Abstract
We investigate how a broker firm’s initial public offering (IPO) affects its analysts’ fiduciary duty of providing independent and objective recommendations. We find that the analysts of newly listed broker firms issue more positively biased recommendations in the first 2 to 3 years after their employers’ IPO than before the IPO. The increase in the recommendation bias is greater among analysts of affiliated brokers and brokers that raise additional capital after their IPO than among other analysts. Newly listed broker firms experience significant increases in revenue and trading commission, and the increases are positively related to recommendation bias, after controlling for many other factors. More importantly, recommendation bias decreases as newly listed broker firms season and as the importance of trading commission declines. This suggests that public exposure through a broker firm’s IPO does not enhance the integrity and professional conduct of its financial analysts. Rather, economic incentives make financial analysts more accepting of unethical behavior. The overall results imply that the public trading of financial gatekeepers compromises the ethical relationship between financial service professionals and society in general.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
See Jennings (2013) and Section II.C for detailed discussions.
The Global Analyst Research Settlement was an enforcement agreement reached in the United States on April 28, 2003, between the ten largest investment banks and the United States Securities and Exchange Commission (SEC), the Financial Industry Regulatory Authority (NASD), and the New York Stock Exchange (NYSE). The ten US investment banks involved in the settlement were fined $1.4 billion because their analysts provided poor and misleading advice about the prospects of internet companies.
See Section II.A for detailed discussions related to the background information in China.
See Section II.A for a detailed discussion on the background information about Chinese IPO and allocation process.
Frank (1985) indicates that employees are willing to accept lower monetary compensations in exchange for the employer’s high status. Focke et al. (2017) find that CEOs of prestigious firms, as proxied by Fortune’s ranking of America’s Most Admired Companies (MAC), earn about 8% less on average than other CEOs of firms not listed on the MACA.
Market forces such as mergers and acquisitions among broker firms have little impact in China because the industry is closely regulated. Allen et al. (2010) show that China was ranked among the worst-scoring countries in terms of legal formalism and property right protection.
This is the latest version of the Notice, which was first released in 2008 and revised in 2009. The changes between these versions, however, are minor; generally speaking, the regulation became more pertinent.
This proportion increased to 30 percent in 2008. For detailed information, see “The Measures for Administration of the Listed Company Issuing New Shares,” which was passed by the CSRC on May 6, 2006, and revised for the first time in October 2008.
For detailed information, please see “The Administrative Measures for the Sponsorships of Securities Issuance and Listing” passed by the CSRC on June 14, 2009. In 2013, the CSRC passed a new stringent regulation. If a newly listed firm’s profit declines more than 50% or suffers losses in the issuing year, the CSRC declines any application from the IPO firm’s sponsor. Thus, IPO sponsors are also under heavy pressure to make sure that the sponsored IPOs perform well in the IPO aftermarkets.
In China, there were 125 investment banks and security brokers by the end of June 2015 and 84 financial consultants and research companies by the end of September 2016. Investment banks in China generally function as security brokers and participate in both underwriting and self-trading. Most importantly, only investment banks and broker firms have trading seats on security exchanges, whereas financial consultants and research companies do not. Thus, financial consultants and research companies provide research and other services rather than trading.
We exclude index funds, bond funds and monetary funds because these funds are passively managed. Both the incentives and pressure for analysts to issue biased recommendations on stocks held by these funds are low. We also exclude funds and broker firms that cannot be matched in these datasets.
Note that although there are 69 broker firms at the end of our sample period in December 2015, the total number of observation does not add up to 2001 (69 × 29) since the number of broker firms varied during the sample period. There were only 8 broker firms at the end of 2002, and the number suddenly increased to 39 in 2003. The number of broker firms peaked at 84 in June 2011. The brokerage firm-year distribution information is available from the authors upon request.
To further investigate whether state ownership in broker firm affects the changes in recommendations, we classify a broker firm as having “positive (non-positive) ownership change” if the percentage of shares owned by state governments increases (does not increase) after the broker firm’s IPO. Similarly, we classify a broker firm as having “large (small) ownership change” if the change in the percentage of shares owned by state governments is greater than (less than or equal to) the median value of changes in the state ownership of all listed brokers in a given year. Then, we compare ∆REC1 and ∆REC2 of broker firms based on the state ownership changes. However, the results are weak and mixed. To save space, the results are not reported, but they are available from the authors upon request.
