If someone has a job that involves stocks and their business card says “analyst,” this probably means one of two things:

1. They work at an investment fund of some sort, like a hedge fund or family office, where they do research that contributes to determining what goes into a portfolio. These are buy-side analysts.

2. They work at a company that works with those funds, like a broker or independent research firm. They get paid not based on whether or not they make the right calls about stocks, but based on how well they contribute to the work of people in job #1. These analysts are on the sell-side.

In this piece, we'll focus on the second category, because it's a somewhat unusual economic niche that is the result of many decades of path dependence, spanning market structure (i.e. how do investment firms pay for the research they get?), career paths, and regulations. It's a durable niche, though; one piece of evidence is that there are people who have been publishing research for so long that the economic model that got them paid has been made illegal—twice!

For sell-side analysts, there's a rough inverse correlation between a) how visible their work is and b) how much their clients value it. The most visible thing that sell-side analysts do is publish research, and that research generally fits into one of three categories:

• Initiations are exactly what they sound like: the analyst is publishing their first report on a company, or on an entire industry. Initiation reports are actually pretty handy as an industry primer; the best ones are basically "everything you'd want to know after reading the prospectus or most recent annual report, plus a few things you hadn't necessarily thought of." Initiations are also a good place for analysts to introduce frameworks for understanding and valuing a business: is the big story growth, or is it margin? Within growth, what's the mix between adding new customers and growing revenue from existing ones? Is there more excitement for the company's existing products, or for new ones? Analysts put these together by doing industry research, reading regulatory filings, and talking to management at the company they’re going to cover and its competitors. These reports will include estimates with varying levels of specificity, and generally a price target based on whatever valuation rubric the analyst thinks makes the most sense.

• Analysts will generally publish a preview ahead of earnings or other catalysts (which could be industry conferences, product launches, regulatory approvals, or something else), and a quick update after the catalyst hits. The preview is partly a chance to update estimates based on what's happened since the last report came out. This might be for fundamental reasons for example, an airline analyst might raise earnings estimates because fuel prices have come down. And sometimes it was because the company wanted to be able to say it "beat" expectations, so it asked analysts to lower them at the last minute ($, WSJ).¹ This, inevitably, is priced in; it's been a long time since there was easy money in looking for a post-quarter estimate revision from whichever analyst had the closest relationship to company management and then betting that every other analyst would make the same move.

• In between catalysts, analysts sometimes produce reports where they update their evaluation of the company—maybe the growth story from their initiation isn't playing out, or maybe it is but there are new concerns about margins. These reports are sometimes a surprise, and can move the market, especially if an analyst switches from a "sell" to a "buy" rating or vice-versa, or if they revise their price target so it's the highest or lowest of all the analysts covering a stock.

But those reports aren’t the key way in which sell-side analysts provide value! Of course, the reports still do something, and can certainly move the market, but buy-side analysts are also updating estimates intra-quarter, setting their own price targets, trying to interpret catalysts, etc. And the better someone is at figuring out whether a company will beat or miss estimates, or whether a given beat/miss represents a significant change in the company's long-term outlook, the more money they'll make if they're at a hedge fund. In other words, if you're trying to maximize the value of your research reports, you’re better off writing them for an audience of one, where the “one” is a fund that collects a performance-based fee for their management of the fund (because they’ll pay you a bonus directly out of that fee).

The key way sell-side analysts provide value comes from a menu of other services they provide. Among these: when they're putting together estimates and price targets, they're building a detailed financial model with historical data, their view on what drives the business, and their estimate of how those drivers work out. These analysts will typically share their models with buy-side clients on demand, which serves two purposes: first, it saves them a lot of data entry. And second, it's a way to see how other investors are thinking about the business, and to figure out how some unusual event would change that assessment. (Recently, that first purpose has been taken over by companies that focus on it—Diff advertiser Daloopa uses AI to generate and update models (and they offer readers a free sample)).

Another service on the sell-side analyst menu is mediating access to company management. Company managers attend bank-sponsored conferences, where they do large-group presentations, small-group meetings, and one-on-one meetings with investors. Access to these meetings is scarce by design: meeting with management is useful, but it's lucrative if it's also partly a Veblen good. That access gets mediated in two ways: first, the analyst has to impress management with some combination of their understanding of the business and their sympathy for management's point of view. It's not impossible for a bank whose analyst has a "hold" rating on a stock to be able to arrange for that company's CEO to spend the day talking to the bank's clients, but it's a lot easier if the analyst has a "buy" rating on a stock.

What's the purpose of these meetings? Matt Levine has written extensively about this: it's partly the unobjectionable process of shareholders talking to the company managers they've entrusted the business to, and partly an effort by those same shareholders to earn a little extra return from having slightly better information than their counterparts. Company management is more interested in talking to big, long-term investors, but "long term" means that those investors don't trade as often or as aggressively as the hedge funds, and thus generate lower fees for the sell-side analyst’s company. So the sell-side analyst has to balance between the meetings that management teams want and the meetings that pay the bills.

