Banks Want Analysts to Earn Their Keep

Once you know the value of research, it's harder to justify a price of zero.

A living, breathing environment.

Photographer: Patrick T. Fallon

I sometimes say that modern finance operates as a gift economy in which you constantly fling free stuff at clients in the hopes that one day they will give you some extremely overpriced paying work. Maybe the purest and best-known example of this is sell-side research. The job of a research analyst is to produce ideas, and then to fling them at clients at no charge. The client, touched by her generosity and impressed by her ideas, will then reward the research analyst by using her bank to trade stocks at a commission of three cents a share. 1  Or that's the idea. Three cents a share may not seem like a great racket as far as extremely overpriced paying work goes, but remember that you can literally trade stocks for free on your phone, and yet somehow institutional investors pay $10 billion a year in stock trading commissions. 2

The problem with this gift economy, from the research analyst's perspective, is that she only gives and never receives. If a client thinks a research analyst is a genius, he rewards her by sending commissions to her bank's salespeople and traders, who may well be dolts. But they are dolts with a P&L; she is a genius cost center. That is not great for her compensation, or her career prospects. It is not unique to her, of course; a lot of bank employees are rewarded out of trading revenue despite the difficulty of tying them directly to that revenue. 3  But it seems especially acute for research, and it is not surprising that banks have long tried to figure out better ways to quantify the value of research analysts. 4

Here's a new one:

Wall Street banks may have finally hit on a way to pinpoint the value of analysts and squeeze more money from their research: Stop making it so easy to share.

Bank of America Corp. has started embedding analysts’ reports into web pages, so it can more easily restrict access than with PDF files that are widely shared with people who aren’t paying clients, said Candace Browning, the firm’s head of research. It’s joining rivals Morgan Stanley and Citigroup Inc. in limiting access, and more plan to follow. The approach also makes it easier to track analysts’ readership and customize products for specific types of clients, according to bank executives and consultants.

I guess this is just changing the file format used to transmit research, but it still feels like a cultural shift. For one thing, it sounds ... really client unfriendly? "Analysts and sales staff have for years made it as easy as possible for clients to get reports," because they are in the client-service business, and making life difficult for clients is a bad client-service strategy. I mean this sounds sort of awful:

Under the new system, clients must access a site to view material that stays there, much as they would peruse a favorite newspaper behind a paywall, according to Daire Browne, Bank of America’s chief operating officer for global research. The pages are more dynamic than a PDF and will have more security, making them harder to recirculate, he said.

“You’re not accessing a static PDF, you’re going into a website and you are authenticating,” Browne said. “That’s the whole premise here, that you have a greater ability to control the access coming in when it’s a living, breathing environment that we control.”

At Citigroup, clients receive an e-mail with just a few sentences about a report. Clicking a link takes them to a website where they can sign in to read the rest. 

So ... can you print it out (or save it on your phone) and read it on the subway? Or are you stuck in the living, breathing, controlled, but possibly somewhat stuffy and claustrophobic environment? 5  

The bigger cultural shift, though, is that if you treat research as a product, then you might start selling it as a product. Or products. With prices:

The next step, still to come for most banks, is to customize offerings to specific clients. Citigroup, Morgan Stanley and others are part of a group that has come up with a coding language intended to make it easier to search reports. Banks also may tailor the reports to, say, macro traders or stock buyers by adding or subtracting components they find most valuable such as charts or models, according to a bank executive.

“Over time, there is a prospect of premium prices,” Frost’s Scarth said.

That seems totally fine and reasonable to me. You have built a valuable thing, people want it, you should be able to sell it. Some people want it more, or are able to pay more, so you should be able to charge them more. Maybe you include some bonus features. More detailed researched reports, say, or live Excel models, or invitations to meetings with company managements, or one-on-one chats with analysts, or early access to the reports before everyone else gets them.

Wait, I feel like I remember something about that? Oh right:

Washington, D.C., April 12, 2012 — The Securities and Exchange Commission today charged that Goldman, Sachs & Co. lacked adequate policies and procedures to address the risk that during weekly “huddles,” the firm’s analysts could share material, nonpublic information about upcoming research changes. Huddles were a practice where Goldman’s stock research analysts met to provide their best trading ideas to firm traders and later passed them on to a select group of top clients.