The number of observations drops in the earlier post-IPO periods because one broker firm covers a few hundred stocks and not all analysts issue recommendations in all periods. Given the fact that committing analyst coverage is a resource allocation decision of brokerage firms, and analysts’ selective coverage itself reflects their true expectation about the covered stocks and their broker firm’s economic incentives (see Das et al. 2006). Another possibility is that if a recommendation does not change from a previous forecast, the new forecast might have been omitted from reporting. To address this possibility, we fill-in the potentially omitted recommendations. Specifically, for example, when analysts of broker firm i issued buy or sell recommendation on stock j in the first half year of 2010 and in 2011 but there was no recommendation in the second half year of 2010, we fill in the same rating in the second half year of 2010 with an assumption that the recommendation remained the same. However, if there are missing observations for more than one consecutive period (say from the second half year of 2010 to first half of 2012), only the first missed recommendation was filled (e.g., in the second half year of 2010). Based on this process, we replicate our tables and obtain qualitatively similar results, which are not reported, but available from the authors upon request.
We also use ∆REC1 as a dependent variable, and the results are qualitatively similar to the results reported here. To save space, the results based on ∆REC1 are not reported but are available from the authors upon request.
Note that changes in state-ownership (\(\varDelta \%\)SO_shares) and in legal person ownership (\(\varDelta \%\)LP_shares) are computed as the change from the year prior to the IPO to the end of the listing year because of data availability.
We also use AF defined at the 25% affiliation level. The results are qualitatively similar to the results based on 50% affiliation level. To save space, we report only the results based on 50% affiliation in the rest of the paper. The results based on 25% affiliation are available from authors upon request.
Note ∆%LP_shares is highly correlated with the other three variables, especially with ∆%SO_shares. For this reason, we exclude ∆%LP_shares from model 5.
Note that, the control variables in the logit regression are measures prior to listing.
Please refer to Appendix 2 for a detailed description of how we constructed the sample for the logit model and how we obtained the matched sample. The empirical results are available from the authors upon request.
We also compute recommendation changes from a 2-year lagged benchmark period (t = − 4, t = − 1) to a 2-year leading period (t = 0, t = 3). We obtain qualitatively similar results, which are not reported due to limited space but are available from the authors upon request. If a longer leading period is used, a broker that is identified as an unlisted broker in period t may be listed in the leading period. For this reason, we prefer to report the results based on the changes from period (t = − 4, t = − 1) to period t = 0.
Note that we do not include SO_share or LP_share because these data are unavailable for unlisted brokers. In addition, Shell-listing and Overseas-listing are irrelevant for unlisted brokers.
We also replace trade commission with total revenue and obtain similar results, which are not reported but are available from the authors upon request.
When the dependent variable is replaced by REC1i,j,t, the results are qualitatively similar to the results reported here. To save space, the results based on REC1i,j,t are not reported, but they are available from the authors upon request.
References
Allen, F., Chakrabarti, R., De, S., Qian, J., & Qian, M. (2010). Law, institutions, and finance in China and India, University of Pennsylvania. Retrieved from http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.496.3503&rep=rep1&type=pdf.
Bogle, J. C. (2017). Balancing professional values and business values. Financial Analysts Journal, 73(2), 14–23.
Boni, L., & Womack, K. L. (2003). Wall Street research: Will new rules change its usefulness? Financial Analyst Journal, 59, 25–29.
Brown, L., Call, A., Clement, M., & Sharp, N. (2015). Inside the ‘Black Box’ of sell-side financial analysts. Journal of Accounting Research, 53, 1–47.
Cain, D. M., Loewenstein, G., & Moore, D. A. (2005). The dirt on coming clean: perverse effects on disclosing conflicts of interest. Journal of Legal Studies, 34, 1-,25.
Chen, S., & Matsumoto, D. A. (2006). Favorable versus unfavorable recommendations: The impact on analyst access to management-provided information. Journal of Accounting Research, 44, 657–689.
Chen, X., Lee, C. J., & Li, J. (2008). Government assisted earnings management in China. Journal of Accounting and Public Policy, 27, 262–274.
Clouse, M., Giacalone, R., Olsen, T., & Patelli, L. (2017). Individual ethical orientations and the perceived acceptability of questionable finance ethic decisions. Journal of Business Ethics, 144, 549–558.
Cowen, A., Groysberg, B., & Healy, P. (2006). Which types of analyst firms are more optimistic? Journal of Accounting and Economics, 41, 119–146.
Das, S., Guo, R. J., & Zhang, H. (2006). Analysts’ selective coverage and subsequent performance of newly public firms. Journal of Finance, 61, 1159–1185.
Davis, M. A., Andersen, M. G., & Curtis, M. B. (2001). Measuring ethical ideology in business ethics: A critical analysis of the ethics positions questionnaire. Journal of Business Ethics, 32, 35–53.
