Here’s an Opportunity: Knowledge Sharing amongCompetitors as a Response to Buy-in Uncertainty∗

Tristan L BotelhoYale School of Managementtristan.botelho@yale.edu

Cite as: Botelho, Tristan L. 2018. Here’s an Opportunity: Knowledge Sharing AmongCompetitors as a Response to Buy-in Uncertainty. Organization Science.

https://pubsonline.informs.org/doi/abs/10.1287/orsc.2018.1214

AbstractAlthough knowledge sharing among competitors is seemingly counterin-tuitive, scholars have found that competitors share knowledge under cer-tain conditions: among actors who have a pre-existing relationship andwho expect direct reciprocity. However, there are examples of knowl-edge sharing among competitors that cannot fully be explained usingthese relational mechanisms. In this study, I propose that in marketswhere competitors are a set of key stakeholders, knowledge sharing is astrategic response to high levels of buy-in uncertainty related to a po-tential opportunity, namely the likelihood that stakeholders will cometo realize the value of a potential opportunity in a timely fashion. Usinga unique dataset of knowledge sharing among investment professionalson a digital platform, this study leverages variation in the platform’sknowledge-sharing structure to test this theory. I find that knowledgesharing among these competitors is most likely when buy-in uncertaintyfor a given opportunity is high and that this knowledge sharing does leadto subsequent buy-in.

Keywords: competition, digital platforms, knowledge sharing, markets, performance, strat-egy, uncertainty

∗This work benefited from the comments of and discussions with Ezra Zuckerman, Roberto Fernandez, RayReagans, Mabel Abraham, Seth Carnahan, and seminar participants at MIT Sloan, Harvard Business School,the ASA Annual Meeting, and the AOM annual meeting. I would also like to acknowledge Gino Cattani andthe anonymous reviewers for improving the paper throughout the review process.

1

Introduction

Communities have long brought people together who have similar interests and passions with

the goal of creating and sharing knowledge, from the guild system of the Middle Ages to more

recent online communities such as Quora and Stack Overflow. Online communities virtually

connect individuals, allowing for larger and more diverse groups of people to come together

to exchange knowledge. Digital platforms are one popular type of online community that

promotes knowledge sharing among participants; this is achieved through various designs,

such as question and answer (e.g., Quora and Stack Overflow), evaluations and ratings (e.g.,

Rotten Tomatoes and Yelp), and free-form discussion (e.g., Facebook and Reddit).

Although most of these digital platforms are casual, firms have increasingly begun using

this emergent organizational form strategically to promote knowledge sharing within the

firm (Constant, Sproull, and Kiesler 1996; Hwang, Singh, and Argote 2015; Jeppesen and

Frederiksen 2006), and there are no signs of this trend slowing down (McKinsey & Company

2013). This appears to be a worthwhile knowledge strategy for firms, since knowledge plays

an integral role in improving firm outcomes (Argote 2012; Argote and Ingram 2000; Hansen

2002; Zander and Kogut 1995; Reagans and McEvily 2008) and digital platforms provide

an efficient means for knowledge exchange. The willingness of employees to engage in these

firm-hosted knowledge-sharing platforms is intuitive given their personal investment in their

firm (Constant, Sproull, and Kiesler 1996), and there is a long history of employees gathering

informally to share knowledge through “communities of practice,” for example (Wenger 1998;

Wenger, McDermott, and Snyder 2002). Given that this organizational form eliminates many

frictions related to knowledge sharing within firms, under what conditions may we also expect

these platforms to lead to knowledge sharing across competing firms?

For knowledge to yield a competitive advantage to a firm, it must be true that it cannot

easily be imitated or learned by a firm’s competitors (Lippman and Rumelt 1982). Therefore,

in addition to the general costs of knowledge sharing, such that it is time consuming and does

not guarantee a positive return ex ante, knowledge sharing among competitors introduces

a strategic cost: the potential loss of competitive advantage. Despite this cost, organization

and strategy scholars have highlighted that competitors do come together to share compet-

itive knowledge. Further, this research elucidates the conditions under which these costs of

knowledge sharing among competitors are minimized, thus facilitating such sharing (Ap-

2

pleyard 1996; Fauchart and von Hippel 2008; von Hippel 1987; Ingram and Roberts 2000;

Schrader 1991). Specifically, pre-existing relationships, expectations of direct reciprocity, and

slow-moving technology commonly accompany instances of knowledge sharing among com-

petitors. This reasoning is primarily relational, explaining the likelihood of knowledge sharing

occurring between a given dyad of competitors.

However, these explanations fall short of explaining more recent instances of knowledge

sharing that are much broader, whereby knowledge is shared with an audience via a common

platform (e.g. Towers Watson 2012; Value Investing Congress 2014b, 2014a). Here, those

sharing have little to no control over who has access to this knowledge making motivations

related to these relational mechanisms less likely. Moreover, in these contexts, technology

is often fast moving, allowing competitors to incorporate this shared knowledge quickly.

Thus, there is a gap in our understanding of the conditions that sustain competitors coming

together more broadly on a platform to engage in knowledge sharing.

I address this gap in our understanding by developing a theory related to a market-based

mechanism, which I call “buy-in uncertainty,” or uncertainty about the likelihood that key

stakeholders will come to realize the value of a potential opportunity in a timely fashion.

This is because, in most contexts, it is extremely difficult to objectively demonstrate an

opportunity’s potential value. For example, the expected future value of investing in a new

technology is challenging to discern. To reap the benefits of having identified a potential

opportunity in one’s market, actors often rely on key stakeholders’ realizing that this oppor-

tunity is worthwhile and buying in to—adopting, endorsing, or committing resources to—the

opportunity. Thus, I posit that in markets where competitors are a set of key stakeholders,

knowledge sharing is a strategic avenue for addressing this buy-in uncertainty, such that as

buy-in uncertainty increases so does the likelihood that an actor shares detailed knowledge

about an opportunity with their competitors. Because this need for buy-in helps alleviate

the potential costs associated with knowledge sharing, competitors are most apt to share in

these cases.

To test this theory, I use data from a digital knowledge-sharing platform, the Real In-

vestors Club (a pseudonym) for “buy-side” (e.g., hedge fund, mutual fund) investment profes-

sionals (e.g., analysts, portfolio managers). On this platform, investment professionals submit

investment recommendations (to buy or short sell a stock) about a potential investment op-

portunity they have identified (i.e., that a stock is under- or overpriced). An important

3

feature of this setting is that it provides a baseline against which to compare knowledge

sharing, as opposed to simply selecting on instances of knowledge sharing. When an invest-

ment professional chooses to submit a recommendation they must include a justification for

this position. The accompanying justification can be detailed, having a minimum of 600

words (averaging over 1,400 words), or simple, having a maximum of 40 words (averaging

24 words). This difference is critical because stating one’s buy or sell recommendation, after

their firm takes that position, does not constitute competitive knowledge sharing beyond

the often publicly available information, since portfolio holdings are commonly available to

others in this industry (e.g., Form 13F). However, recommendations with a detailed justifi-

cation provide competitive knowledge, namely elements from the analysis that supports the

recommendation, which competitors are typically not privy to. This analysis is a fundamen-

tal component of the value an investment professional brings to their firm and thus their

firm’s competitive advantage (Groysberg and Lee 2008).

I specifically focus on the relationship between buy-in uncertainty, measured as the level

of scrutiny and attention that a focal firm/stock1 faces from evaluative institutions in the

market (e.g., the media, sell-side analysts), and the likelihood that an investment profes-

sional shares knowledge in the form of their analysis through a detailed justification when

they submit a recommendation for an investment opportunity. The main contribution of

this research is enhancing our understanding of the dynamics of knowledge sharing among

competitors and the growing use of digital platforms by firms to improve their strategy

more broadly. While extant research has focused on the role of relational mechanisms in

supporting knowledge sharing among a pair of competitors, my theory focuses on the role

of a market-based mechanism in helping sustain much broader instances knowledge sharing

among competitors. Specifically, I show that competitors are more likely to broadly engage

in knowledge sharing with one another when there is considerable uncertainty with regards

to the likelihood that key stakeholders will similarly identify a given opportunity in a timely

fashion. The findings from this research also complement work that has focused on the co-

operative behavior among competing firms and the general motivations for contributing to

digital platforms.

1. A stock refers to the shares that are issued by a publicly traded firm. In most cases, I use the terms“stock” and “firm” interchangeably.

4

Knowledge Sharing within Firms

Knowledge, such as best practices, expertise, and firm routines, is a strategic resource for

firms. Developing an ability to effectively manage and transfer such knowledge within a

firm is a lucrative investment, as evidence suggests that it helps improve firm outcomes

(Argote 2012; Argote and Ingram 2000; Hansen 2002; Zander and Kogut 1995; Reagans and

McEvily 2008). Managers who have realized this fact have encouraged their employees to

come together to engage in knowledge sharing, such as through “communities of practice”

(e.g., Thompson 2005). These communities are normally informal and bring together like-

minded people within the firm who have similar expertise, interests, or passions, with the

goal of building, debating, and exchanging knowledge around an issue (Wenger 1998; Wenger,

McDermott, and Snyder 2002). For example, employees from across the firm may come

together to discuss a firm’s response to an emerging technology in its market. Leveraging

their experience and expertise, members can discuss different perspectives on how best to

embrace and respond to this strategic shift.

Over time, these within-firm knowledge-based communities have moved beyond in-person

gatherings, leading to a new organizational form using online communities that bring em-

ployees together to share knowledge (Constant, Sproull, and Kiesler 1996; Hwang, Singh, and

Argote 2015; Jeppesen and Frederiksen 2006). These digital platforms enable larger, often

global, sets of individuals to come together, creating an efficient medium for communication.

These platforms are more likely to assemble a diverse set of knowledge and disseminate it

throughout the firm instead of keeping it siloed, or local (Hwang, Singh, and Argote 2015).

Further, by institutionalizing knowledge sharing through a digital platform, firms can build

a repository of shared knowledge for future use. For these reasons, digital knowledge-sharing

platforms are aligned with a firm’s goal of maximizing their return to knowledge and are

increasingly being leveraged by firms. Results from a McKinsey & Company survey (2013),

for example, demonstrates that the number of firms using a digital platform to facilitate

knowledge sharing nearly doubled over a five-year period.

Knowledge Sharing across Firms

Although knowledge is a significant resource for firms, there are limitations to the amount of

novel knowledge that can be created and exchanged within a specific firm (e.g., Powell, Koput,

5

and Smith-Doerr 1996), with competitors serving as an important source for knowledge

outside of the firm. Recognizing this value, firms routinely watch their competitors, analyzing

any information that becomes available (Baum and Dahlin 2007; Haunschild and Sullivan

2002; Madsen and Desai 2010). Although there is evident value in a competitor’s knowledge, it

is not surprising that firms may hesitate to directly engage in knowledge sharing with their

competitors. An important task for managers is to preserve their competitive advantage

by ensuring that knowledge does not “leak” to competitors (Liebeskind 1996), especially

when this knowledge can be used (Lippman and Rumelt 1982). Thus, in addition to the

general costs of knowledge sharing—that it is time consuming and does not guarantee a

positive return ex ante—knowledge sharing among competitors introduces a strategic cost:

the potential loss of competitive advantage.

Even in the face of these costs, however, organization and strategy scholars have high-

lighted examples of firms in the same industry coming together to engage in knowledge

sharing. At one end of the spectrum, similar firms that operate in different geographic mar-

kets, and that therefore compete for different consumer bases, have been found to form

knowledge-sharing communities, openly sharing detailed knowledge with one another (e.g.,

Zuckerman and Sgourev 2006). In this case, the loss of competitive advantage is mitigated

by geographical distance, and these firms benefit from having their employees discuss and

refine their strategies and best practices.

At the other end of the spectrum, knowledge sharing occurring among direct competi-

tors has also been documented and is most common under conditions minimizing associated

costs. Specifically, knowledge sharing among two competitors has been found to occur when

there is a pre-existing relationship between them, when there is an expectation of direct reci-

procity, or when the competitors are in a slow-moving industry (Appleyard 1996; Fauchart

and von Hippel 2008; von Hippel 1987; Schrader 1991; Stein 2008). A pre-existing relationship

between those sharing fosters trust and provides a vehicle for social sanctioning, thereby re-

ducing informational frictions (Coleman 1988; Greif 1993; Ingram and Roberts 2000; Inkpen

and Tsang 2005; Stiglitz 1990). An expectation of direct reciprocity, or that actor A shares

with actor B because A expects that B will initiate similar sharing in the future, mitigates

the risk of uncertain return (Appleyard 1996; Fauchart and von Hippel 2008; von Hippel

1987; Schrader 1991; Stein 2008). In this way, direct reciprocity helps guarantee that no

single actor receives a disproportionate benefit relative to the cost they incur within a given

6

exchange relationship (i.e., it prevents a free-rider problem). Finally, a slow-moving industry

ensures the difficulty of implementing the knowledge shared, hence protecting the sharer’s

competitive advantage (Appleyard 1996).

