In defence of investment research

Fixing the secular decline of the sell-side research business

The problem with investment research is that it lies in the no-man’s land between being a white glove service and something mass-produced like widgets in the factory.

Marty Chavez once said that at its height in 2000, Goldman Sachs employed 600 US cash equity traders and that two decades later there were almost none as the business was done with algorithms. By comparison, sell-side investment research has changed very little.

At the same time, the sell-side research business is more commoditized than investment banking. How differentiated and value-added are deal makers is a moot point but the CEO’s and Boardrooms that they advise demand a white-glove service. Plus advisory fees that corporate clients pay their Wall Street bankers don’t show up in their day-to-day operating costs but are lost further down the P&L next to the integration costs of the deal.

Sell-side investment research is something, which many more people value than those who will admit to it. But very few people are willing to pay substantial amounts for it. I still remember trying to explain to family and friends in the late nineties that our equity research clients didn’t pay us directly but via using our trading desk. Even more difficult to explain or justify was the way that we were paid by the investment bankers who advised the companies we wrote good or bad things about.

I wrote about Why equity research fails over and over (and isn’t coming back) in FT Alphaville exactly two years ago… “there are four reasons why it is still shriveling in size and credibility: declining information advantage, new competitors, margin pressure in secondary cash equities and a shrinking client base.”

Despite buoyant stock market levels and an equity trading boom since then, things have played out as I expected. Looking over the long run sell-side investment research headcount has fallen 35% over the last decade with many large banks making cuts through last year. HSBC joined the bandwagon a few weeks ago with a massive reduction in their team and stock coverage. Independent brokers like Redburn have struggled in recent years with declining revenues and continued losses.

The defenders of sell-side research say that it doesn’t need to be a revenue source. It is a shop window for a large bank – a good marketing tool. Others say it comes in the bundle where clients pay for other services including execution, financing, and deal advisory.

But here’s the thing. The collapse in the number and proportion of stock investors willing to pay for sell-side investment research is further eroding the quality and profile of the profession. Moreover, watching the even greater dependence of sell-side investment research cost bases on funding by other conflicted areas of banks gives me a sense of dot com era Déjà vu.

What’s wrong with the investment bankers covering an increased proportion of the cost base of sell-side investment research? Beyond the obvious conflict of interest, it is notable that 2025 year-to-date IPOs and equity fundraising have been 13% of all investment banking revenues. Most revenues come from DCM, M&A, and loan syndication. The only investment banking business line that needs equity research is ECM, and the IPO business is under pressure not just cyclically but structurally with fewer companies coming public. I was head of corporate strategy at a UK-listed company where we paid tens of millions of dollars to investment bankers for M&A deals. I worked on several large billion-dollar deals for many years and our M&A advisors never covered us from an equity research point of view although ten other investment banks did. The growing market share of M&A boutiques illustrates just this as virtually all of these firms don’t have equity research departments.

If the investment banking division can’t justify the size and cost of investment research departments, then who exactly can? Prime brokerage and trading overlap with the same customer base. Large banks like to be seen as thought leaders. But even their banks are struggling to differentiate themselves.

Whether it is to get paid for their services like most other businesses or to add to the bundle like Teams did for Microsoft or Bloomberg News does for the Bloomberg terminal, sell-side research departments need to dig far deeper moats. Thinking about moats reminds me of Peter Thiel’s four barriers to entry in Zero to One: branding, network effects, economies of scale, and proprietary technology. Could sell-side investment research departments harness these better?

Someone very senior at a major bank said to me the other day “There is no FOMO effect with clients for our investment research.” I call this the Mike Mayo effect. Love him or loath him, back in the day before research analysts were encouraged to speak to the media, there was one analyst who mastered the art of leveraging the media to get his message out and protect his franchise and that was Mike Mayo, the US bank’s analyst who was fired several times after run-ins with investment bankers, bank management teams and the research civil service but always managed to bounce back more strongly.

The value of a social media platform like Facebook has always increased exponentially as the network has grown. Platforms also benefit from having a combination of celebrities and well-respected people mixing with the ordinary public. The role of a research analyst is half analyst and half salesperson. There are distinct network effects in terms of your ability to generate ideas based on a wide range of feedback from clients and the role of information gatekeeper. Ever since I started in the industry in the late nineties but more prominently when the hedge fund industry grew, the buy side has wanted to reverse-broke their ideas and stock picks to the sell side. Of course, they want to do it to a top-rated analyst (analysts are rated in Institutional Investor Magazine surveys) at a top investment bank, who back in the day could move stock prices with their recommendations. But within all this noise there were also valuable signals.

