We would like to thank George Baker, Adam Brandenburger, Peter Cappelli, Robert Gibbons, Jerry Green, William Greene, Paul Healy, Casey Ichniowski, David Levine, Will Mitchell, Nitin Nohria, Paul Oyer, Mikolaj Piskorski, Joel Podolny, and Ezra Zuckerman, as well as participants in seminars at Columbia Business School, Econometric Society 2004 North American Winter conference, Harvard Business School, Harvard University Applied Statistic series, London Business School, Ohio State University, Stanford Business School, the University of Pennsylvania, and the University of Washington. We appreciate William Greene’s help with estimating the recently developed ordered probit fixed-effects models. We are grateful to professionals at CS First Boston, Deutsche Bank, Goldman Sachs, Institutional Investor, J. P. Morgan, Lehman Brothers, Merrill Lynch, Morgan Stanley, Salomon Smith Barney, and Sanford C. Bernstein for interviews and comments on previous drafts. We also wish to thank Kathleen Ryan and James Schorr for research assistance. We gratefully acknowledge the Division of Research at the Harvard Business School for financial support for this study. 1 Can They Take It With Them? The Portability of Star Knowledge Workers’ Performance Forthcoming Management Science Boris Groysberg, Linda-Eling Lee, and Ashish Nanda October, 2007 Abstract This paper examines the portability of star security analysts’ performance. Star analysts who switched employers experienced an immediate decline in performance that persisted for at least five years. This decline was most pronounced among star analysts who moved to firms with lesser capabilities and those who moved solo, without other team members. Star analysts who moved between two firms with equivalent capabilities also exhibited a drop in performance, but only for two years. Those who switched to firms with better capabilities and those who moved with other team members exhibited no significant decline in short-term or long-term performance. These findings suggest that firm-specific skills and firms’ capabilities both play important roles in star analysts’ performance. In addition, we find that firms that hire star analysts from competitors with better capabilities suffered more extreme negative stock-market reactions than those that hire from comparable or lesser firms. These findings suggest that hiring stars may be perceived as value-destroying and may not improve a firm’s competitive advantage. JEL Classification J24, J44
37
Embed
Can They Take It With Them? The Portability of Star Knowledge Workers' Performance
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
We would like to thank George Baker, Adam Brandenburger, Peter Cappelli, Robert Gibbons, Jerry Green, William Greene, Paul Healy, Casey Ichniowski, David Levine, Will Mitchell, Nitin Nohria, Paul Oyer, Mikolaj Piskorski, Joel Podolny, and Ezra Zuckerman, as well as participants in seminars at Columbia Business School, Econometric Society 2004 North American Winter conference, Harvard Business School, Harvard University Applied Statistic series, London Business School, Ohio State University, Stanford Business School, the University of Pennsylvania, and the University of Washington. We appreciate William Greene’s help with estimating the recently developed ordered probit fixed-effects models. We are grateful to professionals at CS First Boston, Deutsche Bank, Goldman Sachs, Institutional Investor, J. P. Morgan, Lehman Brothers, Merrill Lynch, Morgan Stanley, Salomon Smith Barney, and Sanford C. Bernstein for interviews and comments on previous drafts. We also wish to thank Kathleen Ryan and James Schorr for research assistance. We gratefully acknowledge the Division of Research at the Harvard Business School for financial support for this study. 1
Can They Take It With Them? The Portability of Star Knowledge Workers’ Performance
Forthcoming Management Science
Boris Groysberg, Linda-Eling Lee, and Ashish Nanda
October, 2007
Abstract
This paper examines the portability of star security analysts’ performance. Star analysts who switched employers
experienced an immediate decline in performance that persisted for at least five years. This decline was most
pronounced among star analysts who moved to firms with lesser capabilities and those who moved solo, without
other team members. Star analysts who moved between two firms with equivalent capabilities also exhibited a drop
in performance, but only for two years. Those who switched to firms with better capabilities and those who moved
with other team members exhibited no significant decline in short-term or long-term performance. These findings
suggest that firm-specific skills and firms’ capabilities both play important roles in star analysts’ performance. In
addition, we find that firms that hire star analysts from competitors with better capabilities suffered more extreme
negative stock-market reactions than those that hire from comparable or lesser firms. These findings suggest that
hiring stars may be perceived as value-destroying and may not improve a firm’s competitive advantage.
JEL Classification J24, J44
2
A Nobel Prize-winning scientist may be a unique resource, but unless he has firm-specific ties, his perfect mobility makes him an unlikely source of sustainable advantage. Managers should ask themselves if his productivity has to do, in part, with the specific team of researchers of which he is a part. Does it depend on his relationship with talented managers who are exceptionally adept at managing creativity? Does it depend on the spirit of the workers or the unique culture of the firm? (Margaret A. Peteraf, 1993, p. 187)
As the labor market shifts toward information- and knowledge-based work, organizations are increasingly
preoccupied with the quality of their workers as a source of competitive advantage (Pfeffer, 1994, 1998). Top
performers—referred to in this paper as stars—are regarded as disproportionately productive and valuable, and
organizations view it as imperative to attract and retain such talent (e.g. Ernst and Vitt, 2000; Hunter, Schmidt, and
Judiesch, 1990; Narin and Breitzman, 1995). Though stars may be extraordinarily productive, they cannot constitute
a sustained competitive advantage if their skills are mobile and transferable across firms.
According to Barney (1991), a firm enjoys a competitive advantage only if its strategic resources are
valuable, rare, difficult to imitate, and lacking in substitutes. Peteraf (1993) specifies four conditions that underlie
competitive advantage: resource heterogeneity, ex-post limits to competition, ex-ante limits to competition, and
imperfect resource mobility. With regard to resource mobility, Peteraf (1993) notes that even a superstar like a Nobel
Prize-winning scientist cannot represent a competitive advantage for an organization unless firm-specific factors
contribute to his or her performance (1993). Hence the resource-based view of the firm has emphasized the tacit
knowledge embedded in organizational structures and relationships among workers (Cohen and Levinthal, 1990;
Kogut and Zander, 1996; Nahapiet and Ghoshal, 1998) rather than capabilities that reside in individual star workers.
Research on knowledge management, meanwhile, has shown that the mobility of individual workers, particularly
high-performing knowledge workers like engineers, is an important means of transferring, dispersing, or buying
knowledge assets (Song, Almeida, and Wu, 2003; Kim 1997; Zander and Kogut, 1995). The extent to which a star’s
performance is transferable across organizations remains a key strategic concern in knowledge-based industries.
