Sell-side analysts' stock recommendations cluster around the consensus, and the clustering intensifies when aggregate uncertainty about future market returns is high (Lin, 2018). What the documenting study could not show is why: the amplification is consistent both with information-based clustering (analysts processing correlated signals or rationally inferring from peers under ambiguity) and with reputation-based herding (career, blame-sharing, and relative-performance incentives to stay near the consensus), which are observationally equivalent in the market-reaction data used. The open problem is to separately identify these channels in recommendation behavior and to establish which one drives the uncertainty-amplified component, and then to quantify whether that component impairs price efficiency — the magnitude of mispricing and the speed of price discovery around recommendation revisions. Later literature has extended the phenomenon to other markets and uncertainty proxies and has examined adjacent price consequences, but no study credibly disentangles the mechanisms for stock recommendations.

Background

Sell-side analysts are brokerage employees who publish research on publicly traded firms; one of their main outputs is the stock recommendation, an ordinal rating from strong sell to strong buy (on the Institutional Brokers' Estimate System (I/B/E/S) database, the standard commercial archive of analyst recommendations and forecasts, coded 1 to 5). Herding means that analysts systematically issue recommendations closer to the prevailing consensus of their peers than their own information would justify. Because an analyst's private information is unobservable, the empirical literature infers herding from market prices: in the test of Jegadeesh and Kim (2010), the one-day abnormal return around a recommendation revision (the revision-day stock return minus the return of a portfolio matched on industry, size, and market-to-book ratio) is regressed on how far the revised recommendation deviates from the consensus of all other outstanding recommendations. If the market reacts more strongly to revisions that break from the herd, the market is inferred to price-protect against herding, and analysts are inferred to herd. On US data this test shows that analysts herd, more strongly in downgrades than upgrades, and more strongly from analysts at large brokerages.

Several distinct motives generate observationally similar clustering. In information-based herding, analysts condition on the same public signals or rationally extract information from each other's actions; Welch (2000) documents such cascade effects (later analysts ignoring their own signals in favor of the observed history of peer revisions) in analyst forecast revisions. In reputation-based herding, career and compensation concerns reward conformity: Trueman (1994) models analysts whose labor-market reputation improves when they are wrong alongside the crowd; Graham (1999) documents herding among investment newsletters whose payoffs depend on relative reputation; Hong, Kubik and Solomon (2000) show that career concerns make inexperienced analysts' earnings forecasts herd more; and Clement and Tse (2005) relate analyst characteristics to forecast herding. A separate literature establishes that firm-level information uncertainty magnifies herding (Zhang, 2006).

Lin (2018) added an aggregate, time-varying condition to this setting: herding in US recommendations increases with aggregate uncertainty, defined as Knightian uncertainty (unknown probability distributions) about aggregate equity returns, measured as the cross-sectional dispersion of professional forecasters' implied market-return forecasts in the Survey of Professional Forecasters (SPF; the construction of Anderson, Ghysels and Juergens, 2009). This aggregate-uncertainty measure is deliberately distinct from quantifiable market risk (the conditional variance of market returns) and from firm-level information uncertainty, and the paper controls for both; the uncertainty amplification is stronger for small stocks and, weakly, for inexperienced analysts. The design establishes that herding co-moves with aggregate uncertainty, but not why: ambiguity-impaired signal processing (an informational story) and reputational self-protection under ambiguity (an incentives story) imply the same regression signs. It also leaves open what the amplified herding does to prices — Jegadeesh and Kim (2010) argue the market anticipates herding and neutralizes it, while later work on recommendation herding documents overreaction and drift — so whether the uncertainty-amplified component of herding slows price discovery (the speed at which stock prices incorporate the information analysts release) or widens mispricing remains an open empirical question.

Problem Statement

Determine the economic mechanism behind the aggregate-uncertainty amplification of herding in analysts' stock recommendations documented by Lin (2018), and its consequences for price efficiency. Concretely: (1) construct a test that separately identifies, in recommendation data, information-based clustering (revisions tracking peers because of correlated signals or rational inference from peers' actions) and reputational herding (revisions tracking the consensus because of career, blame-sharing, or relative-performance incentives), and establish which channel drives the component of herding that rises with measured aggregate uncertainty (such as SPF-forecast-dispersion or policy-uncertainty measures), keeping that component distinct from market risk and firm-level information uncertainty; and (2) quantify whether this uncertainty-amplified herding component impairs price efficiency, specifically its effect on mispricing magnitude and on the speed of price discovery around recommendation revisions. The two deliverables share one overarching objective: a causal account of why aggregate uncertainty intensifies analyst herding in recommendations and of what that intensification costs market efficiency. (Generalization to other markets and alternative uncertainty proxies is excluded from this statement because subsequent literature has addressed it.)

