Skip to main content

Seeing is believing: analysts’ corporate site visits

Abstract

This study examines the impact of corporate site visits on analysts’ forecast accuracy based on a sample of such visits to Chinese listed firms during 2009–2012. We find that analysts who conduct visits (“visiting analysts”) have a greater increase in forecast accuracy than other analysts. Consistent with the notion that site visits facilitate analysts’ information acquisition through observing firms’ operations, we find that the results are stronger for manufacturing firms, firms with more tangible assets, and firms with more concentrated business lines. Moreover, we find that the effect of a site visit is greater when the site visit is an analyst-only visit, when the current visit is preceded by fewer visits, and when visiting analysts are based far from the visited firms. Furthermore, we find that site visits partially mitigate nonlocal analysts’ information disadvantage. Collectively, these results indicate that site visits are an important information acquisition activity for analysts.

This is a preview of subscription content, access via your institution.

Notes

  1. According to the corporate site visit policies disclosed by listed firms, IR managers, including board secretaries and securities affairs representatives, are usually the liaisons for site visits. They are responsible for approving site visit applications, organizing field tours, and accompanying the visitors during the site visits. Our interviews suggest that the practice is consistent with these corporate policies.

  2. Cheng et al. (2015) hand-collect the information about firm executives’ participation in site visits from the detailed records of 4425 site visits from the SZSE website in 2013. Please note that firms started to provide the detailed minutes of site visits only in 2013, including whom the visitors met with during the visits. This information is not available during our sample period.

  3. Prior studies (e.g., Bae et al. 2008) provide evidence consistent with the argument that analysts are more likely to have access to top executives of the firms located in the same area.

  4. Please see Jacob et al. (1999), Mikhail et al. (1997), Malloy (2005) and Cohen et al. (2010) for examples.

  5. Prior research also examines other types of selective access events, such as investors’ private meetings with firm executives (Solomon and Soltes 2015).

  6. We would like to point out that China has adopted the U.S. version of Regulation FD by mandating that, if an issuer discloses material nonpublic information to certain persons, it must make public disclosure of that information. According to the Article 41 of the CSRC’s Regulation FD, which took effect on Jan. 31, 2007, “A listed company shall, hold conference calls, analysts’ meetings, road shows, accepting investors’ field investigation, etc., to communicate with the institutions and individuals about the business operations, financial status and other events, but it shall not provide any inside information.”

  7. We follow a strict interview protocol, asking the same set of open-ended questions in the same order across all interviews.

  8. We focus on forecast accuracy because it is the most frequently studied performance metric of analysts in the accounting literature. The information obtained from site visits likely affects forecast accuracy more than other performance metrics, such as recommendation profitability. Analysts also have incentives to improve their forecast accuracy. In China, analysts’ compensation is largely determined by their annual rankings, as determined by some media such as New Fortune and Today Investment. While New Fortune’s star analyst ranking is largely based on the votes of institutional investors, Today Investment’s ranking is more objective, and one of the awards is explicitly designated for the analysts with the most accurate earnings forecasts in every industry.

  9. As discussed in Cheng et al. (2015), a firm’s top executives participate in only 15.2 % of the site visits.

  10. Analysts likely have other non-official means to obtain information on their peers’ site visits. First, occasionally there are voluntary disclosures on some firms’ websites about visits. But it is unclear how timely or comprehensive such disclosures are. Second, based on the authors’ conversations with the analysts who have conducted site visits, some analysts may obtain such information through their networks (e.g., their friends in other brokerages covering the same industry) or through their peers’ research reports issued after site visits.

  11. We identify sell-side brokers in the following way. First, we assign one unique broker ID to each broker, even when they take different formats in different firms’ site visit records (e.g., “CITIC Securities,” “CITIC Securities Company,” and “CITIC Securities Co. Ltd.,”) or when they change names over the sample period. Second, we exclude the buy-side analysts’ visits based on a manual check of the brokers’ websites. This process leads to a total of 167 unique brokers, 114 of which are Chinese brokers and 53 of which are foreign brokers. Of the 114 Chinese brokers, 102 brokers’ forecasts are covered in the CSMAR database.

  12. We match the brokers’ names in the analyst forecast database with those in the site visit database. Because one broker usually has only one analyst covering a specific firm, we use “broker” and “analyst” interchangeably when discussing forecasts.

