About a decade ago, when Ryan Shea worked as the head of research at a sovereign wealth fund, buy-side representatives regularly flocked to meetings with a particular strategist at a big bank.
This had nothing to do with the strategist being a star, Shea recalls: "It was because he was terrible at predicting markets." Schadenfreude explained a good part of the buy-siders' interest; meetings with the analyst were something like the fixed-income world's answer to the dîner de cons. But there was also valuable information to be gleaned from the hapless strategist's views, however misguided they were. "We flipped him," Shea says. "We always took the other side of the trade. His research was terrible, but it was very valuable because we could invert his calls."
What is the value of an idea? The answer depends, of course, on the idea, and the context in which it is set. Write a hot take for a national media outlet, and you can expect to make about $200; patent a pioneering piece of scientific research or build a social network that traffics in human vanity, and you could be looking at billions.
In the financial industry, evaluating ideas has long been done through the lens of the trading desk: A good idea is one that makes money. But many factors influence the chain of idea formation — general news, market color, the prevailing political context, traditional equities research, the behavior of stupid money. Figuring out how to assign a value to all this information produced away from the trading desk has been trickier, or ignored altogether. That's mostly because investment research has been either produced in-house, done by the buy-side, or given away by the sell-side for free.
That is changing.
On January 3 the world of investment research entered a new era. Brokers and banks used to subsume the cost of research into their execution fees; in practice this meant they charged nothing for it, treating it as a marketing bauble to attract trading. A new European Union law — the Markets in Financial Instruments Directive, or MiFID II — is seen as a global harbinger and has, since the beginning of the year, required sell-side institutions to explicitly charge for research instead of rolling it into brokerage costs. Where once there was no priced market for research, suddenly there is one. A desire for more-transparent and fairer markets drove the EU to implement MiFID II. The aim is to institute a system in which the buy-side places trades with brokers on the basis of best execution, rather than in exchange for freebies and inducements like research and corporate access.
But despite this noble aspiration, uncertainty has been the predictable result. Sell-side institutions are scrambling to figure out the right price for their content — and which asset classes and sectors it makes the most sense for them to cover. Buy-side firms, long drowning in a never-ending stream of (often ignored) sell-side analysis, are facing the question of how much outside information — and what kind of information - they really need to generate market-beating returns. And into the confusion has rushed an opportunistic new host of independent analysts and tech-savvy content aggregators, buoyed by the belief that finance is about to undergo a disruption every bit as meaningful and industry-redefining as those seen in recent decades in music, media, and television.
With the U.K.'s Financial Conduct Authority giving firms some compliance leeway to get their post-MiFID II houses in order, most of the market participants Institutional Investor spoke to for this story feel that it will not be until the middle of the year that the true shape of change materializes. Yet three things seem certain. First, independent analysts put the size of the global investment research industry, once it is broken out, at anywhere from $16 billion to $20 billion per year. Second, when the effects of MiFID II are felt, the buy-side's annual spend on research will likely be much smaller than that, although the market for research will remain inefficiently priced for months, perhaps years, to come. And third, amid the inefficiency — of both pricing and coverage — pockets of opportunity will open up for investors in a way that did not exist under the old regime.
Will the financial content industry Netflixify or ossify?
The volume of content the sell-side produces is staggering. J.P. Morgan Chase & Co. says it published 131,000 research reports in 2017, including research notes, emails, model updates, and client alerts; other big banks and brokerage houses report output at similar levels. As a point of comparison, The New York Times publishes about 80,000 pieces of content — stories, graphics, interactives, blog posts — per year.
But what most distinguishes investment research, especially in equities, is how little it is read.
Only 2 to 5 percent of research emails are opened by clients, according to industry insiders. Buy-side firms have no incentive to pay attention when they're receiving information for free, but analysts do little to help their own cause. Research reports — usually delivered in 1999's hottest new technology, the PDF file — often resemble each other. There's little thought given to presentation, style, or, to adopt the Silicon Valley lingo, "user experience." Duplication of coverage across investment houses is rampant. According to Bloomberg, there are 45 analysts publishing research on Apple alone.