In knowledge-intensive industries like finance, accounting, law, and technology, companies and workers
often share the assumption that stars and their talent are highly portable—that is, that they can apply their skills just
as well for one employer as for another. The business media promote this assumption with reports on stars who
defect to competing firms and presumably take their innovative ideas with them (e.g., see Sessa, 1999). Stars’ ability
3
to transfer their value to a new firm is viewed as particularly pronounced in industries in which clients’ loyalties flow to
the individuals who provide a service rather than the firm that employs those individuals (Greenwood, Hinings, and
Brown, 1990). Furious bidding for star performers (Lazear, 1986) also promotes the perception that stars are
essentially free agents, not unlike star athletes whose ability can be bought by the highest bidder.
Human-capital theory distinguishes between general human capital, which is applicable to many
organizational contexts, and firm-specific human capital, which is valuable only to a specific organization (Becker,
1962). The firm-specificity of workers’ skills is the degree to which the human capital they acquired at a particular firm
is idiosyncratic and therefore useless at other firms. Investment in general training raises workers’ productivity in a
manner equally valuable to all comparable employers, whereas firm-specific training increases the value of workers’
marginal products at only one firm. Firm-specific human capital includes familiarity with unique routines and
procedures, tacit knowledge embedded in interpersonal relationships and corporate culture, skills specific to internal
networks (team production), and the content of in-house training programs and on-the-job experience peculiar to the
firm. As Mailath and Postlewaite (1990, pp. 369-370) put it, a firm is “a network of people, each with an
understanding about how information and goods move within the firm. They know whom to contact about particular
problems that may arise and they know the strengths and weaknesses of their co-workers.”
The perception that star knowledge workers can take their talent with them to competing firms assumes that
their performance draws entirely on general human capital. This paper will address the question of when star
knowledge workers can in fact take their performance with them to competing firms. We will argue that the human
capital of star performers has a firm-specific component that cannot be easily transferred across firms. We will
consider how much one component of firm-specific human capital—the knowledge embedded in relationships with
colleagues—may contribute to the portability of a star’s performance. And we will address whether hiring stars
constitutes a value-creating or value-destroying proposition for firms.
We tested our hypotheses on star performers in the security-analyst market. Because an analyst’s core
activities, clients, and geographic location typically remain unchanged after changing employers, analysts and the
firms that employ them share a conviction that their performance is highly portable. As research executive Fred
Fraenkel put it, "Analysts are one of the most mobile Wall Street professions because their expertise is portable. I
4
mean, you’ve got it when you’re here and you’ve got it when you’re there. The client base doesn't change. You need
your Rolodex and your files, and you’re in business.”
We have compiled a unique data set that captures the performance and turnover of 799 equity analysts and
254 fixed-income analysts across 78 firms over a period of nine years. This longitudinal data allows us to capture
both short- and long-term changes in performance. The quality of our data presents a unique opportunity to advance
empirical knowledge on human capital and drivers of performance. Previous empirical investigations of portability of
performance have been hobbled by incomplete data on the workers in a given firm, job, or industry; performance that
is difficult to observe given the nature or structure of the jobs in question; and a low incidence of turnover (Chapman
and Southwick, 1991). Thus it has been impossible to disentangle the effects on worker productivity of human capital
and other factors (Garen, 1988). Measuring the performance of knowledge workers has proven particularly difficult
(e.g. Ernst and Vitt, 2000). What little performance data were available relied on self-reports or were limited to a
single firm, preventing cross-firm comparisons (Barrett and O'Connell, 2001). The external rankings we use in this
study provide a clear and public distinction between outstanding and ordinary performers, and the availability of rich
data at the demographic, departmental, firm-specific, sectoral, and macroeconomic levels makes it feasible to control
for all these variables together. Furthermore, few other studies have employed longitudinal data on individual
performance, which is crucial to disentangle individual and organizational contributions to performance.
Star Performers
Two factors distinguish star performers from average or merely competent performers. First, star performers are
disproportionately more productive. Studies of research-and-development scientists and academic researchers have
found that a few individuals are many times more productive than their colleagues (e.g., Cole and Cole, 1973; Ernst,
Leptein, and Vitt, 2000; Narin and Breitzman, 1995). In knowledge work, the ability and experience of top performers
are assets that cannot be compensated for by larger number of poorer performers, or by non-human assets (Narin,
1993; Rosen, 1981).
Second, stars are more visible in their labor markets than more ordinary performers. Any industry will have
many high-performing individuals, but the handful of superstars at the top will receive disproportionate attention from
competitors and clients (and, in some industries, the media), making their performance public and observable.
stocks, bonds, and other investment products for their clients' portfolios. Wall Street distinguishes between
investment banks and money-management firms by referring to investment banks as “the sell side” of the investment
process, and money-management firms as “the buy side.” (The buy side also employs analysts, whose research is
far less specialized; buy-side analysts thus rely heavily on sell-side research.) Our study focuses exclusively on sell-
side equity and fixed-income analysts, who work at investment banks.
Investment banks make money in three principal ways: merger-and-acquisition advisory services,
underwriting (raising funds for companies by issuing stocks or bonds), and trading of securities for institutional
clients. The ultimate purpose of research at an investment bank is to help generate business from trading securities
for institutional clients (Cowen et al., 2006). Security analysts write in-depth reports on the companies they cover and
issue recommendations as to whether investors should buy or sell the companies’ securities and bonds, or hold on to
stocks and bonds already in their portfolios. On average, an analyst writes 53.1 company-specific research reports
(two pages or longer) annually on 13.6 companies that they cover. It is not unusual for analysts to have more than
800 institutional clients who receive their reports and seek their advice.
An analyst’s compensation and visibility in the industry both depend on his or her ranking in an annual
survey conducted since 1972 by the trade journal Institutional Investor (II). Every year the magazine asks sell-side
analysts’ clients (portfolio managers and analysts at buy-side money-management firms, mutual funds, and pension
funds) to rate analysts on their effectiveness. Those ranked highest are celebrated in the magazine as members of
Institutional Investor’s All-America Research Team.1 The II survey reaches a large proportion of the investment
community. In October 1996, for example, II ranked analysts in each of 80 industry sectors and investment
specialties, based on ratings from roughly 1,300 individual respondents whose firms represented approximately 68
percent of the 300 largest institutional investors in the United States. These institutions include state pension funds,
which invest enormous pools of money, along with other investment-management institutions like Fidelity
Investments, Putnam Investments, and Wellington Management Company.
