They are supposed to be among Wall Street’s most closely guarded secrets: changes in research analysts’ views, up or down, of a company’s prospects. But some of the nation’s biggest brokerage firms appear to be giving a handful of top hedge funds an early peek at these sentiments — allowing them to trade on the information before other investors get the word.

The signals come from questionnaires that analysts answer and submit electronically, either monthly or quarterly, to some of their firms’ largest hedge fund clients. Chief among the questions posed to the analysts are those about possible earnings surprises at companies they follow.

What analysts tell investors about the companies they follow — and when — is central to the concept of a level playing field on Wall Street. When disseminated, analyst downgrades and upgrades can make a stock sink or soar. Getting that information early can be very profitable for traders. As a result, regulatory rules require brokerage firms to restrict the information flow from research departments to prevent the potential for trading ahead of research reports.

Questions about the selective release of analysts’ views came up when the brokerage firms charged with selling Facebook’s initial shares were found to have warned large buyers about some analysts’ doubts regarding the company’s prospects. That irked many small investors who had not received the guidance and sustained losses in their Facebook shares. The Securities and Exchange Commission is investigating these disclosures.

But documents obtained by The New York Times indicate that the hedge fund practice of trawling for analysts’ shifting views is systematic and growing on Wall Street. Questionnaires completed by analysts that can telegraph their thinking are being used by hedge funds run by BlackRock; Marshall Wace, a large British hedge fund company; and Two Sigma Investments, a United States hedge fund concern.

The funds say they ask only for public information, but in at least four cases, documents from Barclays Global Investors, now a unit of BlackRock, state the goal is to receive nonpublic information. Two documents state that the surveys allow for front-running analyst recommendations.

To the degree that these surveys alert select hedge fund clients to future shifts in researchers’ views or estimates, analyst participation in them may contradict firms’ policies stating that research is distributed to all clients simultaneously.

Analysts at many companies, including Citigroup, Credit Suisse, Deutsche Bank, Goldman Sachs, JPMorgan Chase, Merrill Lynch and UBS, have participated in the programs. The analysts’ answers are fed into the hedge funds’ trading algorithms, determining which stocks to buy or sell.

As one of the largest payers of trading commissions to Wall Street, BlackRock is a client that brokerage firms want to please to keep those commissions flowing. For a period during 2008 and 2009, the firms participating in an expanded survey received cash based on their results.

BlackRock Scientific Active Equity, the money management unit that was part of Barclays Global Investors before it was bought by BlackRock in 2009, has been a leader in these surveys. One of the company’s survey questions, titled “earnings surprise direction,” asks analysts whether a company’s coming profits “are more likely to surprise on the upside or downside.” Another asks: “Do you think the current consensus earnings forecast” for a specified period “will likely move upwards or downwards?”

A more recent question involves takeovers: “How likely is it that the company will be taken over in the next 6 months?”

The BlackRock surveys are careful to ask that analysts supply only those views that they have already stated publicly. But in various confidential documents describing the surveys, company officials state that nonpublic information is what they are after. “We expect the earnings surprise direction to be able to capture the information not released to the market,” stated a confidential BlackRock memo from November 2008, detailing its analyst surveys of nine brokerage firms in Asia. “The question may give the clue on the direction of the analyst’s future revisions.”

A 2009 document on the firm’s analyst surveys is even more explicit. “We are trying to front-run recs,” it said, referring to trading ahead of analysts’ recommendations.

James Badenhausen, a BlackRock spokesman, said, “The language in the Barclays Global Investors internal memos is sloppy and inaccurate and totally inconsistent not only with the stated purpose of the survey but also with the high ethical standards by which BlackRock does business. The survey explicitly states that it requests only information that sell-side research analysts have already disseminated publicly and an analyst cannot even take part in the survey without first clicking a button to confirm that answers would be based solely on public information.”

He continued in a statement, “The surveys allow the group to quantify information from tens of thousands of analyst research reports so they can feed that data into the computer models used in the group’s investment process. The data from these online surveys are just one of more than 100 different factors that the group feeds into the models to determine investment decisions.”

Mr. Badenhausen declined to make any other BlackRock executive available for comment.

Representatives for Citigroup, UBS and JPMorgan Chase said they had strict policies in place for analysts participating in client surveys requiring that their comments were consistent with their publicly held views. Representatives at Credit Suisse, Deutsche Bank, Bank of America and Goldman Sachs declined to comment.

Lewis D. Lowenfels, an authority on securities law in New York City, said that while such analyst surveys might be legal, they could be improper if they were being used to mask the transfer of nonpublic information. “The firms have to be meticulously careful about how they monitor and supervise these interactions,” Mr. Lowenfels said, “because it would be so easy to cross the line.”

Plumbing for Information

Making sure that ordinary investors are not duped by bad research or by opinions disseminated long after big investors have already had a chance to trade on them has been a concern of regulators in years past, leading to more rigorous rules about analysts’ compensation and their independence.

The pronouncements of powerful Wall Street analysts have long been known to move stock prices.

New recommendations to buy or sell a stock from major brokerage firms produce big initial returns.

A 2004 study, titled “The Value of Client Access to Analyst Recommendations” and written by T. Clifton Green, a professor at Emory University’s business school, confirmed the profit potential. Responding rapidly to announcements of changes in stock recommendations gave brokerage firm clients average two-day returns of 1.02 percent and annualized gains of more than 30 percent, the study found.

