Relying on peer-effects theory in the labor economics and management literature, we examine whether peer effects exist among regular (non-star) sell-side financial analysts, who represent the majority of employees in the sell-side industry (Mas and Moretti 2009; Chan, Li, and Pierce 2014). Research in organizational behavior and applied psychology research suggests that lateral relationships are linked to important individual employee outcomes (e.g., Chiaburu and Harrison 2008; Kim and Yun 2015). The contributions of higher-status individuals are often given too much weight while those of lower-status individuals are often overlooked. Similarly, the all-star analyst is recognized as having a superior social status in the sell-side analyst labor market. To address the issues of status differences among analysts, in this study we focus on the effect of non-star analyst transitions on their equal-status counterparties—non-star incumbent analysts. To make it clear, we focus on incumbent analysts who remain in the same brokerage house but experience changes in their peers, which arguably is an exogenous shock to their current knowledge base.
As pointed out by Hasan and Koning (2019), an essential driver of peer effects identified in the literature is spatial proximity to coworkers and peers who may possess diverse knowledge or skills. While the magnitude of these effects differs across contexts, proximate peers rather than distant ones are more likely to shape performance. Consistent with this argument, we make use of the fact that analysts are industry experts to measure proximity among regular analysts. Knowledge sharing is more likely to occur if two analysts share common industry coverage. Intuitively, an analyst covering the technology industry is more likely to gain relevant industry knowledge from a colleague who covers the same industry than from a colleague who covers a completely different industry.
Different from all-star analysts who may impact a larger group of incumbents in the brokerage, regular analysts are more likely to influence a smaller group of coworkers who interact with them. Accordingly, we compare the performance of incumbent analysts who cover at least one of the same industries as the transiting analyst (“affected incumbent analysts”) with the performance of those who have no overlapping industry coverage as the transiting analyst (“unaffected incumbent analysts”). By comparing the incumbents of different groups within the same broker (and for the same time periods), we mitigate the possibility that structural or cultural changes at the broker level lead to general performance differences in incumbents employed at different brokers. We define an analyst transition as the case where the transiting analyst leaves employment at one brokerage house and finds employment at another brokerage house. This requirement addresses the issue that the transiting analyst could be a rookie analyst without sell-side industry experience, or that the turnover decision is involuntary. Additionally, this definition ensures that a transiting analyst comes from the sell-side industry rather than a different industry (e.g., corporate firms, credit rating agencies, or banks), which satisfies our “transitions within sell-side industry” requirement.
To construct our sample, we start with quarterly earnings forecasts for U.S. firms from the I/B/E/S database over the period 1994 to 2018. We identify an analyst transition when we observe a change in the broker identifier that the analyst is associated with. We employ two standardized measures of analyst forecasting performance: accuracy and timeliness. Our results show that regular analyst transitions do not have a statistically significant impact on all analysts in a particular broker on average. Instead, transiting analysts affect incumbents of different groups in the same broker in different ways. Specifically, we find that affected incumbents issue more accurate and timely forecasts than unaffected incumbents after a transiting analyst arrives. Similarly, affected incumbent analysts issue less accurate but not less timely forecasts than unaffected incumbents after an analyst departs. The findings indicate that affected incumbent analysts in the new broker obtain valuable knowledge from the transiting analyst and produce information more accurately and quickly.
Motivated by prior studies on horizontal knowledge sharing, we next examine potential channels of knowledge spillover (Wang and Noe 2010). We find that when a transiting analyst switches from a larger broker to a smaller broker, she can exert a more significant impact on the affected incumbents’ performance in the new broker. This finding suggests that transiting analysts possibly share their practice and experience from the larger broker to incumbents in the new broker. Moreover, we show that the knowledge spillover effect is more pronounced when transiting analysts cover multiple industries. Larger industry scope creates more knowledge-sharing opportunities for affected incumbents. For example, when a transiting analyst covers three industries, two of which share common coverage with incumbents, then incumbents can obtain knowledge about two industries rather than one industry if the transiting analyst only covers one industry. Finally, we find that the knowledge spillover effect is more salient when transiting analysts cover geographically linked firms as incumbents, indicating that transiting analysts may also spillover local economic knowledge to incumbents.
Our study offers implications for academics, practitioners, and investors. First, the paper adds to the growing literature that examines how colleagues’ quality can affect analyst forecasting accuracy. Instead of relying on unidentified connections between analysts, we utilize a setting where incumbent analysts’ peers change with analyst transitions. Additionally, the transition effect on incumbents’ subsequent performance can be directly identified. We add to this line of literature by showing that the transition of regular analysts can impact incumbent analysts’ performance.
Next, we provide a potential explanation for high turnovers in the sell-side equity analyst industry. The analyst labor market is a knowledge-intensive market, where analysts collect industry-related and firm-specific information to produce research output. While practitioners suggest that industry knowledge is perhaps the most important quality an analyst can possess and is critical to an analyst’s job, there is little systematic evidence on how industry knowledge affects analyst performance, possibly because industry knowledge is inherently difficult to measure. We shed light on how analysts can transfer their industry knowledge into value-added information to peers. Affected incumbent analysts in the new (or old) broker benefit (or suffer) from the portable and transferrable information sets owned by the transiting analyst. There are also possible implications for brokerage-house hiring decisions. For example, brokers will potentially enjoy more benefits if they absorb new analysts who cover firms or industries related to the coverage portfolios of their incumbent analysts.
Moreover, our study extends research on the consequences of analyst job changes. Most prior studies focus on the subsequent performance of transiting analysts whose turnover decisions are endogenous. It is challenging to establish a causal relation between an analyst’s turnover and her subsequent performance changes. To mitigate this inherent endogeneity, later studies use broker closures or mergers to examine the impact of analyst job changes on forecasting performance. However, mergers and closures of brokerage houses are rare events, while analyst transitions happen regularly. Focusing on incumbent analysts (and employing analyst, broker, and quarter fixed effects) alleviates endogeneity because the incumbents’ subsequent performance is more likely to be influenced by an analyst transition instead of the other way around.
Ole-Kristian Hope is the Deloitte Professor of Accounting at the University of Toronto Rotman School of Management
Xijiang Su is a PhD student at the University of Toronto Rotman School of Management
This post is adapted from their paper, “Peer-Level Analyst Transitions” available on SSRN.