In each industry sector, the four analysts with the highest aggregate ratings are ranked first, second, third,
and runner-up. (More than four analysts may be ranked if there is a tie.) In 1996, less than 3 percent of all U.S. 1 We use the terms ranked analysts and star analysts interchangeably to refer to sell-side analysts ranked by Institutional Investor.
analysts were ranked. The II survey constitutes a core component of the internal performance evaluation for security
analysts at every investment bank. The II survey enables securities firms to determine how their most important
clients rate the service of their analysts (Siconolfi, 1992). Prior research has provided direct empirical evidence on
the relationship between rankings and forecast accuracy (e.g., Brown and Chen, 1991). Within the firm, being ranked
by II wins an analyst enhanced credibility, greater power, and substantially higher compensation (Dorfman, 1991). At
the major investment banks, top analysts could easily earn from $2 million to $5 million per year during the period of
our study (Laderman, 1998). In the larger financial world, II ranking brings heightened visibility and prestige.
Newspaper and TV reporters frequently seek out star analysts’ opinions on companies and stocks in their sectors.
There are many advantages to testing our hypotheses in the security analysts’ market. The II rankings
provide clear, observable, industry-accepted performance measures that are comparable across firms on a yearly
basis. Because star analysts enjoy high visibility in the financial industry, it is possible to compile thorough data on
each analyst’s performance and career history. The skills of security analysts are perceived in the labor market as
highly portable: the star power of top-performing analysts is attributed to individual ability that transcends
organizational support (Howard, 1967; Lurie, 1967; Institutional Investor, 1991). And even after moving to a new firm,
analysts continue to run the same financial models, follow the same sectors and companies, and maintain the same
client lists. (Institutional investors do not have exclusive relationships with particular investment banks; they receive
research reports from more than a dozen banks at the same time). For all these reasons, this market represents a
near-ideal sample against which to test our hypothesis that star performance is based in part on non-transferable,
firm-specific human capital.
Data
From the annual All-America Team issues of Institutional Investor published between 1988 and 1996, we collected
the following information for equity analysts (ranked in October) and fixed-income analysts (ranked in August): name;
industry specialty/sector; type (equity or fixed income); rank; year of the ranking; and company affiliation. The nine-
year period of the study produced a total of 4,200 analyst-year combinations (3,514 in equity and 686 in fixed
income). If each analyst were counted only once, the list would include 799 equity analysts and 254 fixed-income
analysts. Ranked analysts were employed by 78 investment banks; the 24 firms employing the most number of
14
ranked analysts accounted for 4,036 ranked analyst-year combinations (96 percent). To collect information on
analysts’ tenure with their current employers and total industry experience, we searched databases maintained by
Lexis-Nexis, the National Association of Securities Dealers, and Dow Jones News. Data on analysts’ tenure at their
firms were available for 3,639 analyst-year combinations (87 percent), and information on analysts’ experience was
collected for 3,653 analyst-year combinations (87 percent).
We accounted for every ranked analyst who left or joined an investment bank within one year of being
ranked during the period 1988–1996. Each of these incidents is identified as an “analyst-year move” in this paper.
Analysts’ affiliations in the year subsequent to being ranked, and the specific dates of their moves, were identified
using the databases of Nelson’s Directory of Investment Research, Lexis-Nexis, and Dow Jones News Service. We
identified 546 analyst-year moves: 500 were from one firm to another, and 46 were promotions or transfers to non-
research positions within the same investment bank. Of the 500 analyst-year moves from one firm, 134 were exits
from research altogether.2 The remaining 366 moves (made by 316 individual analysts) were switches to
competitors’ research departments. The average annual turnover of analysts in this sample was 11.9 percent.
News announcements on the hiring of ranked analysts indicated that ranked analysts occasionally moved in
teams, a phenomenon referred to by investment bankers as “block trading in people” or lift outs. Of the 366 analyst-
moves, 100 involved such colleagues as other ranked analysts, junior analysts, institutional salespeople, and traders.
Variables
Dependent Variable
The dependent variable in this study is the analyst’s performance, operationalized as rank in the Institutional Investor
All-America Research Team poll, Rankt+1. Each year the magazine’s editor sends a letter asking institutional
investors to rank the analysts who "have been most helpful to you and your institution in researching U.S. equities
over the past twelve months." Voters are asked to evaluate analysts using six criteria: earnings estimates, servicing
initiatives, accessibility and responsiveness, stock selection, industry knowledge, and written reports. The
respondents award a single overall numerical score to each analyst in each industry sector. Votes are cumulated 2 One hundred thirty-four analysts left sell-side research: 69 joined buy-side firms; 30 founded new companies; 20 retired; 8 took non-research positions in the securities industry; 5 joined a company that they had covered as analysts; one died in a car accident; and one died of a heart attack.
15
using weights based on the size of the voting institution. The identities of the respondents and their institutions are
kept confidential. A very small percentage of analysts achieve rankings in multiple sectors. Some, but not all, stars in
a given year continue to be ranked in subsequent years. In each sector, there are five levels of the dependent
variable (categorical and ordinal): the first, second, third, and runner-up ranks and unranked.
Independent Variables
Switching firms. We test whether the analyst’s movement across firms has an effect on his or her
Institutional Investor ranking. The independent variable of interest is the analyst’s mobility. The Analyst move variable
is 1 if a ranked analyst switched to another sell-side research department and 0 if a ranked analyst did not move
during the year.
Firm capability. For each of the 366 job changes, we determine whether the destination firm had lesser,
equivalent, or better capabilities relative to the originating firm by using the bulge-bracket distinction. Bulge-bracket
investment banks are those responsible for the bulk of securities underwritten in the United States; historically, the
six bulge firms are CS First Boston, Goldman Sachs, Lehman Brothers, Merrill Lynch, Morgan Stanley, and Salomon
Brothers. These banks and their employees enjoy a competitive edge due to the firms’ exceptional capabilities:
economies of scale in marketing, sales, and technology; a broad product base; wide distribution network for its
services, including analysts’ research products; and specialized units (investment banking, sales, trading, and
research) capable of solving complicated customer problems (Eccles and Crane, 1988).