  13. We also require that, for each site visit event, nonvisiting analysts do not conduct any other site visit to the same firm during the period beginning from 6 months before to the end of the first month afterward. This constraint is imposed to ensure a clean sample of nonvisiting analysts as the benchmark group for such a visit.

  14. In China, forecast errors, when scaled by stock prices, are usually very small. This is due to the very high PE ratio in Chinese stock markets. Our statistics are comparable to those reported in other studies of Chinese financial analysts, such as those of Gu et al. (2013). In an untabulated additional analysis, we use the relative forecast accuracy score developed by Hong and Kubik (2003), and our inferences remain the same.

  15. The inferences remain the same when we exclude these analysts from the analyses. Separately, some analysts issue a post-visit forecast but not a pre-visit forecast. For these cases, we assume that the pre-visit forecast to be the same as the mean value of all other analysts’ pre-visit forecasts. The inferences remain the same if we exclude these analysts from the analyses.

  16. The objective of H3 is to test whether the usefulness of site visits varies with the characteristics of site visits or visiting analyst groups, which are not relevant for nonvisiting analysts. When we include the main effect of these variables in an untabulated analysis, the inferences remain the same. Also while we include the interaction terms separately in the regression, the results are quantitatively similar when we include all interaction terms in the same regression model (untabulated).

  17. Our interviews suggest that the site visits that occur in the month after quarterly earnings announcements or the initial announcements of mergers and acquisitions are likely to be initiated by the firms, rather than by the analysts. Our conclusions still hold after excluding these visits.

  18. To the extent that selective disclosure is more likely to occur during the visits of the analysts who have favorable opinions of the firm, the robust results after excluding the site visits conducted by the analysts who have issued strong-buy recommendations recently for the firm or by those with investment banking relationships with the firm, as reported in Column (2) of Table 7, also suggest that selective disclosure is not driving our results.

  19. Due to the data requirement for calculating Visit_freq (i.e., information related to site visit frequencies in the past 6 months), the sample of earnings forecasts includes those with forecasting dates of July 2009 onward.

  20. Based on a sample of 250 randomly selected analyst reports that were issued by the visiting analysts in the month after their site visits, we find that 186 of these reports prominently use the term “site visit” in the report titles, in various forms such as “site visit briefing,” “site visit report,” or “site visit bulletin.” For the remaining reports, eight reports mention “site visit” as one of the information sources in the textual body of the reports. In total, 77.6 % (= (186 + 8)/250) of the randomly selected analyst reports explicitly disclose analysts’ recent site visits.

  21. Following Bae et al. (2008), we require that at least one local analyst and one nonlocal analyst follow the sample firm-years. This requirement helps alleviate the concern that local and nonlocal analysts choose to follow firms with different fundamentals. In addition, we exclude the stale earnings forecasts, i.e., those issued more than 300 days before earnings announcements. The final sample for this test consists of 17,714 earnings forecasts from July 2009 to 2012.

  22. We acknowledge that some of the factors affect both the costs and benefits. For example, while the cost of conducting site visits increases with geographical distance, the benefit also increases with it, as analysts generally know less about distant firms and thus the information obtained from visits is more important. As such, the results reflect the net effect. Also, this is by no means a comprehensive list of proxies for the costs and benefits of conducting site visits.

  23. Companies occasionally include telephone interviews, webinars, email exchanges, nondeal road shows, investor conferences, industry summits, industry forums, annual broker conferences, and one-on-one meetings with managers in this section. We include only the site visits (held at company headquarters or subsidiaries) in the current study.

References

  • Abarbanell, J. S., Lanen, W. N., & Verrecchia, R. E. (1995). Analysts’ forecasts as proxies for investor beliefs in empirical research. Journal of Accounting and Economics, 20(1), 31–60.

    Article  Google Scholar 

  • Asquith, P., Mikhail, M., & Au, A. (2005). Information content of equity analyst reports. Journal of Financial Economics, 75(2), 245–282.

    Article  Google Scholar 

  • Bae, K. H., Stulz, R. M., & Tan, H. (2008). Do local analysts know more? A cross-country study of the performance of local analysts and foreign analysts. Journal of Financial Economics, 88, 581–606.

    Article  Google Scholar 

  • Bradshaw, M. (2011). Analysts’ forecasts: What do we know after decades of work? Unpublished working paper. Boston College.

  • Brown, L. D., Call, A. C., Clement, M. B., & Sharp, N. Y. (2015). Inside the “Black Box” of sell-side financial analysts. Journal of Accounting Research, 53(1), 1–47.

