Brokerage research has a built-in tension: investors want an honest assessment of a security, while the firm publishing that assessment may earn money from trading it, selling it or advising its issuer. A research report can contain useful analysis and still come from a business with competing interests.
The history of research conflicts is therefore not simply a record of inaccurate stock recommendations. It concerns who pays analysts, who controls their work and what happens when an unwelcome opinion threatens a profitable relationship. Within the broader history of stock brokers, research became both a service for investors and a commercial asset for securities firms.
The American experience provides the clearest thread, from commission-funded research to the analyst scandals surrounding the technology boom. Later European and British payment reforms addressed another part of the same problem: how to pay for analysis without distorting investment decisions.
Professional Analysis Did Not Remove Commercial Pressure
Investment analysis developed professional institutions well before the internet stock boom. Benjamin Graham and David Dodd published Security Analysis in 1934. American analyst societies established a national federation in 1947, and the first CFA examination followed in 1963. These milestones appear in the CFA Institute’s historical timeline.
Professional standards addressed how analysts should investigate businesses and present their judgments. Commercial independence posed a different question. An analyst could be skilled at reading accounts yet work within an organization that benefited from a particular recommendation.
Consider the distinction between analytical error and a conflict of interest. An analyst might overestimate a manufacturer’s future sales after making a reasonable but unsuccessful judgment. That is not, by itself, evidence of misconduct. A different problem arises if the analyst softens a warning because the employer wants to arrange the manufacturer’s next share offering.
The first concerns the quality of a forecast. The second concerns the incentives behind it. A sound assessment of brokerage research needs to examine both, rather than treating every losing recommendation as dishonest or every profitable recommendation as independent.
Fixed Commissions and the Economics of Research
Before negotiated commissions became standard in the United States, fixed exchange commission schedules left brokers competing through services as well as execution. Research could help attract institutional orders. Fund managers also used commission arrangements to reward brokers for investment ideas.
On May 1, 1975, the SEC ended fixed brokerage commissions. Congress subsequently introduced Section 28(e), a conditional safe harbor allowing investment managers to pay more than the lowest available commission for qualifying brokerage and research services. The manager had to determine in good faith that the commission was reasonable relative to the services received. The SEC’s inspection report on soft dollar practices traces both the earlier arrangements and the resulting conflicts.
These “soft dollar” arrangements connect an investment manager’s research purchases with commissions paid from client accounts. The manager benefits from research without necessarily paying for it from the management firm’s own resources. That creates a tension between obtaining useful analysis and controlling clients’ trading costs.
A hypothetical example makes the distinction clearer. A fund manager might choose a broker charging a higher commission because its industry research helps evaluate several portfolio holdings. The research could be valuable. But the manager should still ask whether the extra cost is justified, rather than treating the research as free because no separate invoice arrives.
The wider consequences of May Day 1975 and the end of fixed US commissions extend beyond research. For this subject, the central point is narrower: changing the price of execution did not settle how analysis should be funded.
The Technology Boom Exposed the Banking Conflict
By the late 1990s and early 2000s, the conflict between research and investment banking had become a major regulatory concern. Analysts could help investors assess companies while also helping their employers win corporate business. Those objectives were not always compatible.
The joint NASD and NYSE report on analyst conflicts documented practices that compromised research independence. These included investment bankers influencing analyst bonuses, analysts supporting banking transactions, personal investments in companies they covered, and pressure from issuers and major shareholders to maintain favorable ratings.
The commercial logic was straightforward. A company considering an initial public offering wanted investor interest. A securities firm wanted the underwriting assignment. A respected analyst could make the firm’s offer more attractive by providing industry knowledge and potential visibility among investors. Trouble arose when an expectation of favorable coverage became entangled with the competition for that assignment.
Pressure need not take the form of an instruction to publish a false number. It can affect which risks receive attention, how much confidence a report expresses or whether an analyst challenges management’s assumptions. A valuation model can contain defensible inputs individually while assembling them into an implausibly cheerful whole.
This is why independence cannot be assessed solely by checking arithmetic. The choice of assumptions matters as much as the spreadsheet’s ability to multiply them.
The 2003 Global Research Analyst Settlement
On April 28, 2003, US regulators announced enforcement settlements with ten major securities firms over conflicts between research and investment banking. The SEC’s announcement of the global analyst settlement put the combined payments at roughly $1.4 billion.
That headline amount was not entirely a fine. It included penalties, disgorgement, funding for independent research and investor education. The distinction matters because the response combined punishment with attempts to change how investors received information.
The settlement also imposed organizational restrictions on the participating firms. These included separating research from investment banking, removing banking influence over research budgets and restricting analysts’ participation in banking pitches and roadshows. The aim was to change working relationships, not simply add another paragraph of disclosure.