We assign the job changes of ranked analysts to three groups: better firm capability, equal firm capability,
and lesser firm capability. The Firm capability direction categorical variable takes the value “no move” if a ranked
analyst does not move during the year, “moving to a weaker firm” if the move is from a bulge to a non-bulge firm,
“moving to a better firm” if the move is from a non-bulge to a bulge firm, and “moving to a comparable firm” if the
move is from bulge to bulge or non-bulge to non-bulge.
Team movement. We also test the effects of moving with a team of colleagues. Such a team can consist of
colleagues of several types. One is other security analysts who cover complementary sectors (such as the food-and-
beverage sector and the gaming-and-lodging sector), and who can benefit mutually from information sharing and
collaboration. Jacob, Lys, and Neale (1999) find that as the percentage of analysts at a firm who follow similar
dummy, Year dummy) are similar to coefficients in models M1 through M5. Therefore Table 3 only reports the
coefficients for the different categories of firm capabilities.
The portability of ranked analysts’ performance is examined by focusing on the sign and the significance
3 The Analyst star tenure coefficient is also affected by top-rated analysts’ inability to move up in rankings. Furthermore, using the Analyst new star variable, we examine whether being ranked for the first time affects performance. Analyst new star takes the value of 1 if an analyst has been ranked for the first time in a given year, and 0 otherwise. Newly ranked analysts might be on their way up or might merely have gotten lucky and are thus expected to lose ranking the next year. The Analyst new star coefficient is positive and insignificant. Finally, in order to reject alternative explanations for our results, we test whether there are significant interactions between Analyst move and Analyst firm tenure, Analyst experience, Analyst star tenure, Analyst performance and Analyst new star. None of the interactions are found to be significant.
20
level of the Firm capability direction variable. The findings suggest that star analysts who move between two
comparable firms (firms of equivalent capabilities) also exhibit a drop in performance but only for two years. Since the
origin and destination firms have similar capabilities, differences in firm capability cannot account for this drop in
performance. This finding provides evidence that the loss of firm-specific skills accounts for the diminished
performance of analysts who move between firms of comparable capabilities. During the first two years after
switching firms those analysts are able to acquire new firm-specific skills, and their performance in the third, fourth,
and fifth years is not statistically different from the performance of ranked analysts who stay put. Thus, Hypothesis 2
is supported for short-term performance.
Our findings suggest that performance decline is most pronounced among star analysts who move to firms
with weaker capabilities. Relative to comparable analysts who stay put, these analysts underperform for five years.
Stars who switch to firms with better capabilities exhibit no significant decline in short- or long-term performance. For
this category of analysts, the loss of firm-specific skills associated with decline in performance might have been
mitigated by their new employers’ valuable capabilities. Even after five years at a new firm, however, analysts who
switched to better firms are unable to outperform analysts who did not move. Star analysts who moved to firms with
worse capabilities appeared to experience the worst outcome among analysts who moved. If an analyst switches to a
firm with better capabilities, for instance, the probability of achieving the first rank in year t+1 is 0.083 (=0.106-0.023),
compared with a probability of 0.058 (=0.106-0.048) for an analyst who moves to a firm with worse capabilities.4
Team Movement and Ranked Analysts’ Short-Term and Long-Term Performance
Next we examine whether moving to a new firm solo or with a team has an effect on analyst performance (Rankt+1,
Rankt+2, Rankt+3, Rankt+4, and Rankt+5). Ranked analysts who move solo show an immediate decline in performance
that persists for at least five years (see Table 3). Switching firms has no significant effect, by contrast, on the
performance of ranked analysts who move with teammates. This finding suggests the value of skills embedded in
4 We also examined whether the effect of comparable moves differ between moves between bulge firms and moves between non-bulge firms, i.e. how better, comparable bulge, comparable non-bulge, or weaker affect performance (Rankt+1, Rankt+2, Rankt+3, Rankt+4, and Rankt+5). We find that star analysts who move between two comparable non-bulge firms exhibit a drop in performance for three years. Stars who switched between comparable bulge firms also exhibit a drop in performance but only for one year. Within the moving to comparable category (both bulge and non-bulge), most of the moves are between non-bulge bracket firms (78%). 22% of moves between comparable firms are moves between bulge firms.
21
colleague or team relationships. For instance, the probability of achieving the first rank in year t+1 is only 0.053 (=
0.106-0.061) for analysts who changed firms solo, but 0.100 (= 0.106-0.006) for those who switched employers with
teammates, suggesting the contribution of team-specific skills to an analyst’s performance. Thus, H3 is supported.5
Figure 1 provides graphical interpretation of these results for all independent variables (Analyst move, Firm
capability direction, Coworker move) for two individual years (t+1 and t+5).
Stock-Market Reaction to Announcements of Movement by Ranked Analysts
Table 4 presents stock-market reactions to announcements of moves by ranked analysts. The direction of cumulative
stock-market reactions is intriguing. The event itself evokes a significant negative stock-market reaction for the hiring
firm, suggesting that, in aggregate, the stock market views such a hire as value-reducing. Hiring firms experience
significant negative abnormal return (-0.74 percent) during the event period. The negative stock-market reaction for
the hiring firm is consistent with corporate-takeover studies suggesting that acquirers tend to suffer negative stock-
market reactions during the announcement period (Haspeslagh and Jemison, 1991).
On average, the size of the stock-market reaction to such a hiring announcement is a loss of about $24
million for the acquiring firm. Thus the negative stock-market reaction far exceeds the analyst’s entire compensation.
During the period of this study, even the highest reported salaries received by ranked analysts did not exceed $3
million per year for more than six years. If discounted, this amounts to just half of the stock-market reaction.
The magnitude of the abnormal returns depends on the type of move. Announcements of a ranked analyst’s
move to a firm with inferior capabilities evoke a strong and statistically significant negative stock-market reaction (-
1.67 percent). Announcements of moves between firms with equivalent capabilities are associated with a -0.63
percent return. Announcements of moves to firms with better capabilities elicit a smaller negative return (-0.26
percent). Clearly, the stock-price impact of hiring announcements by lesser firms is more severe than that of
comparable announcements by firms with equivalent or better capabilities (p < 0.05). Investors may believe that
inferior firms have to pay a high premium to entice a star analyst from a superior firm. Investors might also be
anticipating a drop in performance on the part of analysts who move to a firm with worse capabilities.