Devos, E. (2014). Are analysts’ recommendations for other investment banks biased? Financial Management, 43, 327–353.
Fang, L., & Yasuda, A. (2009). The effectiveness of reputation as a disciplinary mechanism in sell-side research. Review of Financial Studies, 22, 3736–3777.
Firth, M., Lin, C., Liu, P., & Xuan, Y. (2013). The client is king: Do mutual fund relationships bias analyst recommendations? Journal of Accounting Research, 51, 165–200.
Fisman, R., Shi, J., Wang, Y., & Xu, L. (2016). Social ties and favouritism in chinese science. Journal of Political Economy (Forthcoming).
Focke, F., Maug, E., & Niessen-Ruenzi, A. (2017). The impact of firm prestige on executive compensation. Journal of Financial Economics, 123, 313–336.
Fombrun, C., & Shanley, M. (1990). What’s in a name? Reputation building and corporate strategy. Academy of Management Journal, 33, 233–258.
Forbes, W. (2013). No conflict, no interest: On the economics of conflicts of interest faced by analysts. European Journal of Law and Economics, 35, 327–348.
Forsyth, D. K., Nye, J. L., & Kelley, K. (1988). Idealism, relativism, and the ethic of caring. Journal of Psychology, 122, 243–248.
Francis, J., Hanna, J. D., & Philbrick, D. R. (1997). Management communications with securities analysts. Journal of Accounting and Economics, 24, 363–394.
Frank, R. H. (1985). Choosing the right pond: Human behavior and quest for status. Oxford: Oxford University Press.
Gasparino, C. (2005). Blood on the street: The sensational inside story of how Wall Street analysts duped a generation of investors. New York: Free Press.
Grant, A., Jarnecic, E., & Su, M. (2015). Asymmetric effects of sell-side analyst optimism and broker market share by clientele. Journal of Financial Markets, 24, 49–65.
Griskevicius, V., Tybur, J., Gangestad, S., Perea, E., Shapiro, J., & Kenrick, D. (2009). Aggress to impress: Hostility as an evolved context-dependent strategy. Journal of Personality and Social Psychology, 96, 980–994.
Gu, Z., Li, Z., & Yang, Y. G. (2012). Monitors or predators: the influence of institutional investors on sell-side analysts. The Accounting Review, 88, 137–169.
Haveman, H., Jia, N., Shi, J., & Wang, Y. (2016). The dynamics of political embeddedness in China. Administrative Science Quarterly (Forthcoming).
Hong, H., & Kubik, J. D. (2003). Analyzing the analysts: Career concerns and biased earnings forecasts. Journal of Finance, 58, 313–351.
Huang, H., Li, M., & Shi, J. (2016). Which matters: “paying to play” or stable business relationship? Evidence on analyst recommendation and mutual fund commission fee payment. Pacific-Basin Journal of Finance (Forthcoming).
Huyghebaert, N., & Wu, W. (2015). What determines the market share of investment banks in Chinese domestic IPOs? China Economic Review, 34, 150–168.
Jackson, A. (2005). Trade generation, reputation, and sell-side analysts. Journal of Finance, 55, 673–717.
Jennings, M. (2013). Ethics and financial markets: The role of analyst. Research Foundation of CFA Institute, 1–7.
Kang, J., Liu, M. H., & Ni, S. X. (2002). Contrarian and momentum strategies in the China stock market: 1993–2000. Pacific-Basin Finance Journal, 10, 243–265.
Kedia, S., Rajgopal, S., & Zhou, X. (2014). Did going public impair Moody’s credit ratings? Journal of Financial Economics, 114, 293–315.
Lin, H. W., & McNichols, M. F. (1998). Underwriting relationships, analysts’ earnings forecasts and investment recommendations. Journal of Accounting and Economics, 25, 101–127.
Ljungqvist, A., Malloy, C., & Marston, F. (2009). Rewriting history. Journal of Finance, 64(4), 1935–3960.
Maber, D. A., Groysberg, B., & Healy, P. H. (2014). The use of broker votes to reward broker firms’ and their analysts’ research activities. Retrieved May 12, 2016, from University of Michigan and SSRN website: http://papers.ssrn.com/sol3/papers.cfm?abstract id = 2311152.
Malmendier, U., & Shanthikumar, D. (2014). Do security analysts speak in two Tongues? Review of Financial Studies, 27, 1287–1322.
Mehran, H., & Stulz, R. M. (2007). The economics of conflicts of interest in financial institutions. Journal of Financial Economics, 85(2), 267–296.
Michaely, R., & Womack, K. L. (1999). Conflict of interest and the credibility of underwriter analyst recommendations. Review of Financial Studies, 12, 653–686.