And analysts don't just cover the same things; they often say the same things. Herd thinking is rife, a monotony of perspective that's arguably unavoidable when you consider that analysts usually come from similar educational backgrounds:
Ph.D. economists become fixed-income analysts, and those with a corporate finance bent find work in equities research. Analysts often share an outlook, an approach to modeling forecasts — the bread-and-butter of all analytical work — and a flat-footedness when attempting to analyze market-moving events outside their immediate areas of expertise. (Try consuming the views of a neoclassically trained economist on politics; the results are rarely pretty.)
Until MiFID II came along, these analysts also shared a readership. That is already beginning to change. "Sell-side research is completely inefficient," says a senior portfolio manager, who wishes to remain anonymous, at an asset management firm with more than $150 billion under management. "There's too much of it, and most of it is no good." Before MiFID II, he adds, "we didn't really care about research — we didn't track it, we didn't know whether it was useful. It was just there. Now that fund managers have to pay for it themselves, they're saying, 'Actually, no, fuck it — it's not worth it.'"
Most buy-side firms have announced they will absorb the cost of sell-side research themselves rather than pass it on to their customers. Fidelity, one of the few holdouts, said in late February that it would reverse course and do the same. Last November the CFA Institute, a body of investment professionals, released a survey in which buy-side participants said they expected to pay between 3.5 and 20 basis points for research — €350,000 to €2 million for every billion euros in assets managed, in other words.
Now that MiFID II has come into effect, it appears many fund managers overestimated their own generosity. The senior portfolio manager, who works for a desk that invests across asset classes, says his group will spend barely more than $500,000 on external research; the fixed-income desk is doing away with sell-side content altogether. The equities managers in his firm have a greater need for research, he notes, but in all, he estimates the $150 billion firm will spend less than $5 million on research. "I can sense a bit of panic on the sell-side," he adds archly.
Will this reduced research spend lead to poorer information flows and worse investment decisions? Many buy-side houses have indicated they will be beefing up their own research desks as the rivers of content from the sell-side run dry, but the senior PM admits to some uncertainty. "We might be kidding ourselves that we can do the job without sell-side research. But if everything goes pear-shaped, that will create an opportunity for smart new entrants in the space."
The response from the sell-side has been both swift and predictable: a good old-fashioned price war. Content, the old cliché has it, wants to be free — and financial content is proving no different. J.P. Morgan kicked the war off in earnest late last year when it offered clients access to all its written research, across all asset classes, for $10,000 per year; many large banks have followed suit in the months since, and the price-cutting shows no sign of abating. "The big banks' bet is that they'll gain market share and squeeze out smaller competitors, then gradually raise prices once the field has been cleared," says Gerard Walsh, head of business development at Northern Trust Capital Markets, a brokerage firm.
This projected revenue curve — a sharp decline in the immediate aftermath of January's big regulatory bang, followed by a steady climb back to levels that make financial research sustainable, not to say profitable — is familiar from the music and media industries. In the past five years, the dollars from music streaming services have offset the decline in physical-album sales and one-off digital downloads; streaming revenues are expected to double between now and 2020. Thanks to Spotify and its imitators, overall revenue in the global music industry is poised to grow for the first time since Sean Parker and Napster went toe-to-toe with Metallica and Dr. Dre, almost 20 years ago.
But easy analogies with the financial industry should be resisted. Investment research is not music, whatever top-ranked equity analysts might tell you, and change imposed by regulation is different from the type of proactive change wrought by shifts in consumer behavior. The music industry had only one unit of production to price efficiently — the song — but took more than a decade to figure out a viable pricing model. Research is considerably more difficult to price in a consistent way; a one-page equity research note previewing a big public company's quarterly earnings report usually involves far less digging, proprietary knowledge, and analysis than, say, a single page of analysis by a risk arbitrage specialist. The idea that the economics of financial content will be worked out quickly seems like so much sell-side wishful thinking.