5 In our analysis, we are not measuring how many colleagues moved. We only measure whether any colleagues moved. We did use the linear data in our analysis, which did not meaningful affect the results.
22
Finally, abnormal returns are subdivided into two categories: (1) entries of individual analysts, or solo
entries, and (2) entries by analysts with other employees, or team entries. News of the hiring of an individual analyst
is greeted with significant negative abnormal returns (-0.62 percent), as is news of the hiring of an analyst with a
team (-0.99). Attracting a star along with his or her team appears to destroy more value, as measured by share price,
than hiring an individual star.
Discussion
Our results indicate that hiring stars is advantageous neither to stars themselves, in terms of their performance, nor
to hiring companies in terms of their market value. The performance of an outstanding worker is not owned by the
worker alone; it is a property of the worker/firm combination, and encompasses firm-specific human capital
embedded in colleague relationships and firm capabilities.
Star analysts who change employers suffer a decline in performance compared to that of analysts who stay
put, thus providing evidence of the importance of firm-specific human capital. Our findings support research on the
socialization of newcomers into organizations showing that such transitions are more disruptive and challenging than
anticipated. Newcomers often find that their expectations don't fit their actual experiences (Hill, 1992) and that their
status shifts from insider to outsider. Organizational socialization, or "learning the ropes," is an integral part of the
joining-up process (Van Maanen and Schein, 1979). Veteran employees, resentful of a newly hired star’s high
compensation and favored status, may withhold crucial information that is needed to succeed (Harding, 1998). A
star's performance would decline if long-tenured employees withheld cooperation in this way. Previous research has
shown that newcomers suffer a drastic decline in performance, especially if their success depends on the
contributions of others, as they cope with adjusting to an unfamiliar organizational setting (Louis, 1980).
Mobile analysts can preserve their performance to some extent under two conditions: when they move to a
firm with better capabilities and when they take firm-specific human capital with them in the form of existing colleague
relationships. The relative quality of the two firms affects a star’s post-move performance. This result supports
research that emphasizes the contribution of organizational context to individual performance (e.g., Allison and Long,
1990). Though our data do not pinpoint the specific components of firm capability that help mitigate performance
decline, research in various domains suggests that factors like a superior human-resource system (e.g., Huselid,
(0.076) (0.089) (0.098) (0.111) (0.130)- Predicted probabilities for analysts who do not move: First rank 0.106 0.150 0.156 0.157 0.154 Second rank 0.234 0.215 0.199 0.186 0.187 Third rank 0.356 0.279 0.237 0.211 0.193 Runner up 0.238 0.252 0.258 0.256 0.258 Unranked 0.066 0.105 0.150 0.191 0.207- Marginal change in probabilities for analysts who move:e
First rank -0.050 -0.053 -0.049 -0.047 -0.054 Second rank -0.064 -0.040 -0.033 -0.029 -0.036 Third rank -0.013 -0.008 -0.010 -0.011 -0.014 Runner up 0.071 0.046 0.031 0.022 0.022 Unranked 0.056 0.056 0.060 0.065 0.083 Avg. Effect 0.051 0.041 0.036 0.035 0.042Cut Point 1f -0.960 -1.194 -1.209 -0.882 -0.624Cut Point 2 0.030 -0.307 -0.407 -0.141 0.106Cut Point 3 0.954 0.408 0.197 0.398 0.602Cut Point 4 1.791 1.098 0.837 1.000 1.212Log (likelihood) -4413.808 -4006.221 -3381.890 -2844.371 -2075.233No. of observationsg 3511 2966 2451 2033 1501Pseudo-R2 0.211 0.157 0.140 0.125 0.130† p < 0.10, * p < 0.05, ** p < 0.01, *** p < 0.001Adjusted standard errors are in parentheses. In all models, the Wald chi-squared test is significant at p < 0.01.
e The marginal effect of the Analyst move variable is calculated as the discrete change in F(x) as this variable changes from 0 to 1: F(x=1) - F(x=0) f The categorical and ordinal dependent variable (Rank t ), represented by first rank, second rank, third rank, runner up, and unranked, generates four cutoff points. In Stata, the values of all cutpoints are estimated assuming that the intercept is 0. In contrast, LIMDEP fixes one of the cutpoints to 0 and estimates the intercept. In Stata, cutpoint 1 is LIMDEP's intercept, but with a reversed sign. The slope coefficients and predicted probabilities are the same under either estimation.
Models 1 through 5 examine the impact of switching firms on ranked analysts’ short-term and long-term performance. Each model is a robust cluster ordered probit specification with ranked analysts as clusters in which the dependent variable is analysts' Institutional Investor rankings. The robust cluster model adjusts standard errors for correlation because it assumes that data is not independent within clusters but independent across clusters. Only adjusted robust standard errors are reported.
g The number of observations decreases with each performance year, for two reasons. First, during the later years (t+1 through t+5), some analysts switched firms for a second time. To properly test whether analysts’ movements across firms affect their long-term performance, we include only analysts who remain with the same firms (“stayers”) for two years (t=0 and t+1) in the Rankt+2 models, three years (t=0, t+1, and t+2) in the Rankt+3 models, four years (t=0, t+1, t+2, and t+3) in the Rankt+4 models, and five years (t=0, t+1, t+2, t+3, and t+4) in the Rankt+5 models, and compare them with ranked analysts who switched firms(t=0) but then stayed with their new firms for one year (t+1) in the Rankt+2 models, two years (t+1, and t+2) in the Rankt+3 models, three years (t+1, t+2, and t+3) in the Rankt+4 models, and four years (t+1, t+2, t+3, and t+4) in the Rankt+5 models. In other words, we consider only performance after an analyst’s initial move during this period and not after any subsequent moves. If an analyst switches firms again during those years, his or her subsequent performance observations are not included. Finally, the number of observations decreases because of departures from the research business.
d The Analyst rank dummy t = 0 variable (first rank, second rank, third rank, and runner up) controls for the current analysts' rankings. Runner up is the base.
a To control for inter-temporal changes, Year dummy variable is used. b Sector dummy is used to control for industry effects. c Firm dummy controls for effects specific to investment banks.
Table 2Effect of Switching Firms on Ranked Analyst's Short-term and Long-term Performance
† p < 0.10, * p < 0.05, ** p < 0.01, *** p < 0.001Adjusted standard errors are in parentheses.
