A peculiar natural experiment has been running in equity markets for two decades, and its results should unsettle every CFO of a small or mid-sized public company. When a brokerage firm closes or merges and its analysts stop covering a stock — for reasons that have nothing to do with the company itself — the affected firm subsequently cuts investment by roughly 1.9 percent of total assets and external financing by 2.0 percent, its cost of debt rises by about 25 basis points, and its probability of a credit event roughly doubles. Research published in the Journal of Finance and the Journal of Corporate Finance established these numbers using broker disappearances as a clean source of exogenous variation, which means the damage is caused by