The conditions identified in extant research focus primarily on relational factors leading to

knowledge sharing, helping to explain why a given dyad of competing actors would exchange

knowledge. But this reasoning does not fully account for the conditions that would sustain

broader, platform-based knowledge sharing among competitors where the exchange moves

from actor-to-actor to actor-to-audience. For example, the Value Investing Congress (VIC)

brings together value investors with the straightforward mission of “providing delegates with

immediately actionable investment recommendations...[and] helping attendees acquire the

wisdom they need to understand and profit in the often-irrational market” (Value Investing

Congress 2014b). An investment professional, reflecting on the conference, stated, “When I

attend the [VIC], I know that I will go home with a ton of great investment recommenda-

tions and some new ways of viewing value investing” (Value Investing Congress 2014a). Here,

an investment professional can attend and gain valuable strategic insights without having a

pre-existing relationship or being expected to directly reciprocate with those sharing their

knowledge. This platform-level knowledge sharing is fundamentally different from that char-

acterized and analyzed in extant research. Thus, it is unclear what conditions help motivate

competitors to broadly engage in detailed knowledge sharing.

In the following, I develop a theory related to how uncertainty in the likelihood that key

stakeholders will similarly identify a potential opportunity in a timely fashion helps sustain

this sharing.

Knowledge Sharing among Competitors as a Response to Buy-

in Uncertainty

Many factors affect the likelihood that a potential opportunity, such as an idea or innovation,

succeeds. However, one thing is certain: if the opportunity does not achieve a certain level

of buy-in from key stakeholders—adopting it, endorsing it, or committing resources to it—it

will fail. Even opportunities retrospectively noted as objectively high quality must receive

this buy-in. In 1999, investment professionals noted that dot-com stocks were overvalued and

that a bursting of this “bubble” was on the horizon (Brunnermeier and Nagel 2004, p. 2). The

7

NASDAQ Composite Index continued to climb, increasing by 108 percent between March

1999 and March 2000, before crashing: decreasing by 62 percent between March 2000 and

March 2001. Thus, even though some investment management firms correctly recognized an

opportunity to short sell related stocks, if they acted too soon they would have lost significant

capital. If they instead waited, closer to the point that buy-in from stakeholders about this

overpricing would be achieved, they would have been bountifully rewarded. The issue at play

here is what I term “buy-in uncertainty,” the likelihood that key stakeholders will come to

realize that an opportunity is not only present but also, conditional on recognizing it, that

they will act on it (i.e., buy-in) in a timely fashion.

Addressing buy-in uncertainty is important for firms because they are resource con-

strained, namely in terms of the amount of capital and time they can invest in hopes of

realizing a profit from a potential opportunity. Investment management firms—like firms in

other industries—cannot often profit from an opportunity without receiving the necessary

buy-in from other stakeholders. Returning to the dot-com bubble, investment management

firms had the ability to short sell dot-com stocks (Geczy, Musto, and Reed 2002), but few

firms—if any—had the resources necessary to weather massive losses while they waited in-

definitely for other stakeholders to agree that this opportunity was present. This buy-in

uncertainty stems from the fact that most markets are imperfect: resources do not neatly

flow to high-quality opportunities while avoiding low-quality opportunities along the way.

This uncertainty parallels a Knightian view of uncertainty, where the expectation of an

opportunity’s quality is difficult to know ex ante (Knight 1921). Although these imperfections

are necessary for profitable opportunities to exist (Barney 1986; Dierickx and Cool 1989),

they introduce significant obstacles for stakeholders. Therefore, once an actor has identified

a potential opportunity, they know a Keynesian Beauty Contest of sorts is occurring for

stakeholders.

Specifically, Keynes (1936) explains a scenario where newspaper readers were asked to

choose the six most attractive faces from a larger set, with the winner being the reader

whose list was closest to the most popular choices across all readers. Keynes argues that

the best strategy is a higher-order thinking approach: a reader should not choose based on

their own preferences, or even in line with what they perceive the preferences of others to

be, but instead based on what they expect that the average reader perceives the preferences

of others to be.

8

This stylized example helps demonstrate the type of evaluation process that stakeholders

undergo when analyzing which potential opportunities are worth their buy-in, affecting the

likelihood that a profit can be realized from a given opportunity. I posit that knowledge

sharing serves as a strategy to help alleviate this buy-in uncertainty by overtly focusing the

attention of key stakeholders on the merits of an opportunity that they may otherwise miss.

In markets where competitors also serve as key stakeholders, the potential benefits of this

knowledge sharing with competitors, when there are high levels of buy-in uncertainty for a

potential opportunity, help offset the potential costs related to loss of competitive advantage

from such sharing. Thus, I expect that as buy-in uncertainty for an opportunity increases,

so does the likelihood that an actor shares detailed knowledge about that opportunity with

their competitors.

Hypothesis: Conditional on identifying an opportunity, the likelihood that knowledge

sharing among competitors occurs increases as levels of buy-in uncertainty for that opportu-

nity increases.

Why Share with Competitors? Competitors as Key Stakeholders

Competitors often occupy a unique position as key stakeholders in their respective markets

in a number of ways. First, they are able to commit resources to an opportunity, such as

taking a similar position in the financial market, investing in a burgeoning technology, or

even helping to legitimize an industry category. Therefore, though counterintuitive, firms

may realize positive externalities from the presence of their competitors. Research has high-

lighted some of these positive externalities, such as common labor markets and social capital

arising from the geographical agglomeration of similar firms (e.g., Krugman 1991; Saxenian

1996; Sorenson and Audia 2000) and promoting self-interest through the use of board di-

rectorates and market-governance institutions (Burris 2005; Mizruchi 1992; Yue, Luo, and

Ingram 2013). Other research has provided numerous examples of competing firms acting

in a more “cooperative” fashion (e.g., Barnett and Carroll 1987; Gulati, Nohria, and Zaheer

2000; Hamel, Doz, and Prahalad 1989; Mowery, Oxley, and Silverman 1996; Navis and Glynn

2010). These competitors are not motivated by altruism; this is a strategic choice: to maxi-

mize the rents from a potential opportunity (Teece 1986; Williamson 1991), especially since

sole ownership of knowledge may not always lead to the optimal outcome (Carter 1989).

9

Navis and Glynn (2010) highlight this behavior in their analysis of the U.S. satellite radio

market. They show that during the nascent period of the satellite radio market, Sirius and

XM, the two major players in the market at the time, initially focused on legitimizing the

broader market category. It was not until this was achieved that they began to shift their

focus to between-firm differentiation.

Second, in terms of knowledge sharing, competitors offer the added benefit that they

can most easily understand the knowledge being shared. Knowledge is regularly difficult to

disseminate; much of it is complex and codified and therefore challenging for a non-expert to

understand (Hansen 1999; Teece 1977; Zander and Kogut 1995). Competitors have sufficient

absorptive capacity to grasp complex knowledge and assess its merits when it is presented to

them (cf. Cohen and Levinthal 1990). In markets where competitors can help solve buy-in

uncertainty by committing resources to an opportunity, they become an attractive set of

candidates to share this knowledge with to achieve this end.

Drawing on an example from the Value Investing Congress (VIC), an event geared towards

knowledge sharing among investment professionals, provides an illustration of the benefits of

strategic knowledge sharing with competitors in an effort to deal with buy-in uncertainty. At

the VIC in 2011, David Einhorn, founder and president of the hedge fund Greenlight Capital,

announced that he had taken a short position in Green Mountain Coffee Roasters’ (GMCR)

stock. He then spent one hour presenting 110 slides of rigorous analysis that supported this

opportunity. To this point, between the end of 2010 and Einhorn’s presentation, GMCR’s

value had been consistently increasing, with its stock price nearly tripling. Although Einhorn

had identified a market opportunity that he was confident in, as evidenced by his firm’s

substantial investment in the opportunity, there was considerable buy-in uncertainty for this

opportunity; specifically, it was unclear if others in the market would similarly recognize this

opportunity in a timely fashion. However, during his presentation the stock price began to

rapidly decrease and GMCR had lost about 10 percent of its total value that day (Comstock

2011). In line with my theory, Einhorn engaged in knowledge sharing with other investment

professionals (i.e., key stakeholders in the industry) to convince them that an opportunity was

present, and in doing so, he was able to successfully achieve buy-in. While effective, Einhorn’s

approach was not overly original. Researchers have long noted that there are benefits from

getting the word out about an investment opportunity you have invested in (Dow and Gorton

1994; Crawford, Gray, and Kern 2017; Ljungqvist and Qian 2016; Zuckerman 2012a, pp.

10

235-236). Furthermore, investment professionals have used “idea dinners,” where a small

group of friends get together to discuss investment opportunities with one another, for years

(Anderson 2005)–although unlike the VIC, these are closer to the dyadic knowledge sharing

based on existing relationships and norms of reciprocity.

The benefits derived from engaging in knowledge sharing with competitors to address

the buy-in uncertainty are not unique to financial markets. In 2014, Tesla Motors announced

that it “will not initiate patent lawsuits against anyone who, in good faith, wants to use

our technology” (Tesla Motors 2014). This was a marked difference from Tesla’s knowledge

strategy before that point: by 2014 Tesla had amassed more than 1,400 patents (Voyles

2014b). Initially, Tesla’s broad knowledge sharing with its competitors was seen as a “strategic

error” and some predicted that Tesla “would suffer as a result” (Voyles 2014b). However,

similar to the Einhorn example, Tesla realized a benefit from this knowledge sharing. At the

time, Tesla was facing high levels of buy-in uncertainty; it was unclear if key stakeholders

saw their innovations as legitimate. Soon after this announcement, discussions of strategic

partnerships between Tesla and its competitors began (Kaufman 2014; Voyles 2014a) and

Tesla’s stock price increased (Voyles 2014b). By engaging in this knowledge sharing, Tesla

motivated competitors to buy-in to this potential opportunity.

Overall, these examples and extant work may lead one to predict that broad knowledge

sharing with key stakeholders, such as competitors, should always follow the identification of

and investment in a potential opportunity. However, given the potential costs of knowledge

sharing, such as a potential loss of competitive advantage, the core contribution of this

theory is understanding the conditions under which knowledge sharing is more likely to

occur, specifically, how buy-in uncertainty motivates knowledge sharing among competitors.

Using a Digital Platform for Knowledge Sharing

A risk of moving competitive knowledge sharing from the dyad, as categorized in extant

competitive knowledge sharing research, to a digital platform is that the potential loss of

competitive advantage magnifies. However, a digital platform provides significant benefits for

knowledge sharing that help offset these risks. A main advantage of engaging in knowledge

sharing through a digital platform is capitalizing on the network effects commonly associ-

ated with such platforms (Evans 2003; Gawer 2011, 2014; Rochet and Tirole 2006). After a

critical mass of participants is active on the platform, the knowledge sharer benefits from

11

disseminating their knowledge to a wider audience. Given the posited motivation of buy-in

uncertainty, this helps increase the likelihood that enough stakeholders similarly recognize an

opportunity’s potential. For the knowledge seeker, organizing via a digital platform offers the

benefit of being exposed to a larger set of opportunities to assess in terms of their worthiness

of the knowledge seeker’s resources.

A second advantage is that digital platforms remove frictions often prohibiting in-person

community-based organizing, namely time and space. Virtually connecting allows actors

from across the globe and different positions within the social hierarchy to come together.

We would expect the diversity of information that one is exposed to on such a digital platform

to be greater than information received in person, from colleagues and friends, who are more

likely to possess redundant information (Burt 1995; Hansen 1999; McPherson, Smith-Lovin,

and Cook 2001; Reagans and Zuckerman 2008). Further, bringing together competitors helps

ensure that these actors have a similar expertise, which has been shown to help the flow of

knowledge and to decrease reliance on individual-level characteristics (Hwang, Singh, and

Argote 2015).

Empirical Context

The setting for this research was the Real Investors Club (RIC, a pseudonym), a private

digital platform that brings together buy-side (e.g., hedge fund, mutual fund) investment

professionals with the goal of openly sharing investment recommendations for individual

securities (e.g., common stock). Buy-side investment professionals analyze securities with

the goal of investing in them on behalf of their firm; sell-side analysts analyze securities with

the goal of disseminating their opinions to their bank’s clients. Prospective members of RIC

must apply for entry, and basic information about each investment professional, such as their

name and place of employment, is visible to other investment professionals on the platform.

The investment recommendations submitted on RIC focus on identifying current market

opportunities rather than on discussing the success or lack thereof of previous opportunities.

When an investment professional submits a recommendation for a stock, they must include

certain basic information: a recommendation; a price target, the price they expect the stock

to reach; and an investment horizon, the estimated time for this price target to be reached

(e.g., one year). They must also include a justification for this recommendation, which is

12

visible to all current and future investment professionals on the platform. The accompanying

justification can be detailed, thoroughly discussing the analysis leading to the recommenda-

tion, at least 600 words long (averaging over 1,400 words), or simple, a statement supporting

the recommendation, at most 40 words long (averaging 24 words). The content included

in a detailed justification is monitored by RIC to ensure a minimal level of quality and

rigor, whereas the content of the simple justification is not strictly monitored, other than

the 40-word limit.