This table reports the coefficients for the different categories of move variables and their effects on ranked analysts’ short-term and long-term performance.
Firm
cap
abili
ty d
irect
ion
Cow
orke
r mov
e
This table examines if ranked analysts’ short-term and long-term performance is affected by whether 1) they move to a weaker firm, move to a comparable firm, or move to a better firm (the Firm capability direction variable); and 2) they move solo or in teams (the Coworker move variable).
Table 3Effect of Switching Firms on on Ranked Analysts’ Short-Term and Long-Term Performance by Direction and Type of Move
Moving to a better firm -0.26%d 0.004 28Moving to a comparable firm -0.63%c, e 0.004 53Moving to a weaker firm -1.67%a, d, e 0.004 20
Moving solo -0.62%b 0.003 70Moving in teams -0.99%b 0.004 31
a Different from 0 at 0.01 level of significance.b Different from 0 at 0.05 level of significance.c Different from 0 at 0.10 level of significance.
e Abnormal returns for moving to a weaker firm is different from abnormal returns for moving to a comparable firm at 0.05 level of significance.This table presents stock market reactions to announcements of movements by the ranked analysts. A short window study of daily excess returns over the event window of -1 to +1 days (market model) was conducted.
Stock-Market Reactions to Announcements of Moves by Ranked Analysts
CategoryAbnormal returns
Standard errors Sample size
Firm capability direction
Coworker move
d Abnormal returns for moving to a weaker firm is different from abnormal returns for moving to a better firm at 0.05 level of significance.
Table 4
Event window (-1, +1)
Firm
Entering firm
30
Figure 1Effect of Switching Firms on the Predicted Probability to Achieve First Rank by Direction and Type of Move: Year (t +1) and Year (t + 5)
This figure provides graphical interpretation of the probabilities of achieving a first rank for all independent variables (Analyst move, Firm capability direction, and Coworker move ) for two individual years (t+1 and t+5).
00.020.040.060.080.1
0.120.140.16
Firs
t Ran
k Pr
obab
ility
Probability of Achieving First Rank (t +1): Stay vs. Move
Ranked analystswho did not changefirms
Ranked analystswho changed firms
0
0.04
0.08
0.12
0.16
Firs
t Ran
k Pr
obab
ility
Probability of Achieving First Rank (t +1): Stay vs. Moving Solo or with Teammates
Ranked analystswho did not changefirms
Ranked analystswho changed firmswith teammates
Ranked analystswho changed firmssolo
00.020.040.060.080.1
0.120.140.16
Firs
t Ran
k Pr
obab
ility
Probability of Achieving First Rank (t +5): Stay vs. Move
Ranked analystswho did not changefirms
Ranked analystswho changed firms
0
0.04
0.08
0.12
0.16
Firs
t Ran
k Pr
obab
ility
Probability of Achieving First Rank (t +5): Stay vs. Moving to a Comparable Firm
Ranked analystswho did not changefirms
Ranked analystswho moved tocomparable firms
0
0.04
0.08
0.12
0.16
Firs
t Ran
k Pr
obab
ility
Probability of Achieving First Rank (t +5): Stay vs. Moving Solo or with Teammates
Ranked analystswho did not changefirms
Ranked analystswho changed firmswith teammates
Ranked analystswho changed firmssolo
0
0.04
0.08
0.12
0.16
Firs
t Ran
k Pr
obab
ility
Probability of Achieving First Rank (t +1): Stay vs. Moving to a Comparable Firm
Ranked analystswho did not changefirms
Ranked analystswho moved tocomparable firms
31
References Acemoglu D., Pischke J.-S. 1998. Why Do Firms Train? Theory and Evidence. The Quarterly Journal of Economics
113(1) 79-119. Allison P. D., Long J. S. 1990. Departmental Effects on Scientific Productivity. American Sociological Review 55(4)
469-478. Asquith P., Mullins D. 1986. Equity Issues and Offering Dilution. Journal of Financial Economics 15(1/2) 61-89. Baks K. P. 2003. On the performance of mutual fund managers. Emory University, Unpublished paper. Barney J. 1991. Firm Resources and Sustained Competitive Advantage. Journal of Management 17(1) 99-120. Barrett A., O'Connell P. J. 2001. Does training generally work? The returns to in-company training. Industrial and
Labor Relations Review 54(3) 647-662. Becker G. S. 1962. Investment in Human Capital: A Theoretical Analysis. The Journal of Political Economy 70(5) 9-
49. Becker G. S. 1976. The economic approach to human behavior. University of Chicago Press, Chicago. Berman S. L., Down J., Hill C. W. L. 2002. Tacit Knowledge as a Source of Competitive Advantage in the National
Basketball Association. Academy of Management Journal 45(1) 13-31. Bradley M., Desai A., Kim E. H. 1988. Synergistic gains from corporate acquisitions and their division between the
stockholders of target and acquiring firms. Journal of Financial Economics 21(1) 3-40. Brindisi L. J. 1985. Shareholder Value and Executive Compensation. Planning Review 13(5) 14-17. Brown L., Chen D. 1991. How good is the All-America Research Team in forecast earnings? Journal of Business
Forecasting 9 14-18. Brown S. J., Warner J. B. 1980. Measuring Security Price Performance. Journal of Financial Economics 8(3) 205-
258. Burt R. S. 1987. Social Contagion and Innovation: Cohesion Versus Structural Equivalence. The American Journal of
Sociology 92(6) 1287-1335. Cassing J., Douglas R. W. 1980. Implications of the Auction Mechanism in Baseball's Free Agent Draft. Southern
Economic Journal 47(1) 110-121. Castanias R., Helfat C. 1991. Managerial Resources and Rents Journal of Management 17(1) 155-171. Chapman K., Southwick L. 1991. Testing the Matching Hypothesis: The Case of Major League Baseball. American
Economic Review 89(5) 1352-1360. Chaudhuri S., Tabrizi B. 1999. Capturing the Real Value in High-Tech Acquisitions. Harvard Business Review 77(5)
123-130. Clark K. B., Fujimoto T. 1991. Product development performance: Strategy, organization, and management in the
world auto industry. Harvard Business School Press, Boston, MA. Cohen W. M., Levinthal D. A. 1990. Absorptive Capacity: A New Perspective on Learning and Innovation.