Barclays Global Investors recognized how profitable such information could be. A confidential internal update from late 2007 describing the company’s European analyst survey noted that revisions in analyst forecasts were a crucial driver of returns. “We observe a contemporaneous one-month correlation between revisions and returns on the order of 15-20 percent over many years and many markets,” the update noted.

Only trouble is, the report said, the moves run out of steam relatively quickly.

Asking analysts questions about earnings direction “may allow us to get ahead of revisions,” the document said.

The 1 to 5 Scale

At the time, Barclays had enlisted six brokerage firms to participate in the analyst survey, providing coverage of 88 percent of large European companies. Asked to respond on a scale of 1 to 5, an analyst giving a 1 to a question about the expected direction of earnings at a particular company signals a significant downward revision. An answer of 5 indicates the opposite. A response of 3 is considered neutral, the document said.

The 2007 presentation identified some of the best returns generated by the program. Many involved financial stocks that analysts were beginning to view negatively as the credit crisis deepened. They included Northern Rock, a British bank that was later nationalized, and Bradford & Bingley, a regional British bank and mortgage lender also taken over by the government.

The document shows how analysts’ responses about Bradford & Bingley not only preceded public downgrades but also most likely proved profitable for those trading on them. Surveys compiled on Aug. 31, 2007, generated negative signals, with an average response of 2 to the earnings questions.

On Sept. 5, not even a week later, Lehman Brothers downgraded the stock, noting uncertainties in asset-backed markets and rising financing costs for the bank. The stock plummeted, and the company was later taken over. Many shareholders were wiped out.

A more upbeat case study from the report concerned the Renewable Energy Corporation, a Norwegian company. In surveys tabulated on May 31, 2007, an analyst from Credit Suisse responded to earnings outlook questions with 5s, positive indications. On June 24, Credit Suisse upgraded its earnings forecast for the company by 7 percent.

A month later, Renewable Energy announced an 11 percent increase in earnings per share from a year earlier. Between the day the analyst surveys came in and the earnings announcement, the company’s shares rose 22 percent.

The predictive nature of the surveys led Barclays to introduce more subjective questions. These included queries about management quality, innovation and competitiveness, according to an April 2008 internal analysis from research officials in its United States, Australian and European units. Among the brokerage firms participating in the “softer” surveys were UBS, Citigroup, Goldman Sachs, Lehman and Merrill Lynch.

“The questions are vague but collectively give us a good sense of the analyst’s overall sentiment towards the company,” the report concluded. “We find that this sentiment manifests itself in future analyst upgrades and in the Australian case, positive EPS revisions.”

In 2008, a paper prepared by Barclays researchers for the firm’s European, United States, Japan and Asian units analyzed the performance of recent surveys. “The results support prior evidence that our signal enables front-running individual analyst recommendations,” the report concluded.

One of the first companies to devise a way to use surveys to generate outsize returns was Marshall Wace, a large British hedge fund operator. Its program, known as Trade Optimized Portfolio System, asked for information from sales traders in addition to analysts.

In 2009, it was capturing investment ideas from 290 brokerage houses and 1,900 equity sales representatives, sector specialists, country specialists and strategists at investment firms around the world, according to a report from Albourne, an independent analysis firm. A Marshall Wace spokesman said that currently the firm’s system has 3,000 brokerage firm sales representatives participating.

At Barclays, conducting a poll of equity analysts for changes in sentiment was first proposed in 2003 in Australia, where financial regulation is more relaxed. It was introduced there in 2006.

The surveys were being tested in the United States in 2005 and were expanded to Europe and Asia in 2007.

Other hedge funds have created their own surveys. Two Sigma Investments, for example, set up a system called Portfolio Idea Contribution System that is patterned after the Marshall Wace program. A Two Sigma spokesman declined to comment.

Spotty Oversight

It is unclear whether securities regulators in the United States are concerned about these brokerage firm surveys. A spokesman for the S.E.C. declined to comment.

But regulators in Europe have discussed the possibility of improprieties. For example, the British securities regulator, the Financial Services Authority, said in a 2006 advisory that it had met with firms using these systems and warned about their potential for transmitting nonpublic information.

An official at the F.S.A. said last week that he was unaware of any investigations conducted into the use of surveys to generate outsize returns.

Many hedge funds in Europe have concluded that the F.S.A. has blessed the programs. Nevertheless, it is clear that regulators view trading ahead of analyst changes to be a misuse of nonpublic information. In July 2007, the Committee of European Securities Regulators said insider information included “the coming publication of research, recommendations or suggestions concerning the value of listed financial instruments.”

Regulators also take a dim view of sharing analysts’ changing views with select customers. In a recent case against Goldman Sachs, the S.E.C. fined the company $22 million for allowing select clients to have private meetings with analysts. The S.E.C. did not charge the company with tipping the customers improperly but cited the potential for analysts’ changing views to be misused.

Other regulators are scrutinizing how brokerage firm research is distributed. The Financial Industry Regulatory Authority proposed an extensive new rule covering equity research reports in late 2008. The rule says firms must have “procedures reasonably designed to ensure that a research report is not distributed selectively to internal trading personnel or a particular customer or class of customers.” But the rule has not been put into place. After receiving comments, the authority has decided to broaden the proposal to cover debt securities. A new version is expected later this year.

Firms that selectively share their analysts’ research could add to unease about fairness on Wall Street, said Michael Clement, a professor of accounting at the University of Texas, Austin, who has a background in banking. “As investors’ portfolios become more global, investors will seek opportunities where they believe the playing field is level,” he said. “Analysts’ surveys open the door to potential problems because firms could start communicating things to some clients that they are not communicating to all of them. And that’s wrong.”