To supplement the data from RIC, I conducted 21 semi-structured interviews with in-

vestment professionals, 12 of whom were members of RIC. These interviews provided more-

detailed information about the investment industry and the platform. When asked what led

them to join RIC or another knowledge-sharing platform in the industry, interviewees almost

always said that they wished to be part of a community of professional value investors. This

elucidates their desire to have a broad audience with which to share their opportunities and

knowledge. Interviewees also expressed gratitude to the community, stating that the invest-

ment recommendations on the platform affected their own view of their portfolio and their

investment strategy, and that the feedback they received on their recommendations helped

them hone their skills. Investment professionals noted that they discussed their desire to

join a knowledge-sharing platform with their firm’s leadership, suggesting that this was a

firm-level decision.

Those interviewed who were not part of at least one knowledge-sharing platform gave

two common reasons for their lack of participation. First, some specified that their firm

did not allow the analysis that led to their investments to be shared outside the firm (i.e.,

the content of a detailed justification). A Director of Research stated that it was impor-

tant to his firm to keep this type of information away from competing firms. Similarly, one

investment professional stated that they had to discontinue their use of knowledge-sharing

platforms when they changed employers. Second, some investment professionals viewed their

knowledge as too valuable to share. For example, a portfolio manager at a mutual fund

specified that he would not want to advance someone else’s career by allowing them to use

his analysis. However, similar to those surveyed by Shiller and Pound (1989), these invest-

ment professionals stated that they engaged in knowledge sharing with a select few friends

(i.e., pre-existing ties), where direct reciprocity was expected (see also Cohen, Frazzini, and

Malloy 2008; Duflo and Saez 2002, 2003; Hong, Kubik, and Stein 2004, 2005). This type

13

of relational knowledge sharing is also consistent with extant research documenting knowl-

edge sharing among competitors more broadly (Appleyard 1996; Fauchart and von Hippel

2008; von Hippel 1987; Ingram and Roberts 2000; Schrader 1991). One analyst described

that he preferred to share his “homework” (analysis) one-on-one only, and the idea of others

incorporating his investment process into theirs kept him from joining such platforms.

Data

The data for this study were all submitted investment recommendations for common stock–

as opposed to debt or options–listed on a U.S. exchange (e.g., NASDAQ and NYSE) between

2008 and 2013 on RIC. The sample includes 19,093 recommendations by 4,521 investment

professionals. Of this total, 4,026 recommendations (about 21 percent) were submitted with

a detailed justification. These data from RIC were supplemented with data about the stocks

featured in the recommendation. Financial market data came from the Center for Research

in Security Prices (CRSP); sell-side analyst coverage data came from I/B/E/S; institutional

ownership data came from the Thomson-Reuters Institutional Holdings (13F) Database; and

industry data came from Compustat.

To be included in the sample, the CRSP database had to cover the stock being recom-

mended at least on the day before the recommendation was submitted. CRSP covers major

U.S. exchanges, therefore, it does not include data on stocks that trade via the Over-the-

Counter (OTC) Bulletin Board. Most of the stocks on the OTC Bulletin Board are “penny

stocks,” which are characterized by their volatility and a lack of scrutiny by key evaluative

institutions in the market (see Ang, Shtauber, and Tetlock 1989, for a discussion), resulting

in a more conservative sample given the goals of this study.

Measuring Knowledge Sharing

The main dependent variable in this study was the indicator variable Knowledge Sharing,

which took the value of 1 for an investment recommendation submitted with a detailed jus-

tification and 0 if a simple justification was used. While some information is being shared in

simple justifications, it stands in stark contrast to the knowledge included in detailed justifi-

cations (Figure 1a and Figure 1b). Simple justifications are limited to 40 words, and average

only 24 words, which severely constrains the ability to convey any meaningful knowledge

about a given opportunity. Moreover, these simple justifications offer tenuous insight into

14

the analysis, a key source of competitive advantage, which led to the specific recommendation

(Figure 1a). It could be argued that gauging another investment professional’s absolute sen-

timent about a stock (buy vs. sell) is useful; however, much these data are publicly available.

For example, the Securities and Exchange Commission (SEC) requires all mutual funds to

report their complete list of holdings each quarter (through forms N-Q and N-CSR), and the

SEC requires other institutional investors (e.g., hedge funds) with more than $100 million in

equity assets under management to report their holdings quarterly via Form 13F.

[Figure 1a]

[Figure 1b]

Detailed justifications, unlike simple justifications, offer valuable insights into an in-

vestment professional’s thought process and analysis that supports a recommendation. The

content of detailed justifications commonly includes an analysis of supporting information,

gleaned from meetings with management (and investor calls), recent news, and company

reports; comparable companies, such as competitors; macroeconomic and industry trends;

and the valuation (Figure 1b). The ability to substantiate one’s potential investment oppor-

tunities with a rigorous analysis is a key source of the value that an investment professional

adds to their firm, hence, it is also a main source of their firm’s competitive advantage. One

portfolio manager made this distinction clear in describing his Chief Investment Officer’s re-

sponse to his having accepted an invitation to lecture at a local business school’s Investments

course: “He told me that I couldn’t use any material that showed analysis. . . [laughs]. . . I was

like ‘what the hell am I going to do then?’ When I told [the professor] we both agreed it was

better for me to skip [the lecture].” While the CIO did not prohibit the portfolio manager

from giving the lecture altogether, he made clear that details of the analyses were off-limits.

To further elucidate why a detailed justification represents knowledge sharing and a

simple justification does not, let us return to the portfolio manager from above who refused to

share knowledge because he did not want someone else using his analysis. He was not opposed

to others knowing the stocks he was invested in—his portfolio is a matter of public record

(his firm publicly published their investment positions). What he wished to safeguard was the

analysis that led to his identifying a given market opportunity—the knowledge contained in

a detailed justification. When asked about the differences between using the simple and the

detailed justifications, an investment professional said, “I see them as two completely different15

vehicles; [the detailed justification] lets me fix the market by sharing my due diligence with

the community while [the simple justification] lets me make a call.” The “fix[ing]” of the

market speaks directly to the strategic use of knowledge sharing: an investment professional

who finds an opportunity but who faces high levels of buy-in uncertainty is motivated to share

knowledge with competitors to combat this uncertainty and help increase the likelihood that

their opportunity receives buy-in. This “fix[ing]” reflects the investment professional’s belief

that they have identified a market opportunity rather than a desire to make the market

efficient. The motivation for using a simple justification was less clear. Those who were

interviewed agreed with the above quote, regarding simple justifications as making a call. In

other words, they valued being able to prove that they had identified an opportunity. This

suggests that investment professionals used simple justifications when they perceived lower

levels of buy-in uncertainty about a given opportunity. Importantly, the platform does not

guide investment professionals to use one type of justification over another—members are

free to do as they please.

Therefore, a detailed justification, relative to a simple justification, offers a rigorous level

of both qualitative and quantitative knowledge. Interviewees often mentioned that these jus-

tifications took months to research and hours to create. Further, one investment professional

stated that the detailed justifications he submitted to RIC were identical to the proposals he

submitted internally to his portfolio manager. Together, all of these factors indicate that the

knowledge included in a detailed justification is much closer to firm-level strategic knowledge

(Grant 1996; Levitt and March 1988).

An alternative explanation to the hypothesized relationship between knowledge sharing

and buy-in uncertainty is that those using the simple justification are simply free riding.

Specifically, investment professionals want access to other investment professionals’ anal-

yses; therefore, they submit a recommendation with a simple justification to gain access

to the full database of recommendations using a detailed justification. However, this type

of free riding is not possible on RIC. To gain access to the database of recommendations

with detailed justifications, one must actively contribute such recommendations. Further, as

mentioned above, RIC checks the quality of detailed justifications to help ensure that invest-

ment professionals are not submitting low-quality analyses, therefore, safeguarding against

recommendations that may be submitted in order to gain access to the analysis of others.

The ability to leverage variation in knowledge sharing within the same context is im-

16

portant methodologically. Previous research has often focused on cases where knowledge

sharing occurred and then highlighted the conditions present in those instances, or relied

on self-reports of knowledge-sharing activities (e.g., Appleyard 1996; Fauchart and von Hip-

pel 2008; von Hippel 1987; Ingram and Roberts 2000; Schrader 1991). While this research

has shed important light on the phenomenon of knowledge sharing among competitors, this

method of analysis introduces the possibility of measurement error. My context, in contrast,

provides a unique opportunity to isolate the conditions leading investment professionals to

share knowledge through recommendations with detailed justification by comparing these in-

stances to an appropriate baseline: simple justifications where knowledge is not being shared

(see Fernandez and Sosa 2005, for a similar discussion related to labor market research).

Measuring Buy-in Uncertainty

Buy-in uncertainty is the likelihood that key stakeholders will come to realize the value

of a potential opportunity in a timely fashion. In this context, an opportunity is when an

investment professional perceives a stock to be underpriced (overpriced): the current price

of the stock is less (greater) than its current value. Although buy-in uncertainty plays a role

in most markets, the financial market context offers a chance to measure it more easily. If

an investment professional finds an opportunity, it is not the case that they can just buy

(or sell) that stock; they need others in the market to similarly realize that this opportunity

is present, or to buy-in (Zajac and Westphal 2004; Zuckerman 2012b). Buy-in uncertainty

was captured using multiple measures related to the level of scrutiny and attention that a

firm faced from key evaluative institutions in the market, which has been shown to affect the

price–value relationship for a stock (Boehmer and Kelley 2009; Fang and Peress 2009; Zhang

2006). The measures Firm Age, Sell-Side Coverage, Institutional Ownership Concentration,

and Media Attention were used to capture the level of buy-in uncertainty for a given stock.

The goal of this approach is to not overemphasize the coefficient of any one measure but

instead to interpret the results collectively.

Firm Age was calculated as the difference between the year the investment recommen-

dation was submitted and the initial year the stock was covered in the CRSP database

(most often the firm’s IPO year), plus 1. Firm Age approximates the availability of histor-

ical data—an overall sense of information availability—for a firm. As time passes there is

more information and certainty about a firm’s strategy, leadership, and performance. Addi-

17

tionally, given that these are public firms, “older” firms have also submitted more financial

documentation (e.g., quarterly reports) to the SEC that has been scrutinized and evaluated.

Sell-Side Coverage was calculated as the sum of the number of unique earnings estimates

for the stock featured in the investment recommendation, in each of the four quarters prior

to that recommendation. When a firm becomes publicly traded on a U.S. exchange, sell-

side analysts may choose to initiate coverage of the firm. These evaluators have been found

to wield influence with the firm they cover, even affecting a firm’s management practices

(e.g., Benner and Ranganathan 2012; Rao and Sivakumar 1999; Zuckerman 2000). Sell-

side analysts will issue periodic reports about the firm. These reports routinely include

an evaluation of historical information, industry outlooks, and earnings estimates, as well

as other analyses. As this coverage increases, so does the scrutiny and evaluation of the

firm, and information about the firm reaches a wider audience. Greater coverage has been

found to be directly related to a firm’s information availability and to increases in price for

underpriced stocks (e.g., Bushman, Piotroski, and Smith 2005; Bushman and Smith 2001;

Francis, Douglas Hanna, and Philbrick 1997; Lang and Lundholm 1996), and this increased

level of scrutiny and evaluation leads to beneficial outcomes in terms of corporate governance

(Yu 2008). For this measure, if three unique analysts covered a stock in each of the last four

quarters (3 × 4 = 12), and two other analysts covered the stock, but for only two of the last

four quarters (2×2 = 4), Sell-Side Coverage would take the value of 16 (12+4). If estimates

were updated within a given quarter, Sell-Side Coverage was not changed.

Institutional ownership was calculated as the ratio of the number of shares of a firm’s

stock owned by institutional investors that file Form 13F to the total number of shares out-

standing. Institutional investors may include banks, hedge funds, mutual funds, pensions, and

endowments. These are investment professionals who manage money for their clients. Given

their large pool of capital for investment, institutional investors can take large investment po-

sitions in a firm, which gives them substantial influence over management practices, such as

aligning compensation with shareholder expectations (e.g., Connelly et al. 2010; Parthiban,

Kochhar, and Levitas 1998) and promoting long-term innovation efforts (e.g., Kochhar and

Parthiban 1998). The number of institutional investors, as well as the number of shares an

institutional investor owns of a firm, varies greatly; therefore, recent research has focused on

the concentration of institutional ownership. Empirical evidence suggests that when insti-

tutional ownership is concentrated among fewer institutional investors a stock’s price–value

18

relationship is affected (Boehmer and Kelley 2009). Following this research, I dichotomized

institutional ownership2 into two separate variables, Institutional Investor Concentration

(Percent, Top 5) and Institutional Investor Concentration (Percent, Other), for each firm in

the quarter preceding the investment recommendation.

To better understand the construction of these two variables, let us consider, for example,

Firm A that has a total of 100 shares outstanding and seven institutional investors, with

the sum of the shares they each own being 80. First, a rank order is created of how many

shares of Firm A’s stock are owned by each of the seven institutional investors. Institutional

Investor Concentration (Percent, Top 5) is calculated as the ratio of the sum of shares

of Firm A’s stock that are owned by each of the five (of the seven) largest institutional

investors to the total number of shares of Firm A’s stock outstanding (here, 100). Institutional

Investor Concentration (Percent, Other) is calculated as the ratio of the total number of

shares of Firm A’s stock that are owned by the remaining institutional investors (here,

two) to the total number of shares of Firm A’s stock outstanding (here, 100). Therefore,

if the five largest institutional investors collectively owned 60 shares, Institutional Investor

Concentration (Percent, Top 5) would take the value of 0.60 (60 / 100), whereas Institutional

Investor Concentration (Percent, Other) would take the value of 0.20 ((80 - 60) / 100). For

stocks that had less than five institutional investors, Institutional Investor Concentration

(Percent, Other) took the value of 0.

