Abstract
We investigate the impact of observing peers’ information acquisition on financial analysts’ allocation of attention. Using the timely disclosure mandate by the Shenzhen Stock Exchange as a setting, we find that, shortly after analysts observe that a firm has been visited by peer analysts, they reduce short-term attention to that firm, as indicated by a reduced tendency to conduct follow-up visits. Nonvisiting analysts who do not conduct follow-up visits are more likely to discontinue coverage of the visited firm. These findings are consistent with the conjecture that the timely disclosure reveals the first-mover advantage of visiting analysts, leading nonvisiting ones to reallocate their limited attention. We also find that, compared to the pre-mandate period, the information environments of visited firms deteriorate immediately after an analyst’s visit but not over the longer term. Further evidence suggests that the timely disclosure mandate has positive externalities in the form of increased immediate attention to and improved short-term information environments of unvisited peer firms.
1 Introduction
Financial analysts acquire information from public and private sources and facilitate communication between firms and investors (Healy and Palepu [2001], Kadan et al. [2009], Loh and Stulz [2018]). Research has shown that analysts allocate their limited attention and resources across firms, which has important implications for investors, companies, and the capital market (deHaan, Shevlin, and Thornock [2015], Harford et al. [2019], Blankespoor, deHaan, and Marinovic [2020], Driskill, Kirk, and Tucker [2020]). Analysts strive to provide new information and insights to distinguish themselves from other analysts (e.g., Hong and Kacperczyk [2010], Crawford, Roulstone, and So [2012], Merkley, Michaely, and Pacelli [2017]). What analysts know about their peers’ information acquisition can thus significantly change how they choose to allocate their attention across firms, a topic about which researchers know little.
This study examines how observing peers’ allocation of attention affects analysts’ own allocation of attention. Since 2012, the Shenzhen Stock Exchange (SZSE) has required listed firms to promptly disclose private meetings with corporate management. This requirement allows analysts to observe information acquisition by peers. Private meetings enable analysts to acquire information by talking with managers and employees and observing operations and facilities (Brown et al. [2015], Chen et al. [2022], Choy and Hope [2024]). In most locales, analysts cannot observe when and with whom their peers meet (Soltes [2014], Solomon and Soltes [2015]). However, the SZSE mandate enables analysts to note their peers’ visits and respond.
Prior studies document that analysts who privately meet with managers gain an information advantage and make better forecasts (Cheng et al. [2016], Han, Kong, and Liu [2018], Choy and Hope [2024]). When a firm promptly discloses analysts’ visits, investors and nonvisiting analysts can see that the visiting analysts have gained an advantage vis-à-vis that firm.1 The visitors have become better informed, and investors will be inclined to pay more attention to their earnings forecasts or private communications. Nonvisiting analysts will then likely expect a reduction in both investors’ requests for information and their chances of discovering new information through their own visits. This may lead them to reduce their attention to a firm a peer has recently visited.2
However, nonvisiting analysts may not reduce their attention to a visited firm. Corporate site visits often occur when a firm has value-relevant information that is not recognized or understood by the public (e.g., Cheng et al. [2019], Choy and Hope [2024]).3 Timely disclosure of visits might underscore the existence of this sort of information, attracting visits from other analysts, who can then form different information mosaics (e.g., Brown et al. [2015], Solomon and Soltes [2015]).4 Observing other analysts’ site visits may thus increase the expected benefits of paying more attention to a given firm. It is therefore an open question whether timely disclosure of corporate site visits affects analysts’ attention allocation.
We use the SZSE's mandate that corporate site visits must be promptly disclosed in an empirical setting. In July 2012, the stock exchange began to require this disclosure within two trading days of a visit, and the disclosure must include participants’ identities, the meeting date, location, and descriptions of meeting topics (hereafter referred to as the “timely disclosure mandate” or “the mandate”). Before 2012, this information needed only to be disclosed in these firms’ regular periodic reports (i.e., annual, semi-annual, and quarterly), and the lag between a visit and disclosure could stretch to several months. Afterward, analysts could observe their peers’ information acquisition much sooner.5
We use a difference-in-differences research design, comparing changes in analysts’ attention to firms visited by other analysts from 2009–2011 to 2013–2015, relative to unvisited firms. We are interested in analysts’ decisions to allocate attention to a certain firm, and visits by other analysts at firms that are not in the analyst's portfolio are unlikely to influence this attention allocation. We therefore focus on firms covered by the analyst in the 52 weeks prior to each week of interest. We find that, following the timely disclosure mandate, analysts are less likely to subsequently visit a firm when the firm has been visited by other analysts during a week, compared with unvisited firms that same week. These results suggest that the timely disclosure mandate reveals the information advantage of the visiting analyst, leading nonvisitors to reduce their attention to visited firms. We also find that, prior to 2012, there were no differential trends in nonvisiting analysts’ attention allocated to visited firms, versus unvisited ones, supporting the parallel trends assumption. We further document a larger reduction in attention to visited firms, relative to unvisited ones, when visiting analysts have stronger information advantages.