Mikhail, M. B., Walther, B. R., & Willis, R. H. (1999). Does forecast accuracy matter to security analysts? Accounting Review, 74, 185–200.
Mola, S., & Guidolin, M. (2009). Affiliated mutual funds and analyst optimism. Journal of Financial Economics, 93, 108–137.
Piotroski, J., Wong, T. J., & Zhang, T. (2015). Political incentives to suppress negative information: Evidence from Chinese listed firms. Journal of Accounting Research, 53, 405–459.
Piotroski, J., & Zhang, T. (2014). Politicians and the IPO decision: The impact of impending political promotions on IPO activity in China. Journal of Financial Economics, 111, 111–136.
Podolny, J. M. (1993). A status-based model of market competition. American Journal of Sociology, 98, 829–872.
Teoh, S. H., Welch, I., & Wong, T. J. (1998). Earnings management and the long-run market performance of initial public offerings. Journal of Finance, 53, 1935–1974.
Veit, E. T., & Murphy, M. R. (1996). Ethics violations: A survey of investment analysts. Journal of Business Ethics, 15, 1287–1297.
Weiss, Y., & Fershtman, C. (1998). Social status and economic performance: A survey. European Economic Review, 42, 801–820.
Yang, Z. (2013). Do political connections add value to audit firms? Evidence from IPO audits in China. Contemporary Accounting Research, 30, 891–921.
Acknowledgements
The authors thank colleagues at Bowling Green State University, RMIT University, Australian National University, and Jiangxi University of Financial Economics for their valuable comments and discussions. Our gratitude also goes to Kenn Phua, Gao Shenghao, and the discussants and participants who provided insightful comments and suggestions at the Asian Finance Association 2017 Annual Meeting (Beijing, China) and at the 2017 Vietnam International Conference in Finance. The usual disclaimer applies.
Funding
This study was funded by 2016 EFM Research Grant at RMIT University.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
Each of the three authors declares that he/she has no conflict of interest.
Appendices
Appendix 1: Variable Definitions
Variable | Definition |
|---|---|
REC1i,j,t | The relative recommendation, computed by subtracting the median value of the recommendation for stock i by all brokers from the raw recommendation of broker firm j for stock i in period t. Raw recommendation takes a value from 1 (strong sell) to 5 (strong buy) |
REC2i,j,t | The relative recommendation, calculated by subtracting the median value of the recommendation for stock i by all unlisted brokers from the raw recommendation of broker firm j for stock i in period t |
AFi,j,t (nonAF) | Broker j is defined as an affiliated broker if stock i covered by broker j is held by fund k, and fund k’s fund management company (FMC) has a common ownership with broker firm j. Other brokers are called non-affiliated, nonAF |
AFi,j,t (25% or 50%) | Takes a value of 1 if broker j owns no less than 25% (50%) the FMC’s total number of shares or if a third party owns no less than 25% (50%) of both broker firm j’s and the FMC’s total shares; the value is 0 for other brokers |
∆AF dummy = 1 | Indicates that stock i is held by at least one of covering broker j’s affiliated funds and that the fund’s FMC has at least 50% (or 25%) joint ownership with broker firm j in at least one of the four periods within the period (t = 0, t = 3) but not in any of the lagged four periods (t = − 4, t = − 1). ∆AF dummy = − 1 indicates the opposite |
Underwriting relation | Takes a value of 1 if a broker has any underwriting business with the covered stock prior to period t and 0 otherwise |
∆Underwriting relation = 1 | Indicates that broker j did not have any underwriting business with firm i in any period prior to period t but has had such underwriting business since period t |
NoS | Number of stocks covered by a broker |
NoB | Number of brokers covering a stock |
%SO_shares | State ownership, measured as the percentage of total shares outstanding |
%LP_shares | Legal person (i.e., non-state institutional) ownership, measured as the percentage of total shares outstanding |
Refinance | Takes a value of 1 if a broker firm conducted refinancing in period (t = 0, t = 3) after the IPO period in t = 0 and zero otherwise |
Shell-listing | Takes a value of 1 if a broker firm conducted IPO through a shell firm and zero otherwise |
Top 10 auditor | Takes a value of 1 if a broker firm’s auditor is ranked in the top 10 auditors by the Chinese Certified Public Accounting Association in year t and 0 otherwise |
Overseas-listing | Takes a value of 1 if a broker firm is listed in foreign stock markets and 0 otherwise |
HQ dummies | Dummy variables control for broker firms’ headquarters locations |
TA | Total assets of a firm, measured in billions RMB |
State ownership | Stock firms’ percentage of shares owned by state governments |
ROA | Return on assets of the firm |
Ind. Adj. CAR | A firm’s industry-adjusted monthly cumulative abnormal returns |
Newly listed | A broker that it has just been listed in period t. The dummy variable takes a value of 1 if the broker is newly listed and 0 otherwise |