So how exactly do those bills get paid? Long ago, the compensation model was, more or less, that analysts were paid based on how much underwriting business the banks did, because that underwriting was partly a function of analyst coverage. So a famous, well-respected sell-side analyst like Mary Meeker or Henry Blodget could get their company lots of underwriting business via IPOs in the 90s because the analyst was likely to rate the stocks their employer took public later on. This model worked reasonably well when everyone understood how it worked, but part of what made it so lucrative in the 90s was an influx of retail investor interest in the market, and retail investors by and large didn't understand what the banks' incentives were.

Ultimately the banks ended up paying giant fines, and agreed that equity researchers would no longer be compensated based on IPO underwriting revenue after all. Instead, they get paid in a somewhat roundabout way: banks charge more for trades than the market would support (if they were purely charging for trade execution), and the amount of commissions that a firm generates determines what services it will get. And in Europe, this model, too, has been mostly banned, and firms have to pay for research and trade execution separately instead of as a bundle. There is a little wiggle-room here: a helpful buy-side analyst at a small firm will probably get better services than someone who generates slightly more revenue and is a pain to deal with. But that's the general idea. (And finance is an iterated game; helpful people who think in positive-sum terms tend to end up doing well over time, so they don't remain small clients forever.)

And that concept of the helpful buy-side analyst brings up yet another function of sell-side analysts, which is one of the valuable services they provide and which explains why their research reports matter even if there are better ways to monetize stock-picking ability: part of a sell-side analyst's job is to articulate different flavors of the current buy-side consensus. A stock price only exists when a buyer thinks it's too cheap and a seller thinks it's too expensive, so there's always at least a bull case and a bear case. Within this, there can be fractal levels of detail: one bull case might be that this small growth company will end up beating a larger incumbent, while a separate bull case might be that the same company will succeed in selling itself to the incumbent. One cohort of bears might be concerned that growth is slowing down, while another might be equally worried that management will destroy value by overpaying for growth.

So buy-side analysts talk to sell-side analysts partly because other buy-side analysts do; research quality ends up being a Schelling Point for arranging even higher-quality discussions with opinionated, well-informed people on either side of the argument. So the end goal of a sell-side analyst publishing a good report is not just to get opinions out there, but to solicit other opinions, synthesize them, and hopefully get to the right answer.²

Earlier in this piece, we talked about how sell-side analysts’ reports can move the market, even if the broad assumption is that these analysts are not as good at predicting stock price movements as the analysts at funds that trade the stocks are. And this is a mechanism for it: publishing a report outlining a bull or bear case for a stock can set off a preference cascade, where people on the buy side know that other people think some new variable or trend matters for the stock. Part of what hedge fund analysts do, especially at high-turnover funds, is figure out “what the company trades on,” i.e. what data points move the stock. When someone on the sell-side starts highlighting such a data point, it’s evidence to the buy-side that this metric affects sentiment and is worth tracking.

Markets thrive at a controlled level of information asymmetry: if everyone agrees on everything, there's no reason to trade. If nobody agrees on anything, the spread between buyers and sellers is too high and there's no liquidity. The function of sell-side analysts is to smooth the flow of information, to articulate what the relevant views are, and to connect people, directly or indirectly, with other people who will change their minds.

Read More in The Diff

The Diff has covered the business of analyzing stocks a few times. (Your author has worked on the buy- and sell-side, albeit in the latter case not at a company that did any management meetings.)

• You And Your Investment Research looks at the high-impact questions a researcher can answer.

• Optimal Research Allocation ($) addresses the individual- and firm-level incentives for doing different kinds of research.

• The Alternative Data Primer ($) and Part Two of the same ($) focusses on one narrow piece of investment research, finding unusual datasets and building predictions with them. This is an area where the boundary between what the sell-side does and what the buy-side is best at remains porous, and it will be interesting to see how the long-term economics shake out.

A Word From Our Sponsors

I’m proud to announce this week’s newsletter is sponsored by Compound.

Compound is personalized wealth management for people in tech. It’s how people at companies like Stripe, OpenAI, and Discord make smarter decisions about their money.

I actually know Jordan, Compound’s CEO, and wrote a piece for their site a while back about equity and options. The company is built and run by people who deeply understand the financial decisions you’re likely thinking about — and who are equipped to help you make those decisions.

Compound focuses on your finances and taxes, so you can focus on what matters most.

1. This is akin to the political ritual where, ahead of presidential debates, a candidate's people will leak all sorts of information about how their candidate is unprepared, doesn't handle debates well, has been busy with more important stuff, etc. All while furiously prepping and rehearsing lines.

2. This has more than a little in common with the newsletter business. Though a Discord is a lot cheaper to run than a conference.

Help us learn more about you!

We're building a media kit and need some anonymous aggregated subscriber data. If you work in finance and have time for a 3-question survey, that would be very helpful:

Join the discussion!