Abstract
We study whether changes in analyst recommendation ratings systems encouraged by the implementation of NASD 2711 in 2002 are associated with improved objectivity and independence in analyst recommendations. Using recommendations issued during windows surrounding major investment banking events (M&A transactions, IPOs and SEOs), we show that reductions in analyst optimism following the reforms concentrate in the recommendations of analysts whose employer adopted a three-tier rating system at the time of the reforms, and that this effect is generally stronger for analysts whom the underlying incentives to engage in unethical behaviour is greatest. We also find evidence that adoption of the three-tier system improved the market’s perception of the objectivity of analyst recommendations issued after SEOs, and that for hold and sell-type recommendations this effect was stronger for analysts subject to the greatest potential ethical conflicts. While we find some evidence of a general post-reform increase in the profitability of recommendations issued following equity transactions, this improvement was only conditioned by changes to the rating system in our IPO sample.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
All but one of the twelve GS signatory firms changed to a three-tier system immediately following the reforms, and it is thus very difficult to separate the impact of the change to rating system from the other provisions of the GS and SRO Rules for these firms.
Analyzing the Analysts: Hearings Before the House Committee on Capital Markets, Insurance, and Government Sponsored Enterprises of the Committee on Financial Services, prepared testimony of Ronald Glantz, 107th Congress, p. 243, June and July 2001.
We test recommendation optimism (ROPTIMISM) using ordered logistic regression and use OLS for all other tests. Standard errors are adjusted for clustering by covered firm in all regressions.
Further details on the measurement of analyst affiliation and optimism are provided in Appendix 2.
Five brokers switched rating systems during 2004. None of our results are substantively affected if we exclude all observations pertaining to these brokers from our sample, nor if we include observations pertaining to brokers who changed ratings systems after 2004.
The average pseudo R2 of the 23 year-specific logit regressions is approximately 16% and a least 2 covariates are significant in each regression. We use an annual caliper equivalent to an absolute p-score difference of 1.6%, which generates a sample that is balanced across all covariates in every year.
The tenor of our results is largely unaffected if we relax the parallel regressions assumption underpinning ordered logistic regression and instead estimate generalised ordered logit models with partial proportional odds. All previously tabulated significant coefficients for ChgtoTier3 × POSTREG × AFF remain significantly associated in the predicted direction with all outcomes in the ordered logit model. Similar results are obtained regarding the coefficient for the two-way interaction (ChgtoTier3 × POSTREG) in the IPO and SEO samples. In the M&A sample only, this test variable only has significant explanatory power regarding the probability that ROPTIMISMMA equals + 1.
References
Aristotle (1966). Nichomachean Ethics (W. D. Ross, Trans. first publ. 1925; reprinted). Oxford: Oxford University Press [in World Classics Series].
Barber, B. M., Lehavy, R., McNichols, M., & Trueman, B. (2006). Buys, holds, and sells: The distribution of investment banks’ stock ratings and the implications for the profitability of analysts’ recommendations. Journal of Accounting and Economics, 41(1), 87–117.
Barniv, R., Hope, O.-K., Myring, M. J., & Thomas, W. B. (2009). Do analysts practice what they preach and should investors listen? Effects of recent regulations. The Accounting Review, 84(4), 1015–1039.
Becher, D. A., Cohn, J. B., & Juergens, J. L. (2015). Do stock analysts influence merger completion? An examination of postmerger announcement recommendations. Management Science, 61(10), 2430–2448.
Billings, B. K., Buslepp, W. L., & Huston, G. R. (2014). Worth the hype? The relevance of paid-for analyst research for the buy-and-hold investor. The Accounting Review, 89(3), 903–931.
Chen, C. Y., & Chen, P. F. (2009). NASD rule 2711 and changes in analysts’ independence in making stock recommendations. The Accounting Review, 84(4), 1041–1071.
Clarke, J., Khorana, A., Patel, A., & Rau, P. R. (2011). Independents’ day? Analyst behaviour surrounding the global settlement. Annals of Finance, 7(4), 529–547.