We next examine whether the reduction in analysts’ attention has consequences for the information environments of visited firms. It is unclear whether visited firms’ information environments would change. On the one hand, the decreased attention from nonvisiting analysts may harm the information environment in the short term. On the other hand, visiting analysts may share insights that aid price discovery, so that the firm's information environment may not change. We follow the literature and use two measures of the information environment. One is the change in forecast accuracy around visits from 26 weeks before to two (or four) weeks after the visit week (e.g., Cheng et al. [2016], Chen et al. [2021]). The second is stock return synchronicity in the subsequent two (or four) weeks (e.g., Durnev et al. [2003], Crawford, Roulstone, and So [2012]). Our results suggest that firms’ short-term information environments deteriorate due to decreased attention from nonvisiting analysts. Specifically, the change in forecast accuracy around visits declines and stock return synchronicity increases from the pre- to the post-period.
Motivated by our finding that nonvisiting analysts reduce attention to visited firms, we next explore whether the timely disclosure of visits has consequences for unvisited firms. Timely disclosure of visits points to the existence of value-relevant information not only for visited firms but also for their industry peers. We find that, following the mandate, analysts increase their attention to unvisited firms when more peer firms have hosted visits. Unvisited firms benefit from more accurate analyst forecasts and lower stock price synchronicity. These results suggest that the timely disclosure mandate has positive externalities in the form of increased analyst attention to and improved information environments of unvisited peers.
Although the nature of the timely disclosure mandate implies a prompt response by analysts after a visit, it is natural to ask whether the short-term effect endures. We find that, from the pre- to the post-period, nonvisiting analysts do not change their tendency to visit the visited firm beyond the initial eight weeks after the visit week. However, nonvisiting analysts—in particular, those who do not conduct follow-up visits—are more likely to discontinue their coverage in the post-period. These results suggest that, in response to the timely disclosure of peer analysts’ visits, nonvisiting analysts who do not conduct follow-up visits are likely to reassess their long-term plans promptly and discontinue coverage of some firms. This increased tendency to drop coverage could cause visited firms’ information environments to deteriorate in the long run. However, if analysts who continue coverage maintain their information discovery efforts, those efforts may compensate for the reduction in analyst coverage. Therefore, it is unclear how firms’ information environments might evolve. We find that the information environments of visited firms do not deteriorate in the long run from the pre- to the post-period. Despite a deterioration of the short-term information environment and increased analyst coverage discontinuation, firms may still benefit from accommodating analyst site visits in the post-period. First, firms may use site visits to communicate information, so these visits continue to contribute to information discovery. Second, coverage tends to be more favorable by the visiting analysts. We find evidence consistent with these predictions.6 Additionally, accommodating site visit requests may help management build relationships with analysts and gain access to their buy-side clients (Bengtzen [2017]).
We perform additional analyses to address potential alternative explanations. First, firms’ responses might explain our findings. Although analysts initiate most visits (Solomon and Soltes [2015], Cheng et al. [2016]), firms can negotiate dates and combine visit requests from multiple analysts into one visit, which allows them to report once and lower compliance costs. However, we find no evidence of this driving our main results. Relatedly, firms occasionally invite analysts in, and these visits typically occur in the month after major corporate events, such as earnings announcements (Cheng et al. [2016], [2019]). Our inferences do not change if we exclude these visits from our sample. Next, although our proposed explanation relies on firms’ timely disclosure of meeting times and visitors’ identities, firms must also disclose meeting topics, which could confound our results. To mitigate this concern, we examine whether nonvisiting analysts learn useful information from the timelier topic disclosures. We do not find any evidence of this, even for firms that disclose more information, following the mandate.