Many banks are hoping to make money off research by going low on text-only access but charging top dollar for one-on-one analyst calls and analyst-brokered introductions to company management teams and policymakers. How much money can be wrung from these premium services remains to be seen, and whether sell-side houses will tolerate research desks operating as cost centers in the event premium services fail to take off is equally unclear. More certain is the prospect of pain for the sell-side, especially in the small to medium-size brokerage houses, over the next two years. "If you're the eighth-best broker in Italy, that's tough," says Jeremy Davies, the co-founder of RSRCHXchange, a London-based research marketplace looking to capitalize on changes sweeping the industry. "In eighth place no one is going to return your calls."
Sell-side research teams will be cut or decommissioned altogether; coverage in equities will concentrate on the most liquid, popular stocks; and analysts will be forced to become more original, according to Walsh. At the end of this quarter or the next, adds Vicky Sanders, Davies's co-founder at RSRCHXchange, "the navel-gazing will begin" as the early results roll in from the broker voting system used across the industry to evaluate analysts and as buy-side clients complete their research assessments. Banks and brokerage firms will start to see which of their analysts have the most traction with a newly alert buy-side. Then, once that's clear, the head count cuts will begin.
Until that happens, Sanders predicts, "most sell-side analysts will stay in their well-paid jobs, with the golden handcuffs on." A few brave souls, however, saw change coming and have taken the decision to strike out on their own.
For analysts, going independent is now a far more viable path than it was in the pre-MiFID II era, when they were stuck with the losing pitch of asking the buy-side to pay for the type of content these firms already received, in unmanageable quantities, for free.
Philip Rush was the chief U.K. economist at Nomura for six years before launching his own macro research consulting firm in late 2016. MiFID II was a major factor in the decision: "The approach most banks have taken — charging a blanket subscription for everything — doesn't suit many clients, who just want to pay for the analysts they like," he explains. Going independent allows for that flexibility and gives Rush space to be more creative in the way he generates and presents his insights. "A lot of the analysis you do on the sell-side is a bit throwaway," he says. "It's short-term, superficial stuff, over Bloomberg chat. There's not much value-add."
As an independent, Rush now has the freedom to speak only when he thinks he has something worth saying. The use of agent-based models and so-called "alternative" data from cloud-based services such as Quandl, an independent data provider, and Simudyne, a simulation-building tool, is now a far greater part of his work than when he was employed at a big bank. Compliance and cultural restrictions — not to mention the pressures of client marketing and herd thinking — meant there was limited scope to broaden his work in this manner when he worked on the sell-side. (Recently, a client wanted to know the effect of potential tariffs between the U.K. and the European Union. Rush parsed a database of 9.5 million U.K. revenue and customs metrics against the full World Trade Organization tariff schedule and mapped out the different scenarios in a dynamic, granular simulation. "At the bank I would have just taken the average tariff rate and applied it to the numbers before running off to some marketing event," he notes. "Now I can take the time and the care to be different.")
Many predicted that the start of MiFID II would be a disaster for independent analysts, as banks moved to slash prices and squeeze out smaller, less well-resourced competitors. Rush, who offers his research at a flat rate of 3,000 per user per year, has found the opposite to be true. He's seen more inbound interest since the beginning of the year than at any other point in his 18 months as a sole operator, and his revenues have grown 47 percent since last year's final quarter. Much of this interest has come not only from large asset management firms, but also from sovereign wealth funds, central banks, and — most surprisingly to Rush — the trading desks of banks, which find themselves underserved as their own research groups are forced to court external business.
Other independent analysts, even those who charge far more for their services, report the same thing. Rod Manalo worked for more than a decade as a risk arbitrage analyst before striking out as an independent three years ago. He charges more than $10,000 for a report on a single M&A deal but says the high cost of his research has been no barrier to getting business. "Going independent, you're not a cost center and you can provide independent analysis, free of influence from trading and sales desks," he explains. "That's extremely valuable to the buy-side."