Goldman agreed to settle the charges and will pay a $22 million penalty. Goldman also agreed to be censured, to be subject to a cease-and-desist order, and to review and revise its written policies and procedures to correct the deficiencies identified by the SEC. 

The theory behind the "huddle" case was that research is not just a product made by a bank for the private consumption of its clients, which can be sold to those clients through the normal workings of market capitalism. It is also "material, nonpublic information," 6  because the research -- or at least the buy/sell recommendation -- can itself move stocks, so banks have some sort of public responsibility to distribute it fairly. (Whatever that could mean: It was always okay for banks to distribute market-moving research only to customers.) So the Securities and Exchange Commission recently charged an analyst and trader with insider trading because the analyst told the trader about his recommendations before he made them public. And BlackRock got in trouble with New York Attorney General Eric Schneiderman for -- this is true -- asking research analysts for their opinions about stocks, on a scale of 1 to 9. "Brokerage customers expect that the information contained in analyst research reports will not be disseminated to a few select brokerage customers prior to those reports being broadly disseminated to all brokerage customers that are entitled to receive those reports," said Schneiderman, strangely. 7 But there is no particular reason for that expectation. It's just how things have always been done.

If research is a thing of value that can trade in the normal channels of commerce -- you know, selling it for money! -- then some longstanding expectations about it will have to change. If a bank can sell research, why can't it decide not to sell it, and use it for itself? Or to sell it in multiple tiers, in which the people who pay the most get the best and fastest research? A lot of intuitions about investment bank research are built on its historical role in a gift economy, and on the somewhat weird assumption that -- like sex or votes or organs but unlike most things produced within investment banks -- research can be given away freely, but never sold. But if banks get better at measuring the value of research, they'll be better able to put a price on it. And if the research is valuable, it may be hard to justify a price of zero.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.
  1. I made that number up. It was about right in 2009, though the commissions pool has declined since. Three cents does seem to be the dream, though.

  2. That's U.S. equity commissions; the total pool is larger. "Money managers cut the amount they spend on commissions by about 26 percent after the 2008 financial crisis to $22.7 billion last year," says Bloomberg

  3. So Brad Hintz says:

    This is simply another chapter in the unending quest for revenue clarity in institutional equity trading. When an institutional client executes a trade in equities and generates commission revenue, everyone on the floor claims his or her share. The equity block desk demands its portion, since it has provided liquidity to the client. The derivatives desk says "It’s mine"-- and points out the structured trade it recently made for the client. The sales force points out the numerous sporting events and lavish dinners that they provided the client. "We took ’em to the U.S. Open." And even the quants ask for a cut, arguing that "their" algos or "their" dark pool are the true reasons the client is even trading with the firm.

    Only the poor equity research staff has nothing to prove that the commission dollar belongs to them. 

    But that last sentence isn't true! You can't prove that the U.S. Open tickets, or even the algos, brought in the revenue, any more than you can prove that the research did. It's all fuzzy attribution. And the trend to "unbundling" in equity trading, with the idea that clients pay for research and algorithms and execution and so forth directly rather than squashing everything into a commission, puts pressure on it everywhere. 

  4. Or that research analysts have occasionally tried to add value in ways that are not, strictly speaking, allowed. If you are not getting love from the equity trading franchise, why not try making yourself indispensable to the underwriting franchise?

  5. People who work in journalism may recognize the tradeoff between revenue, measurement and user experience that banks are pondering here.

  6. In the huddle case, Goldman was accused of violating Section 15(g) of the Exchange Act, which requires that "Every registered broker or dealer shall establish, maintain, and enforce written policies and procedures reasonably designed, taking into consideration the nature of such broker’s or dealer’s business, to prevent the misuse in violation of this chapter, or the rules or regulations thereunder, of material, nonpublic information by such broker or dealer or any person associated with such broker or dealer."

  7. I wrote a long footnote (footnote 1 here) at the time interrogating that sentence. Felix Salmon was also pretty skeptical of the Schneiderman/BlackRock settlement.

To contact the author of this story:
Matt Levine at

To contact the editor responsible for this story:
Zara Kessler at

Before it's here, it's on the Bloomberg Terminal.