Apparently FINRA is looking into whether sell-side research analysts are doing some naughty things, which is an evergreen topic, though you might almost imagine that the current round is being prompted by the return to public life of Eliot Spitzer, who gets a quote in the DealBook article. Eliot Spitzer: not a fan of research-analyst naughtiness.
It's hard to tell if the analysts are doing naughty things but, probably, right? Basically the analysts are meeting with potential issuers before those issuers' IPOs, which is fine. But at those meetings, which tend to be arranged by "so-called I.P.O. advisers" like Solebury, Rothschild or Lazard,1 they might be talking about the IPO and the analysts' views of the issuers, which is not fine. They're "supposed to discuss only broad industry trends at these meetings" and defer to their bankers for "specific views on a company, like earnings models or potential I.P.O. pricing," because the idea is that the analyst meetings are not supposed to be used by the issuers to select underwriters. They're just a chat! It's like, hey, I'm in this industry, you cover this industry, let's talk about broad industry trends! For my general education! Because while, yes, me and my IPO advisers sitting next to me are picking underwriters for our IPO, and while your bankers are pitching us "within hours" of this meeting, right now I'm entirely focused on a general chat about broad industry trends. That's all this is.
Etc. etc., which is sort of absurd but it also feels like the absurdity is in the rules as well as the evasions. It's weird to care so much about this stuff in a world where - well, here, look, issuers can meet with credit rating agencies all they want. When people get exercised about credit rating agencies it's not about meetings and subtle hints. It's that the agencies are paid one hundred percent by issuers and zero percent by anyone else, so they have some fairly straightforward incentives to do what the issuers want.
That seems less true with sell-side research? The analysts here work for investment banks, and the investment bank's whole job in an IPO is to mediate a conflict of interest between two sets of clients. It has a client - the issuer - who wants to sell its stock for a lot of money. And it has some clients - the investors - who want to buy that stock for not a lot of money. And the bank's job is to cajole those two sides to a place where there's a transaction. The way it does this by having two groups of employees, each of which is assigned to one set of clients, and whose job is (1) to cajole its client(s) and (2) to advocate on behalf of its client(s) to the other group. So you have bankers whose job is to manage the issuer's expectations and to convince the salespeople that the deal is worth selling at a high price, and you have salespeople whose job is to drum up investor enthusiasm and to convince the bankers to give the investors a break on price.
And the way you motivate that is also pretty straightforward: the bankers get paid for bringing in banking - issuer-side - business. The salespeople get paid for bringing in trading - investor-side - business. On any given trade there might be a powerful temptation to blow up a client, but in the long run you need repeat customers and so you have good reason to advocate for them
Research analysts have a weird job because they don't have a client. Which means everyone thinks they're the client. Issuers want to hire a bank whose analysts will say nice things about them, and genuinely don't understand that that's sort of illegal. (Because: of course you'd want that!) Investors want to do business with a bank whose analysts provide them valuable research. And the regulators want analysts to just express their own personal opinion, divorced from any sense of responsibility to issuer or investor clients.2
The right answer is what everyone knows about the rating agencies, and what the banks know about motivating their bankers and salespeople, and what Eliot Spitzer et al. knew when they did the research settlement: you get what you pay for. If you want sell-side research analysts to feel beholden to their investor clients, have them paid based zero percent on banking revenues and one hundred percent on ... investor-client stuff, trade generation and client surveys and whatnot. That's what the settlements provide. It's not entirely what happens - coverage of banking clients does seem to correlate with pay - but it is mostly what happens; good research seems to get rewarded more than bad but issuer-favoring research.3
Which seems about right? The conflict that banks have in IPOs is unresolvable: some clients want a high price and an easy path to a public listing, others want a low price and a lot of scrutiny, and making one side perfectly happy tends to involve upsetting the other. Recognizing that seems more sensible than a pretense of objectivity, or of undivided loyalty to one set of clients.
1.DealBook adds that "banks rarely send compliance officers to monitor the discussions," which is sort of a fascinating aside. Like, on the one hand, the idea of sending a compliance officer to a client meeting is basically crazy, right? It would never happen with a banker or salesperson except in the most horrifying of circumstances; probably salespeople are taking compliance officers to their SAC meetings these days. But on the other hand: yeah, the research settlement does involve a lot of compliance chaperoning of all research interactions with anyone. It is a tough life, really, research, these days; you're basically not allowed to talk to anyone.
2.Which is sort of absurd on its face: what good does an honest personal opinion do anyone? "Jenkins, you gave that company a Sell rating, which not only cost us the lead-left role on the deal, but also cost our investor clients billions of dollars when the stock quintupled in the first month." "Well, it was just, like, my opinion, man." "Clean out your desk."
3.Though I continue to be fond of the theory that just sucking up to a company to get access might be better for your investor clients than giving the company a "correct" rating or price target. If investors are mainly using you for access, not your reports, then making everyone a Conviction Buy and getting your clients in the room with them might be the best service you can provide.