutional investors documented in Puckett and Yan (2011) are mostly driven by local trades. The institutional investors in our

sample earn superior intraquarter profits in local stocks, even after adjusting for common risk characteristics and transaction

costs. The average characteristic- adjusted intraquarter return of local stocks that institutions buy is higher than the post- trading

return of local stocks that they sell by 21–34 basis points. This superior local performance is observed even in relatively large

and liquid stocks. In contrast, the average non- local intraquarter trading profits are not significantly different from zero after

accounting for risk characteristics and transaction costs. These findings suggest that even after the epochal changes in firms’

information environment at the turn of the millennium, geographic proximity remains a key determinant of the short- term in-

formation advantage of a typical institutional investor.

Our analysis contributes to the intersection of three literatures: information environment, local bias, and fund performance.

First, we document that changes in the quality and competitiveness of the information environment affect the behavior of

market participants (investors and sell- side analysts) who rely on geographic proximity. Second, our results indicate that the

existing local bias literature underestimates the degree of local information advantage of institutional investors given its focus

on mandated quarterly snapshots of institutional portfolio holdings. With the availability of more efficient communication

technologies as well as the adoption of more stringent corporate disclosure rules, analyses of the local information advantage