Media Attention was measured as the number of articles in which a firm was discussed

in the month preceding the investment recommendation, plus 1. These counts were hand-

collected from a leading website focused on financial markets that aggregates and publishes

news about stocks. A greater amount of media attention implies that a firm faces a higher

level of scrutiny and attention. Similarly, media attention has been found to affect a firm’s

stock price, even when no genuine news is supplied (Fang and Peress 2009). Further, sub-

stantial (even unrelated) media attention in a previous period has been found to attenuate

the effect of negative future events, such as protests, on stock returns (King and Soule 2007).

For these measures, besides the institutional ownership concentration variables, buy-in

uncertainty decreases as the value of these measures increases. Therefore, buy-in uncertainty

is maximized when the value of these measures, X, is minimized. For consistency with the

2. There are instances in which institutional ownership is reported as exceeding 100 percent. As discussedby Asquith and colleagues (2005), there are legitimate reasons for this occurring. In this sample, institutionalownership was capped at 100 percent. Results are robust to the removal of these observations.

19

hypothesized relationship, the reciprocal of these measures, (1/X), was used. To more con-

cretely illustrate this transformation, consider the variable Firm Age. During a firm’s first

year, Firm Age is equal to 1 (0 + 1), with its reciprocal being 1/1, and after four more years

Firm Age for this same firm is equal to 5 (4 + 1), with its reciprocal being 1/5. Since buy-in

uncertainty should be higher, ceteris paribus, for smaller values of Firm Age, we would expect

the values of Firm Age used in the analyses to correspond to this notion, that is, 1/1 > 1/5.

Control Variables

Variables at the investment-professional and investment-recommendation level were used as

controls. At the investment-professional level, these measures included education, using a

ranking of both undergraduate and graduate institution, and the investment professional’s

physical location. At the investment-recommendation level, investment horizon, firm size,

industry fixed effects, and year fixed effects were included.

For undergraduate education, the 2013 U.S. News College Ranking (U.S. News and World

Report 2014b) was used to match an investment professional’s undergraduate institution to

its ranking. This was also done for graduate education. For U.S. business schools, the 2013

U.S. News MBA Ranking was used (U.S. News and World Report 2014a), and for non-U.S.

business schools, the 2013 Financial Times Global MBA Ranking (Financial Times 2014)

was used. Investment professionals’ education was grouped into two variables, Elite Under-

graduate and Elite Graduate, if they attended a school ranked 1−20, respectively. Education

quality was controlled for because certain institutions may train investment professionals to

act in a certain way with regard to sharing knowledge. Given that an investment profes-

sional’s city has been found to affect their investment choices (e.g., Hong, Lim, and Stein

2000) and available resources, location was included as a control. Major City represents

large metropolitan cities in the U.S. that are often thought of as financial hubs (e.g., Boston,

Chicago, New York City, and San Francisco) and took the value of 1 for all investment pro-

fessionals working in these cities. Additionally, the indicator variable Non-US took the value

of 1 for all investment professionals located in a city outside the U.S.

At the investment-recommendation level, certain firm sizes, investment horizons, indus-

tries, or time periods may be more suitable for knowledge sharing. Firm Size was calculated

as the market capitalization (share price × shares outstanding) of the stock featured in the

investment recommendation (in billions) on the day prior to the recommendation being sub-

20

mitted, using the shares outstanding reported in the previous quarter. Although firm size

may be correlated with the level of scrutiny and attention that a firm receives, it also cap-

tures many other factors, making it a necessary control in all models. Similar to the above

reasoning, the reciprocal of Firm Size was used in the analyses. An investment professional’s

investment horizon may affect their likelihood to share knowledge; therefore, the indicator

variable for recommendations of Short Investment Horizon took the value of 1 if the invest-

ment professional has an investment horizon of under one year. Table 1 provides summary

statistics for each of the key variables, separated by justification type, and Table 2 provides

correlations.

[Table 1]

[Table 2]

Empirical Model

I estimated the following logit regression to evaluate the main hypothesis:

Knowledge_Sharingi = βMeasures of Buy-in Uncertaintyi + γXi + δi + λi + εi,

where i indexes the investment recommendation, the unit of analysis. Xi is a vector of

recommender- and recommendation-level controls; δi is a fixed effect for the industry of the

stock being recommended and includes 24 two-digit North American Industry Classification

System (NAICS) sectors, and an indicator for a missing NAICS sector; and λi is a fixed effect

for the year the recommendation was submitted. Robust standard errors were clustered at

the investment-professional level given the possibility that the choice of sharing knowledge

and of which types of stock to recommend may be correlated within investment professional.

Results

Buy-in Uncertainty and Likelihood of Knowledge Sharing

Figure 2 presents initial descriptive evidence that investment professionals were more likely

to share knowledge (i.e., use a detailed justification) when recommending stocks that were

more likely to have higher levels of buy-in uncertainty (i.e., smaller firms). The average mar-

ket capitalization of a stock recommended with a detailed justification was approximately21

$6.8 billion, almost one-third the size of the average market capitalization of the stocks listed

on the Standard & Poor’s 500 (S&P 500; $19.1 billion, at the midpoint of the period under

study). Because this context includes both the simple and detailed justification, I am able to

distinguish between my hypothesized explanation, that investment professionals share knowl-

edge in response to high levels of buy-in uncertainty, and the alternative explanation that

investment professionals who share knowledge are more likely to prefer stocks with higher

levels of buy-in uncertainty. Comparing the market capitalization of the firms discussed

across the two justification types—detailed and simple—begins to adjudicate between these

alternatives. From Figure 2, it is evident that there is a strong correlation between a firm’s

size and the likelihood that knowledge sharing occured. The average market capitalization of

stocks recommended with a simple justification was about $18.4 billion, similar to that of the

S&P 500, and 2.7 times larger than that of stocks recommended with a detailed justification

(p < 0.001).

[Figure 2]

To more rigorously test my main hypothesis, I used measures for buy-in uncertainty,

related to the level of scrutiny and attention that a focal firm faced—namely Institutional

Investor Concentration, and the reciprocal of Sell-Side Coverage, Firm Age, and Media At-

tention. The results from these analyses (Table 3) support the main hypothesis, suggesting

that engaging in knowledge sharing about an investment opportunity was more likely when

an opportunity had high levels of buy-in uncertainty. The odds that an investment profes-

sional recommended a stock using a detailed justification were approximately 22 percent

higher when the firm’s Sell-Side Coverage was one standard deviation (44.99) below the

mean (60.93) than at the mean, in other words, when fewer sell-side analysts were covering

the stock, 1.307 (exp[(1/15.94)×4.267]) versus 1.073 (exp[(1/60.93)×4.267]) (Table 3, M2).

Similarly, the coefficient of Firm Age (Table 3, M3) suggests that an investment recommen-

dation for a firm’s stock that more recently IPO’ed was more likely to include a detailed

justification than a recommendation for a stock that had been public for several years. Sim-

ilarly, the odds of knowledge sharing were 42 percent higher when a firm received the mean

(1.65) amount of media attention than when a firm received media attention that was one

standard deviation (7.18) above the mean (Table 3, M4), 1.534 (exp[(1/1.65)×0.706]) versus

1.083 (exp[(1/8.83)×0.706]).

22

Buy-in uncertainty measured as the concentration of institutional investors yielded con-

sistent results, with Investor Concentration (Percent, Top 5) strongly predicting the odds

of knowledge sharing (Table 3, M5). The log odds of knowledge sharing were higher when

there was a greater concentration of ownership among five or fewer institutional investors

for a given stock. This effect was moderated as concentration increased outside this small

set of institutional investors, Investor Concentration (Percent, Other), meaning institutional

ownership was more dispersed. Overall, these results offer strong empirical support for the

main hypothesis: that, conditional on identifying an opportunity, an investment professional

is more likely to share knowledge with their competitors when an opportunity they have

identified has higher levels of buy-in uncertainty. Figure 3 presents the marginal effects from

the regressions in Table 3.

[Table 3]

[Figure 3]

In subsequent analyses, I introduced the measure Short Recommendation to the above

models. Short Recommendation is a dichotomous variable that takes the value of 1 if the

investment recommendation was to short sell the stock and a value of 0 if the recommendation

was to buy-and-hold the stock. Short selling is a feature of the U.S. stock market, and its

presence offers a unique empirical opportunity to further test the main hypothesis, because

the costs of not achieving buy-in are much greater for a short selling opportunity (see the

dot-com bubble discussion above) than for a buy-and-hold opportunity. Short selling allows

an investor who is pessimistic about the current price of a given stock (i.e., believes it to be

overvalued) to borrow shares of a stock that they do not own, for a fee, and sell them back

to the market. The investor agrees to return the shares at a later date, along with interest,

and any distributions (e.g., dividends) that occur during the borrowing period. A short

seller profits when a stock price decreases relative to when they borrowed the shares. While

the availability of short selling is an important mechanism for an efficient market (Asquith,

Pathak, and Ritter 2005; Curtis and Fargher 2014), it is frequently criticized because the

need for buy-in is especially high relative to buying a stock (Abreu and Brunnermeier 2003;

Brunnermeier and Nagel 2004; Zuckerman 2012b). Unlike when buying a stock, an investor

who short sells a stock does not own part of the firm and faces the possibility of unbounded

loss—there is no limit on the upward movement of a stock’s price.23

Following from the main hypothesis, we would expect that the high level of buy-in un-

certainty surrounding short selling makes it more likely that these recommendations include

knowledge sharing relative to buy-and-hold recommendations. The evidence across all model

specifications (Table 4) supports this relationship; the log odds of knowledge sharing are sig-

nificantly higher for short-selling recommendations than for buy-and-hold recommendations.

On average, the odds that an investment professional engaged in knowledge sharing about

an investment opportunity are between 1.675 (exp[0.516]) to 1.802 (exp[0.589]) times higher

when a recommendation was to short sell a stock relative to buy-and-hold a stock.

[Table 4]

Overall, I find strong evidence that competitors incur the costs of knowledge sharing as

buy-in uncertainty increases for an investment opportunity they have identified.

Robustness Checks

A possible alternative explanation is that these investment professionals needed to provide

more detail for the firms facing a lower level of scrutiny and attention from key evaluative

institutions in the market because such firms are obscure and insignificant in the economy.

While this does not invalidate the main hypothesis, it is worth addressing. An initial safe-

guard against this concern is that the data exclude stocks that trade via the OTC Bulletin

Board, or “penny stocks.” Further, although on average investment recommendations with a

detailed justifications feature smaller firms, these are still substantial firms in the economy.

The average market capitalization of firms featured in a detailed justification is $6.8 billion

(median $0.94 billion), which is larger than the average publicly traded firm in the U.S. ($5.3

billion, median $0.56 billion).3 Further, since all firms in my sample are publicly traded, they

file financial statements and other documents with the SEC. Moreover, the average firm rec-

ommended with knowledge sharing in my sample had about 50 (median of 40) unique sell-side

estimates in the sum of the four quarters prior to the recommendation. Therefore, these are

not obscure firms but firms that at the time of recommendation had higher levels of buy-in

uncertainty than firms recommended with simple justifications. This alternative also fails

to explain the following: what is the goal of this knowledge sharing? What these investment

professionals are doing is sharing knowledge by providing a rigorous analysis that supports

3. Data from Wilshire 5000 Total Market Index Characteristic Sheet (June 30, 2014).

24

a potential opportunity they have identified in an effort to motivate other stakeholders to

buy in, thereby increasing the likelihood that the opportunity will yield a profit. Conversely,

opportunities still exist when there are lower levels of buy-in uncertainty; however, other

stakeholders are more likely to organically realize these opportunities in a timely fashion,

making it less likely that investment professionals opt to incur the costs associated with

knowledge sharing for those potential opportunities.

There are also important challenges related to selection that must be addressed. While

simple justifications present an important baseline for analysis, it is still possible that selec-

tion into knowledge sharing is affecting the above results. For example, it is possible that

buy-in uncertainty is correlated with a certain type of investment professional or that the

above results are driven by unobserved investment-professional heterogeneity, such as dif-

ferences in their prior training or career motivations. To address this concern, I identified

investment professionals who submitted at least one investment recommendation with a de-

tailed justification and at least one recommendation with a simple justification. Individual

fixed effects were then included in all models, allowing for a within-investment-professional

analysis. The analyses presented in Table 3 were replicated in Table 5 and the analyses

presented in Table 4 were replicated in Table 6. These results are robust, with evidence

still strongly suggesting that investment professionals are more likely to bear the costs of

knowledge sharing with their competitors when there are high levels of buy-in uncertainty

surrounding an investment opportunity.