Cliff, M. T. (2007). Do affiliated analysts mean what they say? Financial Management, 36(4), 5–29.
Corwin, S. A., Larocque, S. A., & Stegemoller, M. A. (2017). Investment banking relationships and analyst affiliation bias: The impact of the global settlement on sanctioned and non-sanctioned banks. Journal of Financial Economics, 124(3), 614–631.
Cote, J., & Goodstein, J. (1999). A breed apart? Security analysts and herding behavior. Journal of Business Ethics, 18(3), 305–314.
Dambra, M., Field, L. C., Gustafson, M. T., & Pisciotta, K. (2018). The consequences to analyst involvement in the IPO process: Evidence surrounding the JOBS Act. Journal of Accounting and Economics, 65(2–3), 302–330.
Davis, A. (2004). Increasingly, stock research serves the pros, not ‘little guy’. Retrieved 1 September 2017 from http://www.wsj.com/articles/SB107843948785646943.
Dechow, P. M., Hutton, A. P., & Sloan, R. G. (2000). The relation between analysts’ forecasts of long-term earnings growth and stock price performance following equity offerings. Contemporary Accounting Research, 17(1), 1–32.
Dunaief, D. (2003). Analysts abandon Wall Street. Retrieved September 1, 2017 from http://www.nydailynews.com/archives/money/analysts-abandon-wall-st-article-1.660050.
Eames, M., Glover, S. M., & Kennedy, J. (2002). The association between trading recommendations and broker-analysts’ earnings forecasts. Journal of Accounting Research, 40(1), 85–104.
Forbes, W. (2013). No conflict, no interest: On the economics of conflicts of interest faced by analysts. European Journal of Economics, 35, 327–348.
Givoly, D., & Lakonishok, J. (1979). The information content of financial analysts’ forecasts of earnings: Some evidence on semi-strong inefficiency. Journal of Accounting and Economics, 1(3), 165–185.
Guan, Y., Lu, H., & Wong, M. F. (2012). Conflict-of-interest reforms and investment bank analysts’ research biases. Journal of Accounting, Auditing & Finance, 27(4), 443–470.
Hilary, G., & Hsu, C. (2013). Analyst forecast consistency. The Journal of Finance, 68(1), 271–297.
Houston, J., James, C., & Karceski, J. (2006). What a difference a month makes: Stock analyst valuations following initial public offerings. Journal of Financial and Quantitative Analysis, 41(1), 111–138.
Huang, H., Li, M., & Shi, J. (2018). Should gatekeepers be publicly traded? Journal of Business Ethics, Online First, 31, 1–26.
Kadan, O., Madureira, L., Wang, R., & Zach, T. (2009). Conflicts of interest and stock recommendations: The effects of the Global Settlement and related regulations. Review of Financial Studies, 22(10), 4189–4217.
Keating, T. J. (2013). Analyzing the analysts: A Survey of the state of wall street equity research 10 years after the Global Settlement. Oklahoma: Keating Investments.
Kolasinski, A. C., & Kothari, S. (2008). Investment banking and analyst objectivity: Evidence from analysts affiliated with mergers and acquisitions advisors. Journal of Financial and Quantitative Analysis, 43(4), 817–842.
Lin, H., & McNichols, M. F. (1998). Underwriting relationships, analysts’ earnings forecasts and investment recommendations. Journal of Accounting and Economics, 25(1), 101–127.
Lu, R., Hou, W., Oppenheimer, H., & Zhang, T. (2016). The integrity of financial analysts: Evidence from asymmetric responses to earnings surprises. Journal of Business Ethics, 151(3), 761–783.
Lys, T., & Sohn, S. (1990). The association between revisions of financial analysts’ earnings forecasts and security-price changes. Journal of Accounting and Economics, 13(4), 341–363.
Malmendier, U., & Shanthikumar, D. (2007). Are small investors naive about incentives? Journal of Financial Economics, 85(2), 457–489.