References

  1. Mei-Chen Lin (2018), "The impact of aggregate uncertainty on herding in analysts' stock recommendations", International Review of Financial Analysis 57:90-105. DOI: 10.1016/j.irfa.2018.02.006, https://doi.org/10.1016/j.irfa.2018.02.006
  2. Narasimhan Jegadeesh and Woojin Kim (2010), "Do Analysts Herd? An Analysis of Recommendations and Market Reactions", Review of Financial Studies 23(2):901-937. DOI: 10.1093/rfs/hhp093, https://doi.org/10.1093/rfs/hhp093
  3. Evan W. Anderson, Eric Ghysels and Jennifer L. Juergens (2009), "The impact of risk and uncertainty on expected returns", Journal of Financial Economics 94(2):233-263. DOI: 10.1016/j.jfineco.2008.11.001, https://doi.org/10.1016/j.jfineco.2008.11.001
  4. Ivo Welch (2000), "Herding among security analysts", Journal of Financial Economics 58(3):369-396. DOI: 10.1016/S0304-405X(00)00076-3, https://doi.org/10.1016/S0304-405X(00)00076-3
  5. Brett Trueman (1994), "Analyst Forecasts and Herding Behavior", Review of Financial Studies 7(1):97-124. DOI: 10.1093/rfs/7.1.97, https://doi.org/10.1093/rfs/7.1.97
  6. John R. Graham (1999), "Herding among Investment Newsletters: Theory and Evidence", Journal of Finance 54(1):237-268. DOI: 10.1111/0022-1082.00103, https://doi.org/10.1111/0022-1082.00103
  7. Harrison Hong, Jeffrey Kubik and Amit Solomon (2000), "Security Analysts' Career Concerns and Herding of Earnings Forecasts", The RAND Journal of Economics 31(1):121-144. DOI: 10.2307/2601032, https://doi.org/10.2307/2601032
  8. Michael B. Clement and Senyo Y. Tse (2005), "Financial Analyst Characteristics and Herding Behavior in Forecasting", Journal of Finance 60(1):307-341. DOI: 10.1111/j.1540-6261.2005.00731.x, https://doi.org/10.1111/j.1540-6261.2005.00731.x
  9. X. Frank Zhang (2006), "Information Uncertainty and Stock Returns", Journal of Finance 61(1):105-137. DOI: 10.1111/j.1540-6261.2006.00831.x, https://doi.org/10.1111/j.1540-6261.2006.00831.x
  10. Min Kim and Fernando Zapatero (2023), "Uncertainty and Dispersion of Opinions", SSRN working paper. DOI: 10.2139/ssrn.4623332, https://doi.org/10.2139/ssrn.4623332
  11. Ming-Ti Chiang and Mei-Chen Lin (2019), "Market sentiment and herding in analysts' stock recommendations", The North American Journal of Economics and Finance 48:48-64. DOI: 10.1016/j.najef.2019.01.007, https://doi.org/10.1016/j.najef.2019.01.007
  12. Min Chen, Zhaobo Zhu, Peiwen Han, Bo Chen and Jia Liu (2022), "Economic policy uncertainty and analyst behaviours: Evidence from the United Kingdom", International Review of Financial Analysis 79:101906. DOI: 10.1016/j.irfa.2021.101906, https://doi.org/10.1016/j.irfa.2021.101906
  13. Zhaobo Zhu, Hang Lin, Min Chen and Peiwen Han (2023), "The spillover effect of economic policy uncertainty: Evidence from analyst behaviors in Hong Kong", Finance Research Letters 52:103570. DOI: 10.1016/j.frl.2022.103570, https://doi.org/10.1016/j.frl.2022.103570
  14. Cao Hoang Anh Le, Yaowen Shan and Stephen Taylor (2024), "International economic policy uncertainty and analysts' earnings forecasts", Pacific-Basin Finance Journal 85:102403. DOI: 10.1016/j.pacfin.2024.102403, https://doi.org/10.1016/j.pacfin.2024.102403
  15. Amanjot Singh, Harminder Singh and Venura Welagedara (2024), "Aggregate uncertainty, information acquisition, and analyst stock recommendations", International Review of Finance 24(4):604-640. DOI: 10.1111/irfi.12455, https://doi.org/10.1111/irfi.12455
  16. Wen Chen, Shen Zhao and Guofu Zhou (2023), "Hide in the Herd: Uncertainty and Informational Inefficiency", SSRN working paper. DOI: 10.2139/ssrn.4342477, https://doi.org/10.2139/ssrn.4342477
  17. Nerissa C. Brown, Kelsey D. Wei and Russ Wermers (2014), "Analyst Recommendations, Mutual Fund Herding, and Overreaction in Stock Prices", Management Science 60(1):1-20. DOI: 10.1287/mnsc.2013.1751, https://doi.org/10.1287/mnsc.2013.1751
  18. Nianhang Xu, Xuanyu Jiang, Kam C. Chan and Shinong Wu (2017), "Analyst Herding and Stock Price Crash Risk: Evidence from China", Journal of International Financial Management & Accounting 28(3):308-348. DOI: 10.1111/jifm.12062, https://doi.org/10.1111/jifm.12062
  19. Binam Ghimire, Callum Reveley, Savva Shanaev and Humnath Panta (2023), "Analyst herding – whether, why, and when? Two new tests for herding detection in target forecast prices", Economics and Business Review 9(4). DOI: 10.18559/ebr.2023.4.892, https://doi.org/10.18559/ebr.2023.4.892
  20. Lan Thi Mai Nguyen, Chee Seng Cheong and Ralf Zurbruegg (2026), "Changes to the Work Environment and Analyst Herding Behavior", Journal of Accounting, Auditing & Finance, online first March 2026. DOI: 10.1177/0148558X261434637, https://doi.org/10.1177/0148558X261434637
  21. Jin Yeub Kim, Yongjun Kim and Myungkyu Shim (2023), "Do Financial Analysts Herd?", Global Economic Review 52(3):202-219. DOI: 10.1080/1226508X.2023.2230984, https://doi.org/10.1080/1226508X.2023.2230984