    Article  Google Scholar 

  • Bushee, B. J., Jung, M. J., & Miller, G. S. (2013). Do investors benefit from selective access to management? Unpublished working paper. University of Pennsylvania, New York University, and University of Michigan.

  • Call, A. C., Chen, S., & Tong, Y. H. (2013). Are analysts’ cash flow forecasts naïve extensions of their own earnings forecasts? Contemporary Accounting Research, 30(2), 438–465.

    Article  Google Scholar 

  • Chen, X., Cheng, Q., & Lo, K. (2010). On the relationship between analyst reports and corporate disclosures: Exploring the roles of information discovery and interpretation. Journal of Accounting and Economics, 49(3), 206–226.

    Article  Google Scholar 

  • Cheng, Q., Du, F., Wang, X., & Wang, Y. (2015). Are investors’ corporate site visits informative? Working paper. Singapore Management University, The University of Hong Kong, and Central University of Finance and Economics.

  • Clement, M. B. (1999). Analyst forecast accuracy: Do ability, resources, and portfolio complexity matter? Journal of Accounting and Economics, 27(3), 285–303.

    Article  Google Scholar 

  • Clement, M. B., & Tse, S. V. (2003). Do investors respond to analysts’ forecast revisions as if forecast accuracy is all that matters? The Accounting Review, 78(1), 227–249.

    Article  Google Scholar 

  • Cohen, L., Frazzini, A., & Malloy, C. (2010). Sell-side school ties. Journal of Finance, 65(4), 1409–1437.

    Article  Google Scholar 

  • Coval, J. D., & Moskowitz, T. J. (2001). The geography of investment: Informed trading and asset prices. Journal of Political Economy, 109(4), 811–841.

    Article  Google Scholar 

  • Green, T. C., Jame, R., Markov, S., & Subasi, M. (2014a). Broker-hosted investor conferences. Journal of Accounting and Economics, 58(1), 142–166.

    Article  Google Scholar 

  • Green, T. C., Jame, R., Markov, S., & Subasi, M. (2014b). Access to management and informativeness of analyst research. Journal of Financial Economics, 114(2), 239–255.

    Article  Google Scholar 

  • Gu, Z., Li, Z., & Yang, G. Y. (2013). Monitors or predators: The influence of institutional investors on sell-side analysts. The Accounting Review, 88(1), 137–169.

    Article  Google Scholar 

  • Healy, P. M., & Palepu, K. G. (2001). Information asymmetry, corporate disclosure, and the capital markets: A review of empirical disclosure literature. Journal of Accounting Economics, 31(1–3), 405–440.

    Article  Google Scholar 

  • Hong, H., & Kubik, J. D. (2003). Analyzing the analysts: Career concerns and biased earnings forecasts. Journal of Finance, 58, 313–352.

    Article  Google Scholar 

  • Ivkovic, Z., & Jegadeesh, N. (2004). The timing and value of forecast and recommendation revisions. Journal of Financial Economics, 73(3), 433–463.

    Article  Google Scholar 

  • 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.

    Article  Google Scholar 

  • Keung, E. C. (2010). Do supplementary sales forecasts increase the credibility of financial analysts’ earnings forecasts? The Accounting Review, 85(6), 2047–2074.

    Article  Google Scholar 

  • Malloy, C. J. (2005). The geography of equity analysis. Journal of Finance, 60(2), 719–755.

    Article  Google Scholar 

  • Mayew, W., Sharp, N., & Venkatachalam, M. (2013). Using earnings conference calls to identify analysts with superior private information. Review of Accounting Studies, 18(2), 386–413.

    Article  Google Scholar 

  • Mikhail, M. B., Walther, B. R., & Willis, R. H. (1997). Do security analysts improve their performance with experience? Journal of Accounting Research, 35(3), 131–157.

    Article  Google Scholar 

  • Morck, R., Yeung, B., & Yu, W. (2000). The information content of stock markets: Why do emerging markets have synchronous stock price movements? Journal of Financial Economics, 58(1–2), 215–260.

    Article  Google Scholar 

  • O’Brien, P., & Bhushan, R. (1990). Analyst following and institutional ownership. Journal of Accounting Research, 28(3), 55–82.