Though they maintain that most of their business continues to come from direct deals with the buy-side, both Manalo and Rush have benefited from the recent blossoming of tech-driven, exchangelike research marketplaces that connect research producers and consumers. Smartkarma and RSRCHXchange, two of the most popular platforms, report a noticeable uptick in interest since MiFID II came into effect. RSRCHXchange has seen its user base grow 30 percent over the past two months, and now has 1,200 buy-side firms and 330 research providers on the platform, according to Sanders. Singapore-based Smartkarma, meanwhile, has about 500 analysts on its platform and has signed up a number of big-name buy-side clients. In late 2016 the platform scored a major coup when Société Générale signed Smartkarma to provide Asian equity research to its clients; in recent months the fintech company has opened offices in the U.K. and the U.S. The platforms differ in their revenue models. RSRCHXchange, like iTunes, sells individual research reports and takes a 20 percent commission on any sales analysts make. The model Smartkarma uses — charging a flat fee of $7,500 per user per year, then pocketing most of that revenue and dividing the remainder among individual analysts based on how much client interaction they generate — is closer to that of Spotify or Netflix. The best analysts can expect to make about $10,000 a month on Smartkarma, but according to Jon Foster, the company's chairman, "the pool that we're paying out to analysts is increasing 11 percent month-on-month. If you're a competent analyst, you would now be making the kind of money a midlevel analyst would be paid in a bulge-bracket investment bank."
The growing popularity of these platforms comes amid a broader surge of interest in alternative data and the insights provided by traditional expert networks, such as Gerson Lehrman Group, Third Bridge, and AlphaSights, and their upstart, tech-driven competitors, such as GlobalWonks. Small, innovative fintech shops also think the winds of change are in their favor. Simudyne, the London-based start-up whose tools Rush incorporates into his research, evaluates the performance of companies and sectors in dynamic, artificial intelligence-driven simulations — a far cry from the static, linear models most equity analysts use to generate their projections today. "Any companies that can offer an entirely different insight into the valuation of companies, or the performance of sectors at a macro level, will have an edge over the competition," says Justin Lyon, Simudyne's CEO.
Fintech and quirky data sets are not the whole solution, of course. As MiFID II begins to bite in earnest, the nature of what counts as interesting insights, or what types of people should be providing them, will also change.
Any fund manager can describe in general terms the characteristics of information that adds value to an investment process: It's anything that's original, challenges one's thinking, and either deepens or attenuates conviction about a thesis. The notion that traditional research analysts are the people best placed to provide this information is increasingly coming into question. An astute observer on the ground in Venezuela might have deeper insight into the Venezuelan debt crisis than an economist sitting high up in a tower on Madison Avenue; a plastics expert could provide sharper color on an individual materials company than a stretched equity analyst attempting to cover 20 names at once.
"Old-school, press-the-flesh sales and information-gathering skills might come back into vogue," notes Northern Trust's Walsh. "It's not necessarily all about tech." But tech — specifically, platforms such as Smartkarma — can play an important role in connecting insight providers previously locked out of financial analysis to investors. Foster says Smartkarma receives four to five applications from aspiring analysts every day but accepts just three per month onto the platform; he argues that the background and profile of the industry's best analysts could change over time, as the buy-side's need to justify research expenditure increases the premium on analytical originality and difference. "You'll need to stand out more and more," he asserts.
The financial industry today is stuck between two worlds; the old regime is giving way, but the Netflixified future of investment research, in which content is made available as a streaming service and the universe of insight production is opened up to a whole new category of analysts, has not yet arrived.
The only constant amid all of this is the enduring appetite of investors for new information that can give them an edge. But new and valuable information does not have to come in the guise of a trading recommendation or a well-turned research report. It can take many different forms, from an account of social media activity during an election campaign to the useful idiocy of the always-wrong analyst whose views are an accurate proxy for what dumb money wants to do. Figuring out how to value these wildly contrasting units of information will be difficult and will take years, and might help explain why a slick, on-demand, one-price-fits-all future for research will be even slower to arrive in finance than it has been in other content industries — if it arrives at all. "Right now we know the price of everything and the value of nothing," argues Walsh. The senior portfolio manager at the $150 billion asset management firm agrees. "We still don't really have a good mechanism for figuring out value. Until we develop one, it'll be like the old test for pornography," he concludes, referring to Justice Potter Stewart's dictum in the 1964 U.S. Supreme Court case Jacobellis v. Ohio: "We know it when we see it."