A “spaghetti against the wall” or feedback alternative may also be at play in the con-

text. Investment professionals may share knowledge about many investment opportunities

in order to gauge the response from these key stakeholders. They could then commit their

firm’s capital to the opportunities that were seen as most favorable. First and foremost,

the overwhelming majority of investment professionals on RIC are value oriented, so such a

scheme seems unlikely. Supporting this counterargument, the average investment professional

only shares knowledge about two to three investment opportunities during their tenure. If

this alternative was true, I would expect this to be a much larger number. To rule out this

alternative more rigorously, I was able to collect data on whether the firm the investment

professional submitting an investment recommendation works for had taken a position in the

stock prior to the recommendation occurring. I find that firms had taken a position prior to

25

the recommendation in an overwhelming proportion of cases (81 percent4) helping rule out a

feedback alternative. This statistic also provides further evidence of the posited mechanism

of buy-in uncertainty. These investment professionals are sharing knowledge after they have

committed to the opportunity.

[Table 5]

[Table 6]

Another alternative explanation for engaging in knowledge sharing relates to career stage.

Junior investment professionals may be motivated to establish a reputation, namely a track

record of outperformance, apart from their firm. An initial counterargument to this alter-

native is that these investment professionals could build their reputation by submitting

recommendations with a simple justification, which would allow them to have a history of

their performance. Empirically, Table 5 and Table 6 include investment professional and year

fixed effects, helping rule out that investment professionals are changing their own sharing

behavior over time, however, it is possible that many investment professionals stay on the

platform for a shorter period of time and thus the use of these fixed effects may not fully

address this career stage alternative.

To address this, I was able to collect title information for 739 investment professionals in

my sample. This accounted for 5,187 recommendations (27 percent) of the sample, 40 percent

of which included knowledge sharing. I coded the titles analyst and associate as junior and the

title portfolio manager, as well as executive titles (e.g., Founder, CIO), as senior, using the

dichotomous variable Senior Investment Professional.5 To first test the validity of this coding

scheme, I analyzed the likelihood that the investment professional’s firm had a position in the

stock before it was recommended as a function of their seniority. We should expect firms to be

more likely to commit firm resources to the investment recommendations of senior members

of the firm, relative to more junior members of the firm. I find that for recommendations made

by senior investment professionals their firm took a position in that security 90 percent of the

time as compared to 77 percent for recommendations made by junior investment professionals

(p < 0.001), which helps validate this dichotomous measure. I then regressed the likelihood

4. This statistic was only available for recommendations with a detailed justification, and all results arerobust to dropping detailed recommendations where the firm does not own the stock prior to the recommen-dation.

5. These results are consistent if associate is treated as a senior position.

26

of sharing knowledge on the investment professional seniority. Counter to a career stage

alternative, I found that there is a positive correlation between being senior and sharing

knowledge but this relationship is not statistically significant (Table 7, R9). Furthermore,

I find evidence that junior and senior investment professionals are similarly more likely to

engage in knowledge sharing as buy-in uncertainty increases (Table 7, R10-R13). Therefore,

it is unlikely that an alternative explanation regarding career mobility is driving knowledge

sharing in this context.

[Table 7]

Does Knowledge Sharing Lead to Buy-in?

The evidence presented thus far strongly suggests that knowledge sharing among competitors

is most likely when a potential opportunity faces high levels of buy-in uncertainty. It is

next important to understand whether this goal is achieved. In other words, in order for

organizations to consider this knowledge strategy, it would be beneficial to know whether

knowledge sharing leads to buy-in in my context. To answer this question, I collected stock

return data around the time an investment opportunity was recommended to analyze how

stock return changes in the period right before and after knowledge sharing occurs. The

dependent variable in this analysis is the cumulative abnormal return (CAR) of a stock

recommended using knowledge sharing, a common methodology for assessing whether an

event (here, knowledge sharing) affects a stock’s price to a greater degree than the market

would predict (e.g., Brown and Warner 1985; King and Soule 2007; Zajac and Westphal

2004).

For each stock i featured in a recommendation with knowledge sharing (a detailed justi-

fication), I calculated its abnormal returns (AR):

ARit = Rit − E(Rit),

where ARit is the abnormal return for stock i in period t, Rit is the actual return of stock i in

period t, and E(Rit) is the expected return for stock i in period t. I estimated this expected

return using an ordinary least squares regression of stock i’s return on the market’s return

27

(Rmt), here, the value weighted index of all stocks in the CRSP database:

E(Rit) = αi + βiRmt,

where αi is the intercept term of the regression (stock i’s return when Rmt is 0, comparable

to the risk-free rate) and βi is a constant term from the regression (the systematic risk, or

beta, of stock i). The CAR is then the sum of stock i’s abnormal returns over an event

window (t to T ):

CARi =T∑1ARit

I defined the event window as 21 days (-10, 10), the 10 days prior to the submission of a

detailed justification for a stock to the 10 days after, with the day of submission being day

0. The regression coefficients for expected return were estimated using a 239-day window,

prior to the start of the event window (King and Soule 2007; Zajac and Westphal 2004). I

find that the cumulative average abnormal return (CAAR) is positively correlated with the

investment recommendation around the time the recommendation with knowledge sharing

occurs: for buy-and-hold recommendations the CAAR is +1.19 percent (p < 0.001) and for

short-sell recommendations the CAAR is -4.19 percent (p < 0.001). Figure 4 plots the CAAR

for the event window, where a sharp change in CAAR can be seen around the day that the

recommendation is submitted. These results are robust to use of other specifications for

the event window, including an immediate window around the knowledge-sharing event (-1,

1). Using this specification, the CAAR for buy-and-hold recommendations is +0.43 percent

(p < 0.001) and the CAAR for short-sell recommendations is -1.24 percent (p < 0.001).

These patterns strongly suggest that submitting a recommendation with detailed knowledge

sharing leads to buy-in from key stakeholders.

This finding is also in line with comments from investment professionals that I inter-

viewed who referred to RIC as a resource to source investment opportunities for their fund.

One hedge fund founder discussed that though he managed a significant amount of capi-

tal, his team was small, and thus his firm often used RIC as a resource for beginning their

due diligence process. Furthermore, for recommendations submitted with a detailed justifi-

cation, RIC members can provide free-form comments. Although a systematic analysis was

not undertaken, most of the commentary were questions that helped clarify the submitted

knowledge, which suggests that opportunity seekers are trying to further analyze the quality

28

of an opportunity.

[Figure 4]

These results, in sum, show that investment professionals on RIC were more likely to

engage in knowledge sharing when there were high levels of buy-in uncertainty surrounding

an investment opportunity they have identified. Further evidence suggests that, on average,

this knowledge sharing did lead to the investment opportunity receiving buy-in.

Discussion

Given the integral role of knowledge in affecting firm outcomes (Argote 2012, 2012; Hansen

2002; Reagans and McEvily 2008; Zander and Kogut 1995), firms take their knowledge strat-

egy seriously (Brown and Duguid 2001; Zack 1999). In fact, firms are increasingly using a

new organizational form to improve their knowledge-sharing effort within their firm, namely

digital platforms (Hwang, Singh, and Argote 2015; McKinsey & Company 2013). Firms have

embraced these digital platforms because they mimic the benefits of in-person communica-

tion, such as communities of practice (Wenger 1998; Wenger, McDermott, and Snyder 2002),

while removing many of the related frictions, such as time and space constraints. When we

consider the promise of these platforms for promoting knowledge sharing across firms, prima

facie the prospects are much dimmer. Extant research would lead us to believe that such

broad sharing would only magnify the costs of knowledge sharing among competitors, specif-

ically loss of competitive advantage (Appleyard 1996; Fauchart and von Hippel 2008; von

Hippel 1987; Schrader 1991). This research has primarily highlighted that knowledge sharing

among competitors is most likely through relational mechanisms, such as pre-existing rela-

tionships and expectations of direct reciprocity, where this cost is best minimized. However,

there are cases in the economy of competitors broadly sharing knowledge with one another

that cannot be explained by these relational mechanisms. In this paper, I develop a theory

that helps reconcile this disconnect and deepen our understanding of the conditions under

which knowledge sharing among competitors can be sustained. To test my theory, I use data

from a digital knowledge-sharing platform that brings investment professionals together from

around the globe.

This study contributes to research on knowledge sharing among competitors (Appleyard

1996; Fauchart and von Hippel 2008; von Hippel 1987; Schrader 1991) by uncovering how29

a new broader mechanism—which I call “buy-in uncertainty”—motivates competitors to

engage in broader knowledge sharing with one another. It seems intuitive that actors are pri-

marily concerned with discovering an objectively high-quality opportunity; in most markets,

however, it is difficult to discern objective quality. Instead, a significant factor predicting an

opportunity’s success is whether key stakeholders come to realize that a potential opportu-

nity is valuable and subsequently buy in, by adopting, endorsing, or committing resources

to it. I show that in markets where competitors are a set of key stakeholders, knowledge

sharing serves as a strategic response to this buy-in uncertainty with the aim of focusing

stakeholder attention on a potential opportunity that has been found. In these cases, the

costs associated with knowledge sharing among competitors are offset by the benefits of po-

tentially gaining the necessary buy-in. By virtually connecting individuals from around the

globe, digital platforms provide a medium for achieving this buy-in.

Importantly, this buy-in uncertainty mechanism does not refute the relational mecha-

nisms discussed in previous work. As highlighted above, interviewees reported that they

often share knowledge among a small group, often before submitting the recommendation

to RIC. The existence of these “idea dinners,” which are “smaller [and] intimate,” in the

investment management industry has been highlighted for some time (Anderson 2005). My

interviews suggest that this relational knowledge sharing was motivated more by receiving

feedback than by gaining buy-in. It is also possible that experience with dyadic knowledge

sharing increases the likelihood that a firm would allow their employees to share knowledge

more broadly, via a digital platform. This provides an opportunity for future research to ad-

judicate how the main goal of knowledge sharing in a particular instance, namely for feedback

versus buy-in, leads to different knowledge strategies, as well as how relational mechanisms

complement the market-based mechanism found in this research.

Relatedly, this research also contributes to the body of research showing instances of

competitors acting more “cooperatively” in a competitive environment (e.g., Barnett and

Carroll 1987; Gulati, Nohria, and Zaheer 2000; Hamel, Doz, and Prahalad 1989; Mowery,

Oxley, and Silverman 1996; Navis and Glynn 2010), by uncovering an additional positive

externality (e.g., Barnett and Carroll 1987; Burris 2005; Krugman 1991; Mizruchi 1992;

Navis and Glynn 2010; Sorenson and Audia 2000; Yue, Luo, and Ingram 2013). Specifically,

competitors are often key stakeholders in their market and therefore occupy a unique position.

They can both easily interpret complex knowledge and judge its quality; furthermore, they

30

have the resources to commit to an opportunity and provide the necessary buy-in. Therefore,

in contexts where a first-mover advantage can be secured, I show that knowledge sharing can

be seen as a type of “cooperation”–a strategic response to high levels of buy-in uncertainty. A

subsequent question is: why do these competitors buy in? This is not a case of altruism, and

if the competitor had nothing to gain they would not buy in. Instead, this behavior can best

be explained by the aphorism “a rising tide lifts all boats.” While buy-in will most benefit

the firm that discovered the opportunity, firms that buy in soon after will also extract profit

from the opportunity, without expending the cost of discovering it.

This study also contributes to research on online communities that has focused on the

motivations for participating on digital platforms more generally (e.g., Bagozzi and Dholakia

2006; Constant, Sproull, and Kiesler 1996; Hwang, Singh, and Argote 2015; Kollock 1999;

Lakhani and von Hippel 2003; Lerner and Tirole 2002; Mizruchi 1992; Wasko and Faraj

2005). Previous work has highlighted individual-level motivations for sharing knowledge via

professional-typed digital platforms, such as reputation building by offering high-quality

feedback, and reciprocity by paying the community back for feedback they have received.

Although these individual-level motivations surely play a role in an actor’s decision to join

a knowledge-sharing platform, they are insufficient for explaining the relationship between

knowledge sharing and buy-in uncertainty in my context. For example, I do not find evidence

that career stage affects knowledge sharing (Table 7). Therefore, this study offers insight

about how a market-based mechanism can also motivate the existence of knowledge sharing

through a digital platform. A promising avenue for future work relates to how market-

based mechanisms, such as buy-in uncertainty, help sustain the emergence of other, more

recent, digital platforms, such as crowdfunding, whose aim is to help those who have found

a potential opportunity. Often, opportunities featured on crowdfunding platforms are at an

early stage when buy-in uncertainty is extremely high. Future work may look at how this

motivates participation in crowdfunding and how it affects key outcomes, such as funds

raised. Further, the need for buy-in may also be a mechanism that affects the likelihood that

employees participate in digital knowledge-sharing platforms within the firm. Employees

may hesitate to widely share an idea to protect their ability to capitalize on receiving credit;

thus, certain ideas, where buy-in is especially needed, may motivate employees to share more

broadly within their firm.

Finally, outside of research on knowledge sharing and digital platforms, this study will

31

be of interest to scholars who have showed that market efficiency cannot be taken as a given

(Beunza, Hardie, and MacKenzie 2006; MacKenzie and Millo 2003; Turco and Zuckerman

2014; Zuckerman 2012b). The need to engage in knowledge sharing in a financial market

context stands in direct contrast to the expectations of neoclassical finance (cf. Shiller and

Pound 1989) and the efficient-market hypothesis (EMH; Fama 1965, 1970). The EMH sets the

expectation that buy-in uncertainty will not be a prevalent issue in the financial markets. My

findings, however, show that a lack of scrutiny and attention from key evaluative institutions

in the market increases the odds of sharing knowledge about these investment opportunities.