    Article  Google Scholar 

  • Solomon, D., & Soltes, E. (2015). What are we meeting for? The consequences of private meetings with investors. Journal of Law and Economics, 58(2), 325–355.

    Article  Google Scholar 

  • Soltes, E. (2014). Private interaction between firm management and sell-side analysts. Journal of Accounting Research, 52(1), 245–272.

    Article  Google Scholar 

Download references

Acknowledgments

We greatly appreciate the helpful comments and suggestions from Sarah Bonner, Brian Bushee, Andrew Call, Kevin Chen, Shuping Chen, Xia Chen, Patricia Dechow (Editor), Mark DeFond, Mei Feng, Ole-Kristian Hope, Marcus Kirk, Jing Liu, Stan Markov, Steve Matsunaga, Chul Park, Joseph Piotroski, Nathan Sharp, Doug Skinner, T.J. Wong, Guochang Zhang, two anonymous reviewers, and conference and workshop participants at the 2013 Alumni Symposium of USC Leventhal School of Accounting, 2014 CKGSB Colloquium, 2014 Chinese University of Hong Kong CiG conference, 2015 EAA Annual Congress, 2015 AAA annual conference, 2015 University of Wisconsin-Madison Alumni Research Conference, the University of Hong Kong, and Hong Kong University of Science and Technology. We gratefully acknowledge the financial support provided by the General Research Fund of Hong Kong Research Grants Council (Project No. 790613) and the National Natural Science Foundation of China (Project No. 71102124). Cheng gratefully acknowledges funding from the Lee Kong Chian Fellowship.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Qiang Cheng.

Appendices

Appendix 1

A site visit example: extract of the 2011 annual report of Shenzhen Airport Co. Ltd

During the reporting period, the company follows the information disclosure guidelines and investor relationship management bylaws issued by the SZSE. The company communicates with investors by hosting site visits and holding one-on-one meetings with institutional investors and by taking phone calls from individual investors.Footnote 23 During the reporting period, the company meets with 54 individuals from various institutions. During these visits, the company discusses its general operations and future strategy with investors based on public information. The company does not selectively disclose information to investors. The site visits are detailed as follows.

Time Visitor Topics of discussion and materials provided
Jan. 5, 2011 Everbright Securities Recent company updates
Jan. 11, 2011 China International Capital Co. Ltd. Recent company updates
Jan. 21, 2011 Changjiang Securities, China Investment Securities Progress of construction and recent operations
Mar. 23, 2011 CITIC Securities Recent company updates
Mar. 24, 2011 Changjiang Securities, Baoying Fund Management, Huatai-PineBridge Investments Company operation and construction expansion
Apr. 12, 2011 GF Securities, China Merchants Fund Construction of T3, and the business circumstances of main operations and nonflight-related operations
Apr. 20, 2011 Guosen Securities, Harvest Fund, Guotai AMC, Sino Life Insurance, Dacheng Fund Company fundamentals
May 12, 2011 Taikang AMC Company fundamentals
May 19, 2011 Changjiang Securities, Chengrui Investment Construction of T3 and recent company updates
June 10, 2011 Ping An Securities Recent company updates
June 13, 2011 Hongyuan Securities Convertible bond and business circumstances
June 23, 2011 Investor Conference hosted by Changjiang Securities Introduction of current business picture and topical issues
Aug. 16, 2011 Bosera Securities, Dacheng Securities Fundamentals and convertible bond
Aug. 19, 2011 Ping An Annuity Insurance Company fundamentals
Sep. 29, 2011 JS Cresvale Securities Company fundamentals
Nov. 7, 2011 UBS Company fundamentals
Nov. 17, 2011 China Merchants Fund Company fundamentals
Dec. 8, 2011 Upstone Capital, Kangqiao Asset, Houde Investment Company fundamentals
Dec. 14, 2011 Guotai Junan Company fundamentals
Dec. 15, 2011 GF Securities Company fundamentals
Dec. 19, 2011 Everbright Securities Company fundamentals