Michaely, R., & Womack, K. L. (1999). Conflict of interest and the credibility of underwriter analyst recommendations. Review of Financial Studies, 12(4), 653–686.
Naffziger, F., & Fox, M. (2004). Securities analysts’ conflicts of interest: Ethical and legal issues and international regulatory reactions. Journal of Legal Studies in Business, 11, 1–40.
New York Attorney General. (2002). Affidavit in support of an application for an order pursuant to general business law Sect. 354. Retrieved from http://www.ag.ny.gov/sites/default/files/press-releases/archived/MerrillL.pdf.
O’Brien, P. C., McNichols, M. F., & Lin, H. (2005). Analyst impartiality and investment banking relationships. Journal of Accounting Research, 43(4), 623–650.
Palazzo, G., Krings, F., & Hoffrage, U. (2012). Ethical blindness. Journal of Business Ethics, 109(3), 323–338.
United States Government Accountability Office (GAO). (2012). Securities research. Additional actions could improve regulatory oversight of analyst conflicts of interest. Retrieved September 1, 2017 from http://www.gao.gov/products/GAO-12-209.
Veit, E. T., & Murphy, M. (1996). Ethics violations: A survey of investment analysts. Journal of Business Ethics, 15(12), 1287–1297.
Wu, M., Wilson, M., & Wu, Y. (2015). Was the Global Settlement effective in mitigating systematic bias in affiliated analyst recommendations? Journal of Business Ethics, 146(3), 1–19.
Acknowledgements
Our paper has benefited from feedback provided by Siva Nathan, and participants at the 2016 AFAANZ Annual Conference, Gold Coast. We gratefully acknowledge this feedback.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no conflict of interest.
Ethical Approval
This article does not contain any studies with human participants or animals performed by any of the authors.
Additional information
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendices
Appendix 1: Variable Definitions
Variables Used in Tests of Analyst Optimism (As Per Eqs. 1 and 2)
OPTIMISM | = | B0 + B1ChgtoTier3i + B2POSTREGi,j + B3ChgtoTier3i × POSTREGi,j + B4DEALVALUEj + B5FIRMCAPj + B6EXPi + B7FOLLOWj + B8DAYSi,j + B9BROKERSIZEi,j + B10LEVELi,j + B11OWNERSHIPj + ε | (1) |
OPTIMISM | = | B0 + B1ChgtoTier3i + B2POSTREGi,j + B3AFFi,j + B4 ChgtoTier3i × POSTREGi,j + B5ChgtoTier3i × AFFi,j + B6POSTREGi,j × AFFi,j + B7ChgtoTier3i × POSTREGi,j × AFFi,j + B8DEALVALUEj + B9FIRMCAPj + B10EXPi + B11FOLLOWj + B12DAYSi,j + B13BROKERSIZEi,j + B14LEVELi,j + B15OWNERSHIPj + ε | (2) |
(1) Proxies for analyst optimism in the M&A context | |||
ROPTIMISMMA,i,j | = | An ordinal variable that is equal to 1 if analyst i upgrades their recommendation for acquirer j within ± 90 days of the M&A announcement, 0 if the recommendation is unchanged and − 1 if there is a downgrade. Further details of the measurement of this variable are provided in Appendix 2 | |
FOPTIMISMMA,i,j | = | The signed error of the first EPS1 (or EPS2) forecast issued by analyst i for firm j after the M&A announcement date, matched to the last recommendation change within the M&A (− 90 day, + 90 day) window. The signed error is measure as forecast EPS minus I/B/E/S actual EPS, deflated by stock price | |
LTGOPTIMISMMA,i,j | = | Analyst i’s long-term growth (LTG) forecast revision optimism for acquirer j observed during the 180-day window surrounding the M&A announcement date, measured as the difference between analyst i’s outstanding LTG forecast at day + 90 and analyst i’s outstanding LTG forecast at day − 90. Each forecast is standardised by subtracting the consensus forecast among unaffiliated analysts | |
(2) Proxies for analyst optimism in the IPO/SEO context | |||