This research underscores the fact that, like most other markets, financial markets are not

immune to missing an opportunity and that this organizational form may be a means by

which a more efficient market is achieved.

While this study explains knowledge sharing in a substantive economic market, the fi-

nancial market, it is important to highlight key scope conditions. The theory outlined above

assumes that the market in question is one where actors are resource constrained; in other

words, they cannot solve buy-in uncertainty by themselves. Furthermore, actors must be

able to reasonably secure a first-mover advantage, and the shared knowledge cannot per-

fectly replicate their competitive advantage, such as sharing one’s algorithm or formula.

Another scope condition relates to the fact that a competitor’s role as a key stakeholder is

not easily substitutable with a non-competing alternative. However, even in markets where

these non-competing substitutes exist, there are still risks in engaging in knowledge sharing.

For example, non-competing firms may transfer shared knowledge to one’s competitor. Re-

latedly, an important limitation of these data is that all investment professionals under study

selected into joining the platform, as opposed to the entire universe of investment profession-

als. To gain access to the platform, these investment professionals have shared at least one

investment recommendation to buy or short sell a stock using at least a simple justification.

Although submitting recommendations solely using simple justifications does not constitute

the detailed knowledge sharing of interest, as aforementioned, investment professionals opt-

ing into this platform may differ from investment professionals who would not even share this

commonly available information. This fact further underscores the importance of conducting

an investment professional fixed effects analysis (Table 5 and Table 6) to help rule out other

issues of selection and individual heterogeneity, such as career concerns.

These findings also offer insights for entrepreneurship, particularly, entrepreneurial oppor-

32

tunity discovery and subsequent opportunity exploitation, or profiting (as seen in the Tesla

example above). While many of these investment professionals are not entrepreneurs in the

traditional sense, they are partaking in a kind of “valuation entrepreneurship” (Zuckerman

2012a) that parallels entrepreneurial opportunity discovery. Specifically, while entrepreneurs

are often categorized as keeping their opportunities “close to the vest,” these results suggest

that a similar knowledge sharing strategy may be advantageous for increasing entrepreneurs’

likelihood of profiting from an opportunity they have discovered, particularly when facing

significant buy-in uncertainty.

Finally, these findings also have important strategic implications for firms more generally.

Firms are increasingly leveraging internal digital platforms as part of their knowledge sharing

strategy (McKinsey & Company 2013). My findings suggest that firm management should

also consider how using digital platforms more broadly, outside of the firm, may help improve

their strategy and performance. For knowledge to yield a competitive advantage it must be

protected; however, its value is directly tied to a firm’s ability to profit from this knowledge.

Therefore, similar types of digital knowledge-sharing platforms that connect competitors may

have strategic value for realizing a profit from opportunity discovery, especially when buy-in

uncertainty is high.

33

ReferencesAbreu, Dilip, and Markus K. Brunnermeier. 2003. “Bubbles and Crashes.” Econometrica 71

(1): 173–204.Anderson, Jenny. 2005. “Potluck à la Wall Street: ’Idea dinners’.” The New York Times.Ang, Andrew, Assaf A. Shtauber, and Paul C. Tetlock. 1989. “Asset pricing in the dark: The

cross-section of OTC stocks.” Review of Financial Studies 26 (12): 2985–3028.Appleyard, Melissa M. 1996. “How Does Knowledge Flow? Interfirm Patterns in the Semi-

conductor Industry.” Strategic Management Journal 17:137–154.Argote, Linda. 2012. Organizational Learning: Creating, Retaining and Transferring Knowl-

edge. Springer Science & Business Media.Argote, Linda, and Paul Ingram. 2000. “Knowledge Transfer: A Basis for Competitive Ad-

vantage in Firms.” Organizational Behavior and Human Decision Processes 82 (1): 150–169.

Asquith, Paul, Parag A. Pathak, and Jay R. Ritter. 2005. “Short interest, institutional own-ership, and stock returns.” Journal of Financial Economics 78 (2): 243–276.

Bagozzi, Richard P., and Utpal M. Dholakia. 2006. “Open Source Software User Commu-nities: A Study of Participation in Linux User Groups.” Management Science 52 (7):1099–1115.

Barnett, William P., and Glenn R. Carroll. 1987. “Competition and Mutualism among EarlyTelephone Companies.” Administrative Science Quarterly 32 (3): 400–421.

Barney, Jay B. 1986. “Strategic Factor Markets: Expectations, Luck, and Business Strategy.”Management Science 32 (10): 1231–1241.

Baum, Joel A. C., and Kristina B. Dahlin. 2007. “Aspiration Performance and Railroads’Patterns of Learning from Train Wrecks and Crashes.” Organization Science 18 (3):368–385.

Benner, Mary J., and Ram Ranganathan. 2012. “Divergent Reactions to Convergent Strate-gies: Investor Beliefs and Analyst Reactions During Technological Change.” OrganizationScience 24 (2): 378–394.

Beunza, Daniel, Iain Hardie, and Donald MacKenzie. 2006. “A Price is a Social Thing:Towards a Material Sociology of Arbitrage.” Organization Studies 27 (5): 721–745.

Boehmer, Ekkehart, and Eric K. Kelley. 2009. “Institutional Investors and the InformationalEfficiency of Prices.” The Review of Financial Studies 22 (9): 3563–3594.

Brown, John Seely1, jsb@parc.xerox.com, and Paul1 Duguid duguid@parc.xerox.com. 2001.“Knowledge and Organization: A Social-Practice Perspective.” Organization Science 12(2): 198–213.

Brown, Stephen J., and Jerold B. Warner. 1985. “Using daily stock returns: The case of eventstudies.” Journal of Financial Economics 14 (1): 3–31.

Brunnermeier, Markus K., and Stefan Nagel. 2004. “Hedge Funds and the Technology Bub-ble.” The Journal of Finance 59 (5): 2013–2040.

Burris, Val. 2005. “Interlocking Directorates and Political Cohesion among Corporate Elites.”American Journal of Sociology 111 (1): 249–283.

Burt, Ronald S. 1995. Structural holes: the social structure of competition. Harvard UniversityPress.

Bushman, Robert M., Joseph D. Piotroski, and Abbie J. Smith. 2005. “Insider TradingRestrictions and Analysts’ Incentives to Follow Firms.” The Journal of Finance 60 (1):35–66.

Bushman, Robert M., and Abbie J. Smith. 2001. “Financial accounting information andcorporate governance.” Journal of Accounting and Economics 32 (1): 237–333.

Carter, Anne P. 1989. “Knowhow trading as economic exchange.” Research Policy 18 (3):155–163.

Cohen, Lauren, Andrea Frazzini, and Christopher Malloy. 2008. “The Small World of Invest-ing: Board Connections and Mutual Fund Returns.” Journal of Political Economy 116(5): 951–979.

Cohen, Wesley M., and Daniel A. Levinthal. 1990. “Absorptive Capacity: A New Perspectiveon Learning and Innovation.” Administrative Science Quarterly 35 (1): 128–152.

Coleman, James S. 1988. “Social Capital in the Creation of Human Capital.” AmericanJournal of Sociology 94:S95–S120.

Comstock, Courtney. 2011. “Shares Of Green Mountain Tank After David Einhorn Says He’sNegative.” Business Insider. http://www.businessinsider.com/david-einhorn-is-negative-on-green-mountain-2011-10.

Connelly, Brian L., Laszlo Tihanyi, S. Trevis Certom, and Michael A. Hitt. 2010. “Marchingto the Beat of Different Drummers: The Influence of Institutional Owners on Competi-tive Actions.” Academy of Management Journal 53 (4): 723–742.

34

Constant, David1, Lee2 Sproull, and Sara1 Kiesler. 1996. “The Kindness of Strangers: TheUsefulness of Electronic Weak Ties for Technical Advice.” Organization Science 7 (2):119–135.

Crawford, Steven S., Wesley R. Gray, and Andrew E. Kern. 2017. “Why Do Fund ManagersIdentify and Share Profitable Ideas?” Journal of Financial and Quantitative Analysis52 (5): 1903–1926.

Curtis, Asher, and Neil L. Fargher. 2014. “Does Short Selling Amplify Price Declines or AlignStocks with Their Fundamental Values?” Management Science.

Dierickx, Ingemar, and Karel Cool. 1989. “Asset Stock Accumulation and Sustainability ofCompetitive Advantage.” Management Science 35 (12): 1504–1511.

Dow, James, and Gary Gorton. 1994. “Arbitrage Chains.” The Journal of Finance 49 (3):819–849.

Duflo, Esther, and Emmanuel Saez. 2002. “Participation and investment decisions in a re-tirement plan: the influence of colleagues’ choices.” Journal of Public Economics 85 (1):121–148.

. 2003. “The Role of Information and Social Interactions in Retirement Plan Decisions:Evidence from a Randomized Experiment.” The Quarterly Journal of Economics 118 (3):815–842.

Evans, David S. 2003. “Some Empirical Aspects of Multi-sided Platform Industries.” Reviewof Network Economics 2 (3).

Fama, Eugene F. 1965. “The Behavior of Stock-Market Prices.” The Journal of Business 38(1): 34–105.

. 1970. “Efficient Capital Markets: A Review of Theory and Empirical Work.” TheJournal of Finance 25 (2): 383–417.

Fang, Lily, and Joel Peress. 2009. “Media Coverage and the Cross-section of Stock Returns.”The Journal of Finance 64 (5): 2023–2052.

Fauchart, Emmanuelle, and Eric von Hippel. 2008. “Norms-Based Intellectual Property Sys-tems: The Case of French Chefs.” Organization Science 19 (2): 187–201.

Fernandez, Roberto M., and M. Lourdes Sosa. 2005. “Gendering the Job: Networks andRecruitment at a Call Center.” American Journal of Sociology 111 (3): 859–904.

Financial Times. 2014. Global MBA Ranking. London, U.K.: Financial Times, Ltd.Francis, Jennifer, J. Douglas Hanna, and Donna R. Philbrick. 1997. “Management commu-

nications with securities analysts.” Journal of Accounting and Economics 24 (3): 363–394.

Gawer, Annabelle. 2011. Platforms, Markets and Innovation. Edward Elgar Publishing.. 2014. “Bridging differing perspectives on technological platforms: Toward an inte-

grative framework.” Research Policy 43 (7): 1239–1249.Geczy, Christopher C., David K. Musto, and Adam V. Reed. 2002. “Stocks are special too:

an analysis of the equity lending market.” Journal of Financial Economics, Limits onArbitrage, 66 (2): 241–269.

Grant, Robert M. 1996. “Prospering in Dynamically-Competitive Environments: Organiza-tional Capability as Knowledge Integration.” Organization Science 7 (4): 375–387.

Greif, Avner. 1993. “Contract Enforceability and Economic Institutions in Early Trade: TheMaghribi Traders’ Coalition.” The American Economic Review 83 (3): 525–548.

Groysberg, Boris, and Linda-Eling Lee. 2008. “The Effect of Colleague Quality on Top Per-formance: The Case of Security Analysts.” Journal of Organizational Behavior 29 (8):1123–1144.

Gulati, Ranjay, Nitin Nohria, and Akbar Zaheer. 2000. “Strategic Networks.” Strategic Man-agement Journal 21 (3): 203–215.

Hamel, Gary, Yves Doz, and C. K. Prahalad. 1989. “Collaborate with Your Competitors—andWin.” Harvard Business Review. Accessed July 27, 2017. https://hbr.org/1989/01/collaborate-with-your-competitors-and-win.

Hansen, Morten T. 1999. “The Search-Transfer Problem: The Role of Weak Ties in SharingKnowledge across Organization Subunits.” Administrative Science Quarterly 44 (1): 82–111.

. 2002. “Knowledge Networks: Explaining Effective Knowledge Sharing in MultiunitCompanies.” Organization Science 13 (3): 232–248.

Haunschild, Pamela R., and Bilian Ni Sullivan. 2002. “Learning from Complexity: Effects ofPrior Accidents and Incidents on Airlines’ Learning.” Administrative Science Quarterly47 (4): 609–643.

Hong, Harrison, Jeffrey D. Kubik, and Jeremy C. Stein. 2004. “Social Interaction and Stock-Market Participation.” The Journal of Finance 59 (1): 137–163.

. 2005. “Thy Neighbor’s Portfolio: Word-of-Mouth Effects in the Holdings and Tradesof Money Managers.” The Journal of Finance 60 (6): 2801–2824.

35

Hong, Harrison, Terence Lim, and Jeremy C. Stein. 2000. “Bad News Travels Slowly: Size,Analyst Coverage, and the Profitability of Momentum Strategies.” The Journal of Fi-nance 55 (1): 265–295.

Hwang, Elina H., Param Vir Singh, and Linda Argote. 2015. “Knowledge Sharing in OnlineCommunities: Learning to Cross Geographic and Hierarchical Boundaries.” OrganizationScience 26 (6): 1593–1611.

Ingram, Paul, and Peter W. Roberts. 2000. “Friendships among Competitors in the SydneyHotel Industry.” American Journal of Sociology 106 (2): 387–423.