Administrative Science Quarterly 35(1) 128-152. Cole J. R., Cole S. 1973. Social stratification in science. University of Chicago Press, Chicago. Cowen A.P., Groysberg B., Healy P. 2006. Which Types of Analyst Firms Are More Optimistic? Journal of Accounting
and Economics 41(1-2) 119-146. Cymrot D. J., Dunlevy J. A. 1987. Are Free Agents Perspicacious Peregrinators? The Review of Economics and
Dorfman J. 1991. Analysts devote more time to selling as firms keep scorecard on performance. Wall Street Journal October 29 1.
Drucker P. F. 1998. On the Profession of Management. Harvard Business School Press, Cambridge, Mass. Drucker P. F. 1999. Management Challenges for the 21st Century. Harper Business, New York. Dyckman T., Philbrick D., Stephan J. 1984. A Comparison of Event Study Methodologies Using Daily Stock Returns:
A Simulation Approach. Journal of Accounting Research 22(Supplement) 1-30. Eccles R. G., Crane D. B. 1988. Doing deals: Investment banks at work. Harvard Business School Press, Boston,
Mass. Ernst H., Leptien C., Vitt J. 2000. Inventors are not alike: The distribution of patenting output among industrial R & D
personnel. IEEE Transactions on Engineering Management 47(2) 184-199. Ernst H., Vitt J. 2000. The influence of corporate acquisitions on the behaviour of key inventors. R and D
Management 30(2) 105-120. Felin T., Hesterly W. S. forthcoming. The Knowledge-Based View, Heterogeneity, and New Value Creation:
Philosophical Considerations on the Locus of Knowledge. Academy of Management Review. Friedberg R. M. 2000. You Can´t Take It With You? Immigrant Assimilation and the Portability of Human Capital.
Journal of Labor Economics 18(2) 221-251. Galbraith J. R. 1977. Organization design. Addison-Wesley, Reading, Mass. Garen J. E. (Ed.). 1988. Empirical Studies of the Job Matching Hypothesis. JAI Press, Greenwich. Glenn A., McGarrity J. P., Weller J. 2001. Firm-specific human capital, job matching, and turnover: Evidence from
major league baseball, 1900-1992. Economic Inquiry 39(1) 86-93. Greene W. H. 2000. Econometric analysis (4th ed.). Prentice Hall, Upper Saddle River, N.J. Greenwood R., Hinings C. R., Brown J. 1990. "P2-Form" Strategic Management: Corporate Practices in Professional
Partnerships. The Academy of Management Journal 33(4) 725-755. Groysberg B., Abrahams, R. 2006. Lift Outs: How to Acquire a High-Functioning Team. Harvard Business Review 84
(12) 133-140. Groysberg B., Lee L.-E. forthcoming. The Effects of Colleague Quality on Top Performance: The Case of Security
Analysts. Journal of Organizational Behavior
Hackman J. R. 1990. Groups that work (and those that don't): Creating conditions for effective teamwork (1st ed.). Jossey-Bass, San Francisco.
Harding F. 1998. Creating rainmakers: The manager’s guide to training professionals to attract new clients. Adams Media Corporation, Holbrook, Mass.
Haspeslagh P., Jemison D. 1991. Managing Acquisitions. Free Press, New York. Heckman J. J. 1979. Sample Selection Bias as a Specification Error. Econometrica 47(1) 153-161. Henderson R., Cockburn I. 1994. Measuring Competence? Exploring Firm Effects in Pharmaceutical Research.
Strategic Management Journal 15(Special Issue: Competitive Organizational Behavior) 63-84. Henning J. F. 1993. Maximizing Law Firm Profitability: Hiring, Training and Developing Productive Lawyers. Law
Journal Seminars-Press, New York. Hill L. A. 1992. Becoming a Manager: Mastery of a New Identity. Harvard Business School Press, Cambridge, Mass. Howard G. G. 1967. What Makes a Good Analyst? Well... Institutional Investor May 10-15.
Huber P. J. 1967. The behavior of maximum likelihood estimates under non-standard conditions, Proceedings of the Fifth Berkeley Symposium on Mathematical Statistics and Probability, Vol. 1 221-233.
Huckman R. S., Pisano G. P. 2006. The Firm Specificity of Individual Performance: Evidence from Cardiac Surgery. Management Science 52(4) 473-488.
Hunter J. E., Schmidt F. L., Judiesch M. K. 1990. Individual differences in output variability as a function of job complexity. Journal of Applied Psychology 75(1) 28-42.
Huselid M. A. 1995. The impact of human resource management practices on turnover, productivity, and corporate financial performance. Academy of Management Journal 38(3) 635-672.
Ibarra H. 1995. Race, Opportunity, and Diversity of Social Circles in Managerial Networks. The Academy of Management Journal 38(3) 673-703.
Ichniowski C., Shaw K., Prennushi G. 1997. The Effects of Human Resource Management Practices on Productivity: A Study of Steel Finishing Lines. The American Economic Review 87(3) 291-313.
Institutional Investor. 1991. The 1991 All-America Fixed Income Research Team. Institutional Investor October 81-87.
Jacob J., Lys T. Z., Neale M. A. 1999. Expertise in forecasting performance of security analysts. Journal of Accounting and Economics 28(1) 51-82.
Kaplan S., Murray F., Henderson R. 2003. Discontinuities and senior management: assessing the role of recognition in pharmaceutical firm response to biotechnology. Industrial and Corporate Change 12(2) 203-233.
Kim L. S. 1997. The dynamics of Samsung's technological learning in semiconductors. California Management Review 39 86-100.
Kitayama S., Markus H. R., Matsumoto H., Norasakkunkit V. 1997. Individual and collective processes in the construction of the self: Self-enhancement in the United States and self-criticism in Japan. Journal of Personality and Social Psychology 72(6) 1245-1267.
Kogut B., Zander U. 1996. What firms do? Coordination, identity, and learning. Organization Science 7 502-518. Kordana K. A. 1995. Law Firms and Associate Careers: Tournament Theory Versus the Production-imperative
Model. The Yale Law Journal 104(7) 1907-1934. Krautmann A. C., Oppenheimer M. 1994. Free Agency and the Allocation of Labor in Major League Baseball.