The settlement was an enforcement agreement affecting named firms, rather than a universal statute governing every research provider. It should also be distinguished from the broader analyst rules developed during the same period.
Its historical importance lies in the regulatory approach: when a conflict is embedded in supervision and compensation, disclosure alone may be an inadequate response. An investor can read a warning, but cannot personally supervise the analyst’s employer.
Certification and Rules for Research Independence
Making analysts accountable for their stated views
A parallel reform addressed whether published opinions actually reflected the analyst’s beliefs. Regulation Analyst Certification, or Regulation AC, took effect on April 14, 2003. The SEC’s Regulation AC adopting release required certifications concerning the analyst’s personal views and disclosures about compensation connected to the recommendations or views expressed.
Certification addresses honesty, not foresight. An analyst can sincerely believe a company is undervalued and still be wrong. Equally, a recommendation does not become acceptable simply because the stock happens to rise afterward.
For readers, this distinction prevents an easy misunderstanding. Regulatory language in a report does not mean that a regulator has approved its valuation, endorsed its recommendation or guaranteed its accuracy. It establishes obligations around the production and presentation of the research.
Controlling supervision and compensation
FINRA Rule 2241 on equity research analysts and reports requires member firms to maintain policies addressing research conflicts. Its provisions include restrictions on investment banking review of reports, banking control over analysts and compensation based on particular banking transactions or contributions to banking activities. It also requires safeguards against pressure and retaliation for unfavorable research.
These controls address several routes through which commercial preferences could reach an analyst. Protecting the final report would accomplish little if someone with a conflicting interest could control the author’s bonus or punish an unwelcome recommendation.
Still, organizational safeguards should not be confused with the removal of every incentive. Research operates within a business. The practical question is whether its arrangements allow analysts to publish an unfavorable assessment when doing so is commercially inconvenient.
Europe Shifted Attention to Who Pays for Research
The American reforms concentrated heavily on the relationship between analysts and investment bankers. European research payment reforms focused on another relationship: the one between brokers, investment managers and the clients whose assets paid for services.
MiFID II research unbundling requirements took effect in January 2018. Under the framework examined in the FCA’s historical review of research unbundling, investment managers generally paid for third-party research from their own resources or through a separately controlled research payment account. Brokers had to price research separately from execution.
The economic purpose was to make the purchasing decision more visible. Rather than treating research as something that arrived alongside trading, managers had to assess what they were buying and how it was funded.
Separate pricing does not establish that a report is correct. It makes a different question easier to ask: is this research worth its cost? It also helps distinguish the decision to purchase analysis from the decision about where to execute a trade.
That creates a trade-off worth stating without exaggeration. Charging separately can improve scrutiny, but research still needs paying customers. A funding model should be judged on both accountability and its ability to support useful work, not on whether its invoices look tidier.
The UK Added Payment Flexibility in 2024
The reform process did not end with mandatory separation. From August 1, 2024, the FCA introduced an additional option allowing qualifying UK firms to make joint payments for third-party research and execution, subject to requirements. The change was set out in FCA policy statement PS24/9 on research payment optionality.
This was a conditional payment option, not an instruction to abandon cost controls. It illustrates why the history is not a simple progression from bundled payments to permanent separation.
There are two questions here. Who should bear the research expense? And what controls should apply when that expense is connected to trading? Different payment structures can answer them differently. None makes scrutiny unnecessary.
How to Read Brokerage Research With This History in Mind
The practical lesson is not to discard brokerage research. It is to separate the analysis from the commercial setting in which it was produced. A report can be useful without being the final word on a company.
Start with the disclosed relationships. Look for information about banking work, financial interests and the source of the research. Ask who produced the report, rather than assuming the platform displaying it is also its author. A third-party label is a starting point for investigation, not a certificate of neutrality.
Then examine the argument beneath the recommendation. Does the valuation depend on unusually strong revenue growth, wider profit margins or a higher valuation multiple? What would have to happen for the thesis to fail? A report that makes those conditions clear is easier to evaluate than one offering a confident target with little explanation.
In a hypothetical comparison, two analysts might assign the same price target to a retailer. One expects rapid store expansion; the other expects better profits from existing stores. The matching targets conceal different risks. Reading only the headline would miss the disagreement that matters.
Finally, keep research conflicts separate from other brokerage incentives. The economics discussed in the history of commission-free brokerage and payment for order flow concern a different part of the business. A low trading charge does not answer questions about research independence, just as an independent report does not establish the quality of trade execution.
Across these historical changes, the most useful test remains straightforward: can the analyst reach and publish an unwelcome judgment, and can the reader see enough of the reasoning to challenge it? Neither a familiar firm name nor a polished price target answers that question on its own.