ROPTIMISMIPO,i,j | = | The standardised level of analyst i’s initial post-issue recommendation within the 2-year window after the IPO issue date of firm j. Recommendations issued under the five-tier ratings system are converted to their three-tier equivalent (‘strong buy’ is recoded to ‘buy’ and ‘strong sell’ is recoded to ‘sell’). Buy-type recommendations are coded 1, hold recommendations are coded 0, and sell-type recommendations are coded − 1 | |
ROPTIMISMSEO,i,j | = | The standardised level of analyst i’s initial post-issue recommendation within the 2-year window after the SEO filing date of firm j. Recommendations issued under the five-tier ratings system are converted to their three-tier equivalent (‘strong buy’ is recoded to ‘buy’ and ‘strong sell’ is recoded to ‘sell’). Buy-type recommendations are coded 1, hold recommendations are coded 0, and sell-type recommendations are coded − 1 | |
FOPTIMISMIPO,i,j | = | The signed error of the first EPS1 (or EPS2) forecast issued by analyst i after the IPO issue date of firm j, matched to the initial recommendation issued in the 2-year window after the issuance of the. The signed error is measure as forecast EPS minus I/B/E/S actual EPS, deflated by stock price | |
FOPTIMISMSEO,i,j | = | The signed error of the first EPS1 (or EPS2) forecast issued by analyst i after the SEO filing date of firm j, matched to the initial recommendation issued in the 2-year window after the issuance of the SEO filing date. The signed error is measure as forecast EPS minus I/B/E/S actual EPS, deflated by stock price | |
LTGOPTIMISMIPO,i,j | Analyst i’s LTG forecast revision optimism for issuing firm j measured by the difference between analyst i’s second LTG forecast in the 24-month post-IPO issue date window and the same analysts first LTG forecast during that window. Each forecast is standardised by subtracting the consensus forecast from Summary file | ||
LTGOPTIMISMSEO,i,j | = | Analyst i’s LTG forecast revision optimism for issuing firm j measured by the difference between analyst i’s second LTG forecast in the 24-month post-SEO filing date window and the same analysts first LTG forecast during that window. Each forecast is standardised by subtracting the consensus forecast from Summary file | |
Independent variables | |||
ChgtoTier3i | = | An indicator variable that equals 1 if the observation pertains to an analyst i, whose employer adopted the three-tier ratings system between May 11 2002 and the end of 2004, and 0 otherwise. All observations for brokers who changed their ratings system are coded 1, including pre-2002 observations | |
POSTREGi,j | = | An indicator variable that equals 1 if the analyst i’s recommendation or forecast for firm j is issued on or after 1 January 2004, and 0 otherwise. While the effective date of the SRO Rules was 23 September 2002, our samples exclude observations in 2002 and 2003 | |
AFFi,j | = | An indicator variable that equals 1 if analyst i’s employer is affiliated with covered firm j, and 9 otherwise. For M&As affiliated analysts are identified as those whose employer acted as an advisor to the covered acquirer in the M&A transaction. For IPOs and SEOs affiliated analysts are identified as those whose employer acted as a lead or co-manager in the underwriting syndicate for the covered firm | |
DEALVALUEj | = | The natural logarithm of the inflation-adjusted deal value (in billions of CPI-adjusted with the year-end dollars in 1992) for firm j. For M&As deal value is the total purchase consideration. For IPOs and SEOs deal value is measured by the proceeds of the offer | |
FIRMCAPj | = | The natural logarithm of the inflation-adjusted market capitalisation of firm j (in billions of CPI-adjusted with the year-end dollars in 1992) as at the last trading day of the calendar month prior to the issue of the recommendation or forecast | |
EXPi | = | The number of years since analyst i first published a recommendation on I/B/E/S | |