Inkpen, Andrew C., and Eric W. K. Tsang. 2005. “Social Capital, Networks, and KnowledgeTransfer.” Academy of Management Review 30 (1): 146–165.

Jeppesen, Lars Bo, and Lars Frederiksen. 2006. “Why Do Users Contribute to Firm-HostedUser Communities? The Case of Computer-Controlled Music Instruments.” OrganizationScience 17 (1): 45–63.

Kaufman, Alexander C. 2014. “Tesla’s Clever Patent Move Is Already Paying Off.” http://www.huffingtonpost.com/2014/06/16/tesla-patent-supercharger-station_n_5500724.html.

Keynes, John Maynard. 1936. General Theory Of Employment , Interest And Money. Lon-don: Macmillan.

King, Brayden G., and Sarah A. Soule. 2007. “Social Movements as Extra-Institutional En-trepreneurs: The Effect of Protests on Stock Price Returns.” Administrative ScienceQuarterly 52 (3): 413–442.

Knight, Frank H. 1921. Risk, uncertainty and profit. N.Y. Hart Schaffner Marx.Kochhar, Rahul, and David Parthiban. 1998. “Institutional Investors and Firm Innovation:

A Test of Competing Hypotheses.” Strategic Management Journal 17 (1): 73–84.Kollock, Peter. 1999. “The Economies of Online Cooperation.” In Communities in Cy-

berspace, edited by Peter Kollock and Marc Smith, 220–239. Routledge.Krugman, Paul. 1991. “Increasing Returns and Economic Geography.” Journal of Political

Economy 99 (3): 483–499.Lakhani, Karim R, and Eric von Hippel. 2003. “How open source software works: “free”

user-to-user assistance.” Research Policy 32 (6): 923–943.Lang, Mark H., and Russell J. Lundholm. 1996. “Corporate Disclosure Policy and Analyst

Behavior.” The Accounting Review 71 (4): 467–492.Lerner, Josh, and Jean Tirole. 2002. “Some Simple Economics of Open Source.” The Journal

of Industrial Economics 50 (2): 197–234.Levitt, Barbara, and James G. March. 1988. “Organizational Learning.” Annual Review of

Sociology 14:319–340.Liebeskind, Julia Porter. 1996. “Knowledge, strategy, and the theory of the firm.” Strategic

Management Journal 17 (S2): 93–107.Lippman, S. A., and R. P. Rumelt. 1982. “Uncertain Imitability: An Analysis of Interfirm

Differences in Efficiency under Competition.” The Bell Journal of Economics 13 (2):418–438.

Ljungqvist, Alexander, and Wenlan Qian. 2016. “How Constraining Are Limits to Arbi-trage?” The Review of Financial Studies 29 (8): 1975–2028.

MacKenzie, Donald, and Yuval Millo. 2003. “Constructing a Market, Performing Theory:The Historical Sociology of a Financial Derivatives Exchange.” American Journal ofSociology 109 (1): 107–145.

Madsen, Peter M., and Vinit Desai. 2010. “Failing to Learn? The Effects of Failure andSuccess on Organizational Learning in the Global Orbital Launch Vehicle Industry.”Academy of Management Journal 53 (3): 451–476.

McKinsey & Company. 2013. Evolution of the networked enterprise: McKinsey Global Surveyresults.

McPherson, Miller, Lynn Smith-Lovin, and James M. Cook. 2001. “Birds of a Feather: Ho-mophily in Social Networks.” Annual Review of Sociology 27:415–444.

Mizruchi, Mark S. 1992. The Structure of Corporate Political Action: Interfirm Relationsand Their Consequences. Harvard University Press.

Mowery, David C., Joanne E. Oxley, and Brian S. Silverman. 1996. “Strategic alliances andinterfirm knowledge transfer.” Strategic Management Journal 17 (S2): 77–91.

Navis, Chad, and Mary Ann Glynn. 2010. “How New Market Categories Emerge: TemporalDynamics of Legitimacy, Identity, and Entrepreneurship in Satellite Radio, 1990–2005.”Administrative Science Quarterly 55 (3): 439–471.

Parthiban, David, Rahul Kochhar, and Edward Levitas. 1998. “The Effect of InstitutionalInvestors on the Level and Mix of Ceo Compensation.” Academy of Management Journal41 (2): 200–208.

36

Powell, Walter W., Kenneth W. Koput, and Laurel Smith-Doerr. 1996. “InterorganizationalCollaboration and the Locus of Innovation: Networks of Learning in Biotechnology.”Administrative Science Quarterly 41 (1): 116–145.

Rao, Hayagreeva, and Kumar Sivakumar. 1999. “Institutional Sources of Boundary-SpanningStructures: The Establishment of Investor Relations Departments in the Fortune 500Industrials.” Organization Science 10 (1): 27–42.

Reagans, Ray E, and Ezra W Zuckerman. 2008. “Why Knowledge Does Not Equal Power:The Network Redundancy Trade-Off.” Industrial and Corporate Change 17 (5): 903–944.

Reagans, Ray, and Bill McEvily. 2008. “Contradictory or compatible? reconsidering thetrade-off between brokerage and closure on knowledge sharing.” In Network Strategy,25:275–313. Advances in Strategic Management 25. Emerald Group Publishing Limited.

Rochet, Jean-Charles, and Jean Tirole. 2006. “Two-sided markets: a progress report.” TheRAND Journal of Economics 37 (3): 645–667.

Saxenian, AnnaLee. 1996. Regional advantage: culture and competition in Silicon Valley andRoute 128. Harvard University Press.

Schrader, Stephan. 1991. “Informal technology transfer between firms: Cooperation throughinformation trading.” Research Policy 20 (2): 153–170.

Shiller, Robert J., and John Pound. 1989. “Survey evidence on diffusion of interest andinformation among investors.” Journal of Economic Behavior & Organization 12 (1):47–66.

Sorenson, Olav, and Pino G. Audia. 2000. “The Social Structure of Entrepreneurial Activity:Geographic Concentration of Footwear Production in the United States, 1940–1989.”American Journal of Sociology 106 (2): 424–462.

Stein, Jeremy C. 2008. “Conversations among Competitors.” The American Economic Review98 (5): 2150–2162.

Stiglitz, Joseph E. 1990. “Peer Monitoring and Credit Markets.” The World Bank EconomicReview 4 (3): 351–366.

Teece, David J. 1977. “Technology Transfer by Multinational Firms: The Resource Cost ofTransferring Technological Know-How.” The Economic Journal 87 (346): 242–261.

. 1986. “Profiting from technological innovation: Implications for integration, collab-oration, licensing and public policy.” Research Policy 15 (6): 285–305.

Tesla Motors. 2014. “All Our Patents Belong To You.” Tesla Motors. http://www.teslamotors.com/blog/all-our-patent-are-belong-you.

Thompson, Mark. 2005. “Structural and Epistemic Parameters in Communities of Practice.”Organization Science 16 (2): 151–164.

Towers Watson. 2012. “Towers Watson Launches Knowledge-Sharing Network.” Towers Wat-son. http://tinyurl.com/o983ffn.

Turco, Catherine, and Ezra Zuckerman. 2014. “So You Think You Can Dance? Lessons fromthe US Private Equity Bubble.” Sociological Science 1:81–101.

U.S. News and World Report. 2014a. Best Business Schools. Washington, D.C.: U.S. Newsand World Report.

. 2014b. National University Rankings. Washington, D.C.: U.S. News and World Re-port.

Value Investing Congress. 2014a. “New York 2013 Highlights.” http://www.valueinvestingcongress.com/the_congress/past_congress_highlights/2013/new_york/.

. 2014b. “Our Mission.” http://www.valueinvestingcongress.com/the_congress/our_mission/.

von Hippel, Eric. 1987. “Cooperation between rivals: Informal know-how trading.” ResearchPolicy 16 (6): 291–302.

Voyles, Bennett. 2014a. “Why Elon Musk Just Opened Tesla’s Patents to His Biggest Rivals.”Accessed September 25, 2015. http://www.bloomberg.com/bw/articles/2014-06-12/why-elon-musk-just-opened-teslas-patents-to-his-biggest-rivals.

. 2014b. “Why Elon Musk Released Tesla’s Patents.” Accessed July 7, 2014. http://knowledge.ckgsb.edu.cn/2014/07/07/china-business-strategy/why-elon-musk-released-teslas-patents/.

Wasko, Molly, and Samer Faraj. 2005. “Why Should I Share? Examining Social Capital andKnowledge Contribution in Electronic Networks of Practice.” MIS Quarterly 29 (1): 35–57.

Wenger, Etienne. 1998. Communities of Practice: Learning, Meaning, and Identity. Cam-bridge University Press.

Wenger, Etienne, Richard Arnold McDermott, and William Snyder. 2002. Cultivating Com-munities of Practice: A Guide to Managing Knowledge. Harvard Business Press.

Williamson, Oliver E. 1991. “Comparative Economic Organization: The Analysis of DiscreteStructural Alternatives.” Administrative Science Quarterly 36 (2): 269–296.

37

Yu, Fang (Frank). 2008. “Analyst coverage and earnings management.” Journal of FinancialEconomics 88 (2): 245–271.

Yue, Lori Qingyuan, Jiao Luo, and Paul Ingram. 2013. “The Failure of Private Regulation:Elite Control and Market Crises in the Manhattan Banking Industry.” AdministrativeScience Quarterly 58 (1): 37–68.

Zack, Michael H. 1999. “Developing a knowledge strategy.” California management review41 (3): 125–145.

Zajac, Edward J., and James D. Westphal. 2004. “The Social Construction of Market Value:Institutionalization and Learning Perspectives on Stock Market Reactions.” AmericanSociological Review 69 (3): 433–457.

Zander, Udo, and Bruce Kogut. 1995. “Knowledge and the Speed of the Transfer and Imi-tation of Organizational Capabilities: An Empirical Test.” Organization Science 6 (1):76–92.

Zhang, X. Frank. 2006. “Information Uncertainty and Stock Returns.” The Journal of Fi-nance 61 (1): 105–137.

Zuckerman, Ezra W. 2000. “Focusing the Corporate Product: Securities Analysts and De-diversification.” Administrative Science Quarterly 45 (3): 591–619.

. 2012a. “Construction, Concentration, and (Dis)Continuities in Social Valuations.”Annual Review of Sociology 38 (1): 223–245.

. 2012b. “Market efficiency: a sociological perspective.” The Oxford Handbook of theSociology of Finance: 223.

Zuckerman, Ezra W., and Stoyan V. Sgourev. 2006. “Peer Capitalism: Parallel Relationshipsin the U.S. Economy.” American Journal of Sociology 111 (5): 1327–1366.

38

Figures and Tables

Figure 1: Example of Justification Types

A. Simple Justification

Note: Each row is an example of a simple justification. These are representations from randomlyselected data and removes any information about the stock.

B. Detailed Justification

Note: This is representative example of the type of detailed knowledge that is included ina detailed justification on RIC. This example was obtained by the author, from a buy-sideinvestment professional, and is not from the RIC database. Further, this analysis is meant toserve as an example and is not a recommendation of this stock.

39

Figure 2: Average Market Capitalization of Investment Recommendations by Justification Type

Note: Market capitalization was averaged across investment recommendations within quarter and withinjustification type from 2008 to 2013. The horizontal lines represent the overall sample mean for the respectivejustification type.

40

Figure 3: Likelihood of Knowledge Sharing on Measures of Buy-in Uncertainty (Marginal Effectsfrom Logit Regressions)

A. Sell-Side Analyst Coverage B. Firm Age

C. Media Attention D. Instit. Investor Concen. (Per., Top 5)

E. Instit. Investor Concen. (Per., Other)

Note: These are the marginal effects of the logit model predicting knowledge sharing on different measuresof buy-in uncertainty (Table 3) with control variables held at mean value. For Figures 3A, 3B, 3C, and 3D,as the value on the x-axis increases so does buy-in uncertainty. For Figure 3E, as the value on the x-axisincreases buy-in uncertainty decreases. The gray shading represents the 95% confidence interval.

41

Figure 4: Plot of 21-Day (-10,10) CAAR for Recommendations with Knowledge Sharing

A. Buy Recommendations

B. Short-Sell Recommendations

Note: These graphs plot the cumulative average abnormal return for stocks recommended witha detailed justification (Figure 4A is for recommendations to buy-and-hold the stock and Figure4B is for recommendations to short sell the stock). The day of the recommendation is representedby the vertical dashed line. If a recommendation occurred outside of market hours, the nextmarket day was defined as the event day. The t-stat is greater than 2 for Figure 4A starting onday -6 and for Figure 4B starting on day -3.

42

Table 1: Summary Statistics of Key VariablesPanel A: Detailed Justifications

Variable Obs. Mean Std. Dev. Min. Max.

Explanatory VariablesSell-Side Coveragea 3,538 50.204 41.225 4.000 272.000Firm Age 4,026 16.653 17.424 0.000 88.000Media Attention 4,026 1.269 7.175 0.000 183.000Instit. Investor Concen. (Per., Top 5)a 3,552 0.288 0.132 0.002 1.000Instit. Investor Concen. (Per., Other)a 3,552 0.364 0.205 0.000 0.778Short Recommendation 4,026 0.161 0.367 0.000 1.000

Control Variables—Recommendation LevelFirm Size (B) 4,026 6.828 29.232 0.006 594.864Short Investment Horizon 4,026 0.425 0.494 0.000 1.000

Control Variables—Investment Professional LevelLocation: Major City 4,026 0.724 0.447 0.000 1.000Location: Non-US 4,026 0.053 0.225 0.000 1.000Elite Undergraduate 4,026 0.330 0.470 0.000 1.000Elite Graduate 4,026 0.332 0.471 0.000 1.000

Panel B: Simple JustificationsVariable Obs. Mean Std. Dev. Min. Max.