Managerial & Decision Economics 15(5) 459-469. Laderman J. M. 1998. Who Can You Trust? Wall Street's Spin Game. Business Week October 5 148-156. Lazear E. P. 1986. Raids and offer matching. Research in Labor Economics 8(A) 141-165. Lippman S. A., Rumelt R. P. 1982. Uncertain Imitability: An Analysis of Interfirm Differences in Efficiency under
Competition. The Bell Journal of Economics 13(2) 418-438. Long J. S., McGinnis R. 1981. Organizational Context and Scientific Productivity. American Sociological Review
46(4) 422-442. Louis M. R. 1980. Surprise and Sense Making: What Newcomers Experience in Entering Unfamiliar Organizational
Settings. Administrative Science Quarterly 25(2) 226-251. Lurie S. B. 1967. What makes a good analyst? Institutional Investor July 33, 42, 44. MacDonald D. N., Reynolds M. O. 1994. Are Baseball Players Paid their Marginal Products? Managerial & Decision
Economics 15(5) 443-457. Maddala G. S. 1983. Limited-dependent and qualitative variables in econometrics. Cambridge University Press,
Mailath G. J., Postlewaite A. 1990. Working Versus Firms: Bargaining Over a Firm's Value. Review of Economic Studies 57(191) 369-380.
Maister D. H. 1993. Managing the Professional Service Firm. The Free Press, New York. March J. G., Simon H. A. 1958. Organizations. Wiley, New York. McGregor J. 2006. I can’t believe they took the whole team. Business Week December 18. Mikhail M. B., Walther B. R., Willis R. H. 1997. Do Security Analysts Improve Their Performance with Experience?
Journal of Accounting Research 35(Supplement) 131-157. Mitchell M. L., Netter J. M. 1994. The role of financial economics in securities fraud cases: applications at the
Securities and Exchange Commission. 49(n2) 545-590. Nahapiet J., Ghoshal S. 1998. Social Capital, Intellectual Capital, and the Organizational Advantage. Academy of
Management Review 23(2) 242-266. Nakosteen R. A., Zimmer M. 1980. Migration and Income: The Question of Self-Selection. Southern Economic
Journal 46(3) 840-851. Narin F. 1993. Technology indicators and corporate strategy. Review of Business 14(3) 19-23. Narin F., Breitzman A. 1995. Inventive productivity. Research Policy 24(4) 507-519. Nelson R., Winter S. 1982. An Evolutional Theory of Economic Change Harvard University Press, Cambridge, MA. Nelson’s Directory of Investment Research. 1988-1998. Nelson Information, Port Chester, NY. Peteraf M. A. 1993. The Cornerstones of Competitive Advantage: A Resource-Based View. Strategic Management
Journal 14(3) 179-191. Pfeffer J. 1994. Competitive advantage through people: Unleashing the power of the work force. Harvard Business
School Press, Boston, Mass. Pfeffer J. 1998. The human equation: Building profits by putting people first. Harvard Business School Press, Boston,
Mass. Puranam P., Singh H., Zollo M. 2003. A Bird in the Hand or Two in the Bush?: Integration Trade-offs in Technology-
grafting Acquisitions. European Management Journal 21(2) 179-184. Puranam P., Singh H., Zollo M. 2006. Organizing for Innovation: Managing the Coordination-Autonomy Dilemma in
Technology Acquisitions. Academy of Management Journal 49(2) 263-280. Ravid S. A. 1999. Information, Blockbusters, and Stars: A Study of the Film Industry. The Journal of Business 72(4)
463-492. Rees A. 1973. The Economics of Work and Pay. Harper and Row, New York. Robinson C., Tomes N. 1982. Self-Selection and Interprovincial Migration in Canada. The Canadian Journal of
Economics 15(3) 474-502. Roll R. 1986. The Hubris Hypothesis of Corporate Takeovers. The Journal of Business 59(2) 197-216. Rosen S. 1981. The Economics of Superstars. The American Economic Review 71(5) 845-858. Sessa D. 1999. All-Star Analysts 1999 Survey: Early Mornings, Late Nights Mark an Analyst’s Days. The Wall Street
Journal June 29 R13. Siconolfi M. 1992. Shearson Research Analysts Finish First on 'All-America Team' for Third Year. Wall Street Journal
October 13 C18. Simonton D. K. 1999. Talent and Its Development: An Emergenic and Epigenetic Model. Psychological Review
106(3) 435-457. Sinha P., Brown L. D., Das S. 1997. A re-examination of financial analysts' differential earnings forecast accuracy.
Contemporary Accounting Research 14(1) 1-42. Sirower M. 1997. The Synergy Trap: How Companies lose the Acquisition Game. The Free Press, New York. Song J., Almeida P., Wu G. 2003. Learning-by-Hiring: When Is Mobility More Likely to Facilitate Interfirm Knowledge
Transfer? Management Science 49(4) 351-365. Spence M. A. 1974. Market signaling: Informational transfer in hiring and related screening processes. Harvard
University Press, Cambridge. Topel R. 1991. Specific Capital, Mobility, and Wages: Wages Rise with Job Seniority. The Journal of Political
Economy 99(1) 145-176. Van Maanen J., Schein E. H. 1979. Toward of Theory of Organizational Socialization. In BM Staw (Ed.), Research in
Organizational Behavior, Vol. 1: 209-264. Warner J. B., Watts R. L., Wruck K. H. 1988. Stock prices and top management changes. Journal of Financial
Economics 20(1) 461-492. Weick K. E., Roberts K. H. 1993. Collective Mind in Organizations: Heedful Interrelating on Flight Decks.
Administrative Science Quarterly 38(3) 357-381. Weisman R. 2006. The business of lift outs. The Boston Globe November 19 B1. Weisman R. 2006. Once taboo, hiring away teams is getting popular. The Boston Globe November 19 D1. White, H. 1982. Maximum likelihood estimation of misspecified models. Econometrica 50(1) 1-25. Williamson O. E. (Ed.). 1986. Transaction-Cost Economics: The Governance of Contractual Relations. Jossey-Bass
Publishers, California. Zander U., Kogut B. 1995. Knowledge and the speed of the transfer and imitation of organizational capabilities.
Organization Science 6 76-92. Zollo M., Singh H. 2004. Deliberate Learning in Corporate Acquisitions: Post-acquisition Strategies and Integration
Capability in U.S. Bank Mergers. Strategic Management Journal 25(13) 1233-1256. Zuckerman H., Merton R. K. 1971. Patterns of evaluation in science: Institutionalisation, structure and functions of the