FOLLOWj | = | The total number of analysts with outstanding recommendations for firm j in the 30 days preceding the M&A announcement date (or the 30 days preceding the announcement of recommendations in post-IPO/SEO issue date/SEO filing date) | |
DAYSi,j | = | The absolute number of days between firm j’s M&A announcement date (IPO issue date, SEO filing date) and analyst recommendation or forecast announcement date | |
BROKERSIZEi,j | = | The number of analysts employed by analyst i’s brokerage house with outstanding EPS1 forecasts for any firm during the calendar month of firm j’s M&A deal announcement date (or of analyst i’s recommendation or forecast announcement date following firm j’s IPO issue date or SEO filing date) | |
OWNERSHIPj | = | The percentage of firm j’s shareholdings held by institutional investors during the quarter of the M&A announcement date, IPO issue date, or SEO filing date | |
LEVELi,j | = | The level of analyst i’s recommendation for firm j’s stock outstanding 90 days prior to M&A announcement date | |
LTGOPTIMISMpriori,j | = | Analyst i’s LTG forecast optimism for issuing firm j observed immediately prior to the beginning of the LTG forecast revision window. LTG forecast optimism is the difference between the analyst i’s forecast of LTG and the mean LTG forecast of analysts unaffiliated with firm j | |
Variable Used in Testing Conditional Informativeness
INFORM | = | B0 + B1ChgtoTier3i + B2POSTREGi,j + B3ChgtoTier3i × POSTREGi,j + B4PASTFIRMPERFi,j + B5PASTMKTPERFi,j + B6DEALVALUEj + B7FIRMCAPj + B8EXPi + B9FOLLOWj + B10DAYSi,j + B11BROKERSIZEi,j + B12OWNERSHIPj + ε | (1) |
INFORM | = | B0 + B1ChgtoTier3i + B2POSTREGi,j + B3AFFi,j + B4ChgtoTier3i × POSTREGi,j + B5ChgtoTier3i × AFFi,j + B6POSTREGi,j × AFFi,j + B7ChgtoTier3i × POSTREGi,j × AFFi,j + B8PASTFIRMPERFi,j + B9PASTMKTPERFi,j + B10DEALVALUEj + B11FIRMCAPj + B12EXPi + B13FOLLOWj + B14DAYSi,j + B15BROKERSIZEi,j + B16OWNERSHIPj + ε | (2) |
(1) Proxies for informativeness in the M&A and IPO/SEO contexts (INFORM) | |||
INFORMMA,i,j | = | Cumulative abnormal returns for firm j, estimated using the market-adjusted model, measured over the 3-day window centred on the issuance of analyst i’s stock recommendations that cause a change in the outstanding recommendation (i.e., upgrades or downgrades). Observations must fall within ± 90 days of firm j’s M&A announcement date. All CARs for recommendation downgrades are reversed in sign to simplify interpretation | |
INFORMIPO,i,j | = | Cumulative abnormal returns for firm j, estimated using the market-adjusted model, measured over the 3-day window centred on the issuance of analyst i’s recommendations for stock j. Observations must fall within the 24-month window after the IPO issue date. All CARs for recommendation downgrades are reversed in sign to simplify interpretation | |
INFORMSEO,i,j | = | Cumulative abnormal returns for firm j, estimated using the market-adjusted model, measured over the 3-day window centred on the issuance of analyst i’s recommendations for stock j. Observations must fall within the 24-month window after the the SEO filing date. All CARs for recommendation downgrades are reversed in sign to simplify interpretation | |
(2) Independent variables not defined above | |||
PASTFIRMPERFi,j | = | Firm j’s stock return in the 6 months before analyst i issues their recommendation for firm j (for IPO firm j’s stock return is measured between the IPO issue date and the recommendation announcement date if the recommendation is announced within 6 months of the issue date) | |
PASTMKTPERFi,j | = | The cumulative market return over the six months prior to analyst i issuing a recommendation for firm j | |
MKTSDi,j | The standard deviation of the daily S&P 500 index 1 month prior to analyst i issuing a recommendation for firm j | ||
Variable Used in Testing Recommendation Profitabilty