The challenge is that these network effects that the sell-side researcher had are being eroded. The sell-side investment researcher no longer sees or speaks to the whole investor universe.

The growth of the retail investor base appears structural, and they are no longer the dumb money at the back of the train. The US has always had an active retail investor base but that increased after the Covid-19 lockdowns, and the recent Trump tariff turmoil led to even greater retail investing. Historically for cultural and compliance reasons sell-side investment research largely ignored creating content or being relevant for the retail investing universe but they need to find a way to use new content delivery methods such as videos and social media channels to connect with this universe in a better but compliance friendly way.

With passive funds making up such a large percentage of the stock market, sell-side investment research can’t ignore pockets of the active investing community. Post MIFIDII the number of large, long only fund managers that don’t use the sell side has kept increasing. Hedge funds have always been big payers given they need a more regular flow of information. They also need to appreciate investor sentiment. For instance, a company could hit consensus earnings expectations when it reports earnings results but have its share price trade down significantly if the whisper number was much higher. A look at share price action leading into results can give an investor some sense of expectations but consensus from market data vendors can’t give the context or the build-up of expectation that the sell-side investment analyst can.

The challenge for investment researchers is that not only do they have fewer long-only clients than ever before, but the hedge fund industry has also fundamentally changed in the last decade. The number and capital behind more long-term equity long/short hedge funds have shrunk substantially and those remaining are doing even more of their research independent of the sell side using expert networks and other sources. They have cut back on their interactions with sell-side research.

Hedge funds 10-15 years ago were around 30-40% of major broker-dealer’s investment research commissions. Today they are the vast majority of research commissions. Within this, there is an increased concentration amongst a smaller number of hedge funds – mostly pod shops with a much shorter investment horizon and often focused on quarterly earnings.

This erosion of network effects has to be reversed. Fishing in the same pool of pod shops may seem like a healthy strategy near-term because of the proliferation of pod shops and the number of pods, for each sector/industry of stocks, within the large pod shops. But the buy side speaks a lot more to each other than a decade or two ago, especially to investors in the same or similar space. The value of the sell-side analyst’s network is its breadth and diversity.

Sell-side investment research has always been a scale game. Big banks able to absorb the costs without caring as much about their revenue model were able to pursue a more scorched-earth pricing strategy. Although there were historically advantages to bringing waterfront coverage across the whole universe of industries and securities to a large, long-only asset manager, hedge funds in particular are willing to pay for other deep content. With so much public information on companies, companies organizing their own investor relations activities, and CEOs out on social media more than ever, the value of the equity research team is in really deep industry knowledge and connectivity to private companies and industry experts. The stuff you can’t find on the Internet!

The “grey hairs” and not the juniorization we saw over the last decade, is what is necessary to win in most cases. But the Substack and the blogger era has also given us great examples of how it is not merely about the length of time an analyst has followed an industry but the deep and differentiated content. A recent case in point is the success of the young team of researchers at SemiAnalysis – Bridging the gap between the world's most important industry, semiconductors, and business who have deep supply chain knowledge and industry connectivity so much so that tech giant CEOs are quoting them regularly.

The use of AI to revolutionize investment research is garnering a lot of media coverage whether it is the threat from ChatGPT-type platforms replacing the analyst community, the ability of the existing sell-side investment researchers to increase speed and reduce costs by using AI, or the delivery of content as being trialed by UBS with avatars in videos summarizing research reports. Could this allow sell-side analysts to reach out to a broader audience including the retail investor community more effectively?

Much of what investment research teams produce in terms of earnings estimates, price targets, and stock recommendations are in the public domain and the big financial market data vendors clean, combine, and leverage this data in their products endlessly. I recently wrote about a firm called RELX that transitioned from being a sleepy publisher of thousands of legal, academic, and science journals to a must-have data giant where RELX was adding all sorts of analytics and industry-specific GenAI on its proprietary data and content libraries. To end this piece on a half-full tone I do wonder what the opportunity is for major research departments to do more using their content libraries. They may not have proprietary data or industry benchmarks, but they have decades of research reports and financial models on thousands of companies and economic metrics.