Appendix 2

Variable definitions

Dependent variable (site visit event-analyst group level variables)
\(\Delta Accuracy_{k,j,t} = - \left( {Post\_Visit\_AFE_{k,j,t} - \Pr e\_Visit\_AFE_{k,j,t} } \right)\) −1 times the change in the absolute forecast error of analyst group k (visiting or nonvisiting analysts) for firm j from the 6 months before the site visit conducted on day t to 1 month after. A positive value implies an improvement in forecast accuracy from the pre- to post-visit periods. For each analyst group, we identify the most recent annual EPS forecast issued by each analyst within the group in the 6 months before the site visit, calculate the group mean as the group consensus forecast, and calculate Pre_Visit_AFE as the absolute difference between the group consensus EPS forecast and actual EPS, scaled by the stock price at the beginning of the year, expressed in percentage. To calculate Post_Visit_AFE, we identify the first forecast made by the analysts in the same group in the month after the site visit and calculate their forecast consensus and the absolute forecast error for the same group in the period after the site visit. For the analysts who do not update their forecasts in the post-visit period, we assume their post-visit forecasts to be the same as their pre-visit forecasts. If an analyst does not issue a pre-visit earnings forecast (but does issue a post-visit earnings forecast), then we use the mean forecast of all of the other analysts’ forecasts in the pre-visit period as the pre-visit forecast for such an analyst when calculating the forecast accuracy change
Key independent variable (site visit event-analyst group level variables)
Visit k,j,t An indicator variable that equals 1 for the visiting analyst group and 0 for the nonvisiting analyst group. An analyst is a visiting analyst if he/she visits firm j on a site visit event day t. The analysts who follow the same firm but do not visit it in the 6 months before or 1 month after the site visit of interest are referred to as nonvisiting analysts
Variables for cross-sectional analyses (firm-year level and visiting group level variables)
Manufacture j,t An indicator variable that equals 1 when the firm is a manufacturing firm and 0 otherwise
Tangibility j,t An indicator variable that equals 1 when the ratio of PP&E over total assets is greater than or equal to the sample median and 0 otherwise
Concentration j,t An indicator variable that equals 1 when the firm’s Herfindahl–Hirschman index (HHI) based on segment revenue is greater than or equal to the sample median and 0 otherwise. Segment revenue HHI equals the sum of squares of the ratio of segment revenue to the total revenue for firm j in the current year
AnalystOnly k,j,t An indicator variable for analyst-only visits. It equals 1 if all of the visitors are sell-side analysts and 0 otherwise
Remote k,j,t An indicator variable for nonlocal analysts. It equals 1 if nonlocal analysts outnumber the local analysts in the visiting groups and 0 otherwise. Nonlocal analysts are those whose brokerages are located more than 400 km (250 miles) from the visited firm’s headquarters
Unpreceded k,j,t An indicator variable for firms with fewer preceding visits within the month before the site visit of interest. It equals 1 if the number of preceding site visits within the 1-month window before the current site visit t is below the sample median for the visiting group and 0 otherwise. This variable is coded as 0 for nonvisiting groups
Control variables
ΔHorizon k,j,t Change in forecast horizon, calculated as the log transformation of the decrease in the average forecast horizon of analyst group k (visiting or nonvisiting group) from the pre- to post-visit periods. The forecast horizon is defined as the number of days between the forecast issue date and corresponding earnings announcement date
Firmexp k,j,t Analyst-firm-specific experience, calculated as the log transformation of the average firm-specific experience of all of the analysts in analyst group k for firm j. Firm-specific experience is calculated as the number of years between an analyst’s first forecast for firm j and his/her current forecast for firm j
Brokersize k,j,t Brokerage size, defined as the average number of analysts working for the brokerages in group k
ANA_group k,j,t Group size, calculated as the log transformation of the number of analysts in group k
MV j,t Firm size, calculated as the log transformation of the market value of equity of firm j at the end of the last fiscal year
Inst_holding j,t Institutional ownership, calculated as the ownership percentage of institutional investors
Indep j,t Board independence, calculated as the ratio of the number of independent directors to the total number of directors for firm j in the current year
BM j,t Book-to-market ratio, calculated as the book value of equity divided by the market value of equity
Growth j,t Revenue growth, defined as the revenue in year t divided by the revenue in year t − 1
Loss j,t Loss indicator that equals 1 if the net income is negative in year t and 0 otherwise
BHAR j,t The buy-and-hold market-adjusted returns in year t

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Cheng, Q., Du, F., Wang, X. et al. Seeing is believing: analysts’ corporate site visits. Rev Account Stud 21, 1245–1286 (2016). https://doi.org/10.1007/s11142-016-9368-9

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11142-016-9368-9

Keywords

  • Site visits
  • Analyst forecasts
  • Information acquisition activities
  • Local advantage

JEL Classifications

  • G10
  • G24
  • M41