PROFIT | = | B0 + B1ChgtoTier3i + B2POSTREGi,j + B3ChgtoTier3i × POSTREGi,j + B4DEALVALUEj + B5FIRMCAPj + B6EXPi + B7FOLLOWj + B8DAYSi,j + B9BROKERSIZEi,j + B10LEVELi,j + B11OWNERSHIPj + ε | (1) |
PROFIT | = | B0 + B1ChgtoTier3i + B2POSTREGi,j + B3AFFi,j + B4 ChgtoTier3i × POSTREGi,j + B5ChgtoTier3i × AFFi,j + B6POSTREGi,j × AFFi,j + B7ChgtoTier3i × POSTREGi,j × AFFi,j + B8DEALVALUEj + B9FIRMCAPj + B10EXPi + B11FOLLOWj + B12DAYSi,j + B13BROKERSIZEi,j + B14LEVELi,j + B15OWNERSHIPj + ε | (2) |
(3) Proxies for profitability in the M&A and IPO/SEO contexts (PROFIT) | |||
PROFITMA,i,j | = | The buy-and-hold market-adjusted abnormal return for the stock of firm j, starting from the day of the issuance of analyst i’s recommendation for acquirer firm j and terminating a maximum of 1 year later. If the stock recommendation is downgraded, the measurement window terminates the day before the downgrade. All recommendations issued within ± 90 days of firm j’s M&A announcement date are included | |
PROFITIPO/SEOi,j | = | The buy-and-hold market-adjusted abnormal return for the stock of firm j, starting from the day that analyst i issues their initial recommendation for issuing firm j, and terminating a maximum of 1 year later. If the stock recommendation is downgraded, the measurement window terminates the day before the downgrade. Only initial recommendations issued within 24-months of the IPO issue date or SEO filing date are included | |
All other variables are as defined above.
Appendix 2: Measurement of Recommendation Optimism
For our M&A sample, we measure optimism using the method described in Kolasinski and Kothari (2008) and Wu et al. (2015). We obtain analyst recommendations issued for acquirer firms within ± 90 days centred on M&A announcement dates between 1993 and 2014. We reverse the uniform numerical coding of recommendations recorded on I/B/E/S such that 5 = ‘strong buy’ and 1 = ‘strong sell’. For each analyst (i), we also obtain the latest available recommendations issued by that analyst for acquirer j immediately before the commencement of the 180-day event window, subject to the constraint that the recommendation was issued within 2 years of the M&A announcement date. We then define the revision in recommendations, ROPTIMISMMA,i,j, as a categorical variable equal to 1 if, during the 180-day window, the analyst updates their recommendation to a more favourable level, 0 if there is no change in recommendations, and − 1 if the analyst downgrades their recommendation to a less favourable level.
Our proxies for recommendation optimism in our IPO (ROPTIMISMIPO,i,j) and SEO (ROPTIMISMSEO,i,j) samples are the standardised level of analyst i’s initial post-issue recommendation for issuing firm j. Following O’Brien et al. (2005), we obtain analysts initial recommendation within a 2-year window following the IPO issue date (SEO filing date). We reverse the uniform numerical coding of recommendations recorded on I/B/E/S such that 5 = ‘strong buy’ and 1 = ‘strong sell’. Finally, following Corwin et al. (2017), to ensure that the results are not driven by the shift in recommendation scales from a five-tier to a three-tier ratings system, we standardise the observed levels of recommendations so they are all consistent with the use of a three-tier system: for ratings issued under a five-tier ratings system, we recode ‘strong buy’ to ‘buy’ and ‘strong sell’ to ‘sell’. The standardised recommendations are then re-coded as 1 for all ‘buy-type’ recommendations, 0 for ‘hold’ recommendations and − 1 for all ‘sell-type’ recommendations.
Rights and permissions
About this article
Cite this article
Tseng, YJ., Wilson, M. Changes in Recommendation Rating Systems, Analyst Optimism, and Investor Response. J Bus Ethics 166, 369–401 (2020). https://doi.org/10.1007/s10551-019-04139-2
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10551-019-04139-2