Our paper contributes to the growing literature on how analysts allocate their attention and resources (Harford et al. [2019], Blankespoor, deHaan, and Marinovic [2020], Driskill, Kirk, and Tucker [2020], Hsu et al. [2023]). Rational inattention models suggest that economic agents must decide how to allocate their limited information processing capacity across firms (e.g., Sims [2003], Sims [2010], Veldkamp [2011]).7 Consistent with this, Driskill, Kirk, and Tucker [2020] find that analysts are less timely and thorough when their covered firms have contemporaneous earnings announcements. Harford et al. [2019] find that analysts strategically allocate more effort to portfolio firms that are more important to their careers. We identify a new factor that affects analysts’ attention allocation—information about peers’ activities. To capture analyst attention, we focus on the site visit, an input-based measure that is typically unobservable. This approach differs from prior studies using output-based measures, such as the timing of forecasts or forecast accuracy. Our study helps illuminate how analysts fulfill their information intermediary role.
This paper extends the literature on strategic interactions among analysts. Studies document herding (e.g., Clement and Tse [2005], Jegadeesh and Kim [2010]) and social learning (e.g., Do and Zhang [2020], Chen, Mayew, and Yan [2024], Huang, Lin, and Zang [2022], Kumar, Rantala, and Xu [2022]) among analysts and examine how competition shapes their forecasts and recommendations (e.g., Hong and Kacperczyk [2010], Merkley, Michaely, and Pacelli [2017]). In comparison, we show that competition also affects how they allocate their attention. Our paper relates closely to the work of Crawford, Roulstone, and So [2012], who study analysts’ coverage initiation. They show that the first analyst who initiates coverage provides low-cost market and industry information, while subsequent analysts doing so for the same firm typically focus on firm-specific information to distinguish themselves. Our results suggest that, when firm-specific information has been exploited by other analysts, analysts choose to reduce attention to the firm.
Our paper also contributes to the understanding of analyst coverage choices, which is crucial for deciphering analyst activities (Bradshaw [2011]). Analysts weigh the benefits and costs of following a firm (Barth, Kasznik, and Mcnichols [1999]). The literature suggests that firm disclosure can either complement or substitute for analyst coverage. Improved disclosure can lower costs for analysts, leading to increased coverage, as evidence shows that analyst coverage decisions relate positively to the informativeness of disclosures (Verrecchia [1986], Kim and Verrecchia [1994], Lang and Lundholm [1996], Healy, Hutton, and Palepu [1999]) and the use of segment disclosures (Botosan and Harris [2000]). Conversely, firm disclosure can reduce analyst coverage by crowding out information production, as shown by Chang, Ljungqvist, and Tseng [2023], who find a reduction in coverage when firms join EDGAR. We contribute to this literature by showing that peer analysts’ information acquisition, implied by the disclosed site visits, can substitute for analysts’ own information production. Therefore, the disclosure of other analysts’ activities can help analysts decide which firms to cover.
More broadly, our paper contributes to the literature on public disclosures. First, our findings answer the call by Leuz and Wysocki [2016] for more research on the externalities of firms’ disclosures (Dye [1990], Admati and Pfleiderer [2000]). Public disclosures by peer firms are associated with a firm's cost of capital and liquidity (e.g., Shroff, Verdi, and Yost [2017], De George, Phan, and Stoumbos [2019]) and affect its voluntary disclosures (e.g., Baginski and Hinson [2016], Seo [2021], Breuer, Hombach, and Müller [2022]). Our paper shows that strategic interactions among analysts can contribute to the externalities of public disclosure. Likewise, in its focus on public disclosure of private meetings, our study also answers the call for more research to examine the consequences of requiring firms to disclose site visits (Lennox and Wu [2022]). Although our setting is China, the findings pertain to capital markets more broadly. As evidenced by anecdotes and surveys (Brown et al. [2015]), private meetings with firm managers are critical to analysts. This is confirmed by the empirical evidence of Choy and Hope [2024], who use New York City's daily taxi trip records to proxy for private meetings. Studies using the U.S. setting (e.g., Hong and Kacperczyk [2010], Crawford, Roulstone, and So [2012], Merkley, Michaely, and Pacelli [2017]) also suggest that competition among analysts affects their behavior. Meanwhile, there have been discussions about whether the U.S. Securities and Exchange Commission should increase the transparency of these meetings (e.g., Bengtzen 2017). Our results may contribute to this debate by demonstrating the consequences of mandated timely disclosure of private meetings.
Click to read more