Already-listed | A broker that was already listed prior to period t. The dummy variable takes a value of 1 if the broker was already listed in period t and 0 otherwise |
Still-unlisted | A broker that is not yet listed as of period t. The dummy variable takes a value of 1 if the broker is still not listed in period t and 0 otherwise |
Ln(nonAF25%_hld) | Stock holding by funds that are not affiliated with the brokers, defined at 25% 25% ownership. Ln(nonAF50%_hld) is defined in a similar way |
Ln(Cap) | Broker firm’s net capital, measured in million RMB |
Ln(Trd.Com) | Broker firm’s trade commission in million RMB |
Ln(Age) | The natural logarithm of broker firm age |
Post-listing | Takes a value of 1 if the recommending broker firm was listed prior to period t and 0 otherwise |
Listed | Takes a value of 1 if the broker firm is listed by the end of the sample period and 0 otherwise |
REC_AF & REC_nonAF | This variable has four variations, REC1_AF50%, REC1_AF25%, REC2_AF50%, and REC2_AF25%. They measure the average recommendation bias (using either REC1 or REC2) of all affiliated brokers (with either 50% or 25% ownership connection). REC_nonAF is defined similarly for non-affiliated brokers |
\(\Updelta\)REC_AF & \(\varDelta\)REC_nonAF | Change in REC_nonAF and REC_nonAF from period t-1 to t |
IPO age | The number of semi-annual periods since a broker firm’s listing |
TCP1 (TCP2) | The proportion of a broker firm’s trading commission to its total revenue (total commission) |
High-TCP1 | Takes a value of 1 if a broker’s TCP1 is greater than the median value of all brokers in a given period and 0 otherwise |
Appendix 2: Propensity Score Matching for Listed Broker Firms
To obtain the listing probabilities for both listed and unlisted brokers, we depend on brokers’ financial performance prior to listing as the major determinants. The sample is created as follows:
-
1.
Annual financial data for listed brokers
We include 2 years of annual financial data for listed brokers before their listing dates since this the basis for the CSRC’s judgement on whether a broker firm meets the listing criteria.
-
2.
Annual financial data for unlisted brokers
We assume that the potential listing candidates comprise all unlisted brokers for each listing year (i.e., the year in which at least one broker is listed). Consequently, all unlisted broker firms are in the sample for every listing year.
-
3.
Compute the pre-listing or hypothesized listing performance
For each financial variable, we calculate the average value over the 2 years before the listing or hypothesized listing periods as the proxy for pre-listing performance.
We then run a logit regression using the sample constructed above. The dependent variable takes a value of one if the broker in question was listed by the end of our sample period and zero otherwise. The independent variables are the logarithm of listing age (i.e., firm age at the time of listing), average profitability (i.e., revenue divided by net capital), the logarithm of the average net capital, and the average growth rate of trading and non-trading commission. There are also two dummy independent variables: (Trd.Com_D1 and Non-Trd.Com_D1). The former equals one if in at least one of the 2 years prior to its listing, a broker’s trade commission was no less than the median value of all broker firms in the corresponding period and zero otherwise. The latter takes a value of one if in at least one of the 2 years prior to its listing, a broker’s non-trade commission was no less than the median value of all broker firms in the corresponding period and zero otherwise. The non-trade commission fee is computed as total commission minus trade commission. Alternatively, we define Trd.Com_D2 and Non-Trd.Com_D2 as dummies that equal one if in both years prior to its listing, a broker’s trade (non-trade) commission is no less than the median value and zero otherwise. In addition, the non-trading commission fee is computed as total commission minus trading commission.
We proceed to propensity score matching after obtaining the listing probability for listed and unlisted brokers. We compare the differences in the estimated probabilities and financial measures between the two groups of brokers according to each of the following matching methods: one-to-one with replacement using Caliper, one-to-one with and without replacement using nearest neighbor, and one-to-three and one-to-five without replacement using nearest neighbor. Finally, we decide to use the matched sample obtained through one-to-one without replacement using the nearest neighbor method by taking into consideration both the number of resulted matching observations and the similarity (or closeness) in financial variables between the two groups.
Rights and permissions
About this article
Cite this article
Huang, H., Li, M. & Shi, J. Should Financial Gatekeepers be Publicly Traded?. J Bus Ethics 164, 175–200 (2020). https://doi.org/10.1007/s10551-018-4044-6
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10551-018-4044-6