Explanatory VariablesSell-Side Coveragea 13,824 63.673 45.497 4.000 272.000Firm Age 15,067 19.115 19.160 0.000 88.000Media Attention 15,067 1.747 7.172 0.000 147.000Instit. Investor Concen. (Per., Top 5)a 13,337 0.276 0.124 0.000 1.000Instit. Investor Concen. (Per., Other)a 13,337 0.408 0.185 0.000 0.812Short Recommendation 15,067 0.053 0.223 0.000 1.000

Control Variables—Recommendation LevelFirm Size (B) 15,067 18.357 47.490 0.002 657.975Short Investment Horizon 15,067 0.127 0.333 0.000 1.000

Control Variables—Investment Professional LevelLocation: Major City 15,067 0.726 0.446 0.000 1.000Location: Non-US 15,067 0.077 0.267 0.000 1.000Elite Undergraduate 15,067 0.381 0.486 0.000 1.000Elite Graduate 15,067 0.315 0.465 0.000 1.000

Notes: This table presents the summary statistics for the 4,026 investment recommendations that included adetailed justification (Panel A) to the 15,067 investment recommendations that included a simple justification(Panel B). The raw values of the variables are presented. The reciprocal values of Sell-Side Coverage, FirmAge, Media Attention, and Firm Size are used in the analysis.aThe number of observations for these measures is less than the maximum sample size because values forthese measures were not available for all of the firms in the sample. Results are robust to changing missingvalues to 0.

43

Tab

le2:

Cor

rela

tion

sof

Var

iabl

esVariables

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11)

(12)

(13)

(1)

Kno

wledg

eSh

aring

1.000

(2)

Sell-Side

Coverage

-0.121

1.000

(3)

Firm

Age

-0.053

0.163

1.000

(4)

Instit.Inv

estorCon

cen.

(Per.,To

p5)

0.040

-0.188

-0.148

1.000

(5)

Instit.Inv

estorCon

cen.

(Per.,Other)

-0.095

0.350

0.157

0.197

1.000

(6)

Media

Attention

-0.027

0.367

0.053

-0.092

0.065

1.000

(7)

ShortRecom

menda

tion

0.167

0.024

-0.041

0.033

-0.023

0.054

1.000

(8)

Firm

Size

-0.106

0.514

0.287

-0.245

0.089

0.505

-0.041

1.000

(9)

ShortInvestmentHorizon

0.310

-0.027

-0.044

-0.001

-0.054

0.037

0.278

-0.034

1.000

(10)

Location

:Major

City

-0.002

-0.013

-0.022

0.014

0.013

-0.015

-0.003

-0.026

0.007

1.000

(11)

Location

:Non

-US

-0.038

0.039

0.016

-0.029

-0.023

0.025

-0.018

0.059

-0.001

-0.454

1.000

(12)

EliteUnd

ergrad

uate

-0.043

0.005

-0.015

-0.006

0.009

-0.003

-0.017

-0.010

-0.039

0.189

-0.119

1.000

(13)

EliteGradu

ate

0.015

0.001

-0.020

0.008

-0.002

0.003

0.026

-0.012

0.024

0.161

-0.053

0.075

1.000

Notes:C

orrelation

sgreaterthan

|0.0145|

aresign

ificant

atp

≤0.

05.T

heraw

values

ofthevariab

lesarepresented.

The

reciprocal

values

ofSe

ll-Si

deC

over

age,

Fir

mA

ge,M

edia

Att

entio

n,an

dF

irm

Size

areused

inthean

alysis.

44

Tab

le3:

Log

itR

egre

ssio

nof

Kno

wle

dge

Shar

ing

onM

easu

res

ofB

uy-i

nU

ncer

tain

tyM1

M2

M3

M4

M5

Sell-Side

Coveragea

4.267

***

(0.484)

Firm

Age

a0.687

***

(0.123)

Med

iaAttentio

na0.706

***

(0.075)

Instit.

Investor

Con

cen.

(Per.,To

p5)

0.933

***

(0.167)

Instit.

Investor

Con

cen.

(Per.,Other)

-1.129

***

(0.143)

Firm

Size

a0.015

***

0.011

*0.014

***

0.012

***

0.008

**(0.003)

(0.005)

(0.003)

(0.003)

(0.003)

ShortInvestmentHorizon

1.338

***

1.334

***

1.330

***

1.349

***

1.290

***

(0.072)

(0.075)

(0.072)

(0.072)

(0.075)

Locatio

n:Major

City

0.005

0.032

-0.002

0.005

0.059

(0.091)

(0.093)

(0.091)

(0.092)

(0.092)

Locatio

n:Non

-US

-0.504

**-0.464

**-0.507

**-0.475

**-0.475

**(0.173)

(0.168)

(0.172)

(0.173)

(0.174)

EliteUnd

ergrad

uate

-0.172

*-0.186

**-0.174

*-0.171

*-0.171

*(0.071)

(0.072)

(0.071)

(0.071)

(0.070)

EliteGradu

ate

0.067

0.058

0.061

0.067

0.070

(0.075)

(0.074)

(0.075)

(0.075)

(0.074)

Con

stan

t-2.244

***

-3.110

***

-2.356

***

-2.882

***

-2.238

***

(0.400)

(0.540)

(0.399)

(0.407)

(0.419)

Observatio

ns19,093

17,362

19,093

19,093

16,889

Log-lik

elihoo

d-8,524

-7,517

-8,504

-8,463

-7,476

Notes:Unitof

analysis

istheinvestmentrecommen

datio

n.Mod

elscontain

year

and

indu

stry

fixed

effects

with

robu

ststan

dard

errors,c

lustered

attheinvestmentprofession

al-le

vel,in

parenthe

ses.Sign

ificancelevels:+

p≤

0.10,*

p≤

0.05,*

*p

≤0.

01,*

**p

≤0.

001.

aThe

reciprocal

ofthis

measure

was

used

intheregression

.

45

Table 4: Logit Regression of Knowledge Sharing on Short-Selling Recommenda-tion

M6 M7 M8

Short Recommendation 0.517 *** 0.589 *** 0.516 ***(0.083) (0.086) (0.086)

Sell-Side Coveragea 4.456 ***(0.483)

Instit. Investor Concen. (Per., Top 5) 0.901 ***(0.167)

Instit. Investor Concen. (Per., Other) -1.113 ***(0.144)

Firm Sizea 0.015 *** 0.011 * 0.009 **(0.003) (0.005) (0.003)

Short Investment Horizon 1.258 *** 1.241 *** 1.205 ***(0.072) (0.076) (0.075)

Location: Major City 0.010 0.040 0.064(0.091) (0.093) (0.092)

Location: Non-US -0.487 ** -0.441 ** -0.456 **(0.173) (0.168) (0.174)

Elite Undergraduate -0.172 * -0.187 ** -0.172 *(0.070) (0.072) (0.070)

Elite Graduate 0.061 0.051 0.064(0.075) (0.074) (0.073)

Constant -2.265 *** -3.136 *** -2.248 ***(0.401) (0.538) (0.420)

Observations 19,093 17,362 16,889Log-likelihood -8,492 -7,479 -7,447

Notes: Unit of analysis is the investment recommendation. Models contain year and industryfixed effects with robust standard errors, clustered at the investment professional-level, inparentheses. Significance levels: + p ≤ 0.10, * p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001.aThe reciprocal of this measure was used in the regression.

46

Tab

le5:

Log

itR

egre

ssio

nof

Kno

wle

dge

Shar

ing

onM

easu

res

ofB

uy-i

nU

ncer

tain

ty(w

ithi

n-In

vest

men

tP

ro-

fess

iona

l)R1

R2

R3

R4

R5

Sell-Side

Coveragea

1.982

**(0.660)

Firm

Age

a0.333

*(0.162)

Instit.Inv

estorCon

cen.

(Per.,To

p5)

0.721

**(0.266)

Instit.Inv

estorCon

cen.

(Per.,Other)

-0.706

***

(0.192)

Media

Attention

a0.449

***

(0.102)

Firm

Size

a0.013

***

0.015

**0.013

***

0.008

*0.011

***

(0.003)

(0.005)

(0.003)

(0.004)

(0.003)

ShortInvestmentHorizon

1.430

***

1.399

***

1.429

***

1.385

***

1.431

***

(0.077)

(0.083)

(0.077)

(0.082)

(0.077)

Observation

s6,395

5,671

6,395

5,657

6,395

Log-lik

elihoo

d-2,763

-2,377

-2,760

-2,377

-2,753

Notes:Sa

mplerestricted

toon

lythoseinvestmentprofession

alswho

used

both

justification

type

s.Unitof

analysis

isthe

investmentrecommenda

tion

.Mod

elscontaininvestmentprofession

al,y

ear,

andindu

stry

fixed

effects

withstan

dard

errors

inpa

renthe

ses.

Sign

ificancelevels:+

p≤

0.10,*

p≤

0.05,*

*p

≤0.

01,*

**p

≤0.

001.

aThe

reciprocal

ofthis

measure

was

used

intheregression

.

47

Table 6: Logit Regression of Knowledge Sharing on Short-Selling Recommenda-tion (within-Investment Professional)

R6 R7 R8

Short Recommendation 0.382 *** 0.416 *** 0.343 **(0.102) (0.107) (0.106)

Sell-Side Coveragea 2.124 **(0.662)

Instit. Investor Concen. (Per., Top 5) 0.715 **(0.266)

Instit. Investor Concen. (Per., Other) -0.711 ***(0.193)

Firm Size 0.013 *** 0.014 ** 0.008 *(0.003) (0.005) (0.004)

Short Investment Horizon 1.371 *** 1.333 *** 1.327 ***(0.078) (0.084) (0.084)

Observations 6,395 5,671 5,657Log-likelihood –2,756 –2,369 –2,372

Notes: Sample restricted to only those investment professionals who used both justificationtypes. Unit of analysis is the investment recommendation. Models contain investment pro-fessional, year, and industry fixed effects with standard errors in parentheses. Significancelevels: + p ≤ 0.10, * p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001.aThe reciprocal of this measure was used in the regression.

48

Tab

le7:

Log

itR

egre

ssio

nof

Kno

wle

dge

Shar

ing

onSe

nior

ity

and

Mea

sure

sof

Buy

-in

Unc

erta

inty

R9

R10

R11

R12

R13

Senior

Inv.

Prof.

0.174

0.136

0.237

+-0.015

0.199

+(0.111)

(0.130)

(0.130)

(0.208)

(0.117)

Sell-Side

Coveragea

2.845

*(1.168)

Firm

Age

a0.754

**(0.269)

Media

Attention

a0.282

+(0.165)

ShortRecom

menda

tion

0.619

***

(0.188)

Senior

XSell-Side

Coveragea

-0.072

(1.425)

Senior

XFirm

Age

a-0.550

(0.379)

Senior

XMedia

Attention

a0.219

(0.232)

Senior

XSh

ort

-0.268

(0.252)

Firm

Size

0.012

+0.008

+0.006

0.008

+(0.007)

(0.004)

(0.004)

(0.005)

ShortInvestmentHorizon

1.171

***

1.149

***

1.162

***

1.164

***

1.099

***

(0.111)

(0.114)

(0.111)

(0.111)

(0.112)

Location

:Major

City

-0.180

-0.138

-0.194

-0.186

-0.182

(0.150)

(0.155)

(0.148)

(0.151)

(0.146)

Location

:Non

-US

-0.024

0.065

-0.039

-0.013

-0.013

(0.287)

(0.261)

(0.282)

(0.283)

(0.283)

EliteUnd

ergrad

uate

-0.247

+-0.273

*-0.253

*-0.255

*-0.255

*(0.129)

(0.134)

(0.127)

(0.129)

(0.126)

EliteGradu

ate

-0.111

-0.154

-0.115

-0.114

-0.120

(0.113)

(0.109)

(0.111)

(0.111)

(0.111)

Con

stan

t-1.370

*-2.178

*-1.464

*-1.631

**-1.435

*(0.621)

(0.907)

(0.625)

(0.633)

(0.634)

Observation

s5,187

4,647

5,187

5,187

5,187

Log-lik

elihoo

d-3,178

-2,816

-3,169

-3,164

-3,161

Notes:Su

b-sampleof

recommen

dation

swhere

theinvestmentprofession

alha

datitleavailable(27pe

rcentof

thesample).

Unitof

analysisistheinvestmentrecommenda

tion

.Mod

elscontainyear

andindu

stry

fixed

effects

withrobu

ststan

dard

errors,

clusteredat

theinvestmentprofession

al-le

vel,in

parentheses.

Sign

ificancelevels:+

p≤

0.10,*

p≤

0.05,**

p≤

0.01,***

p≤

0.00

1.aThe

reciprocal

ofthis

measure

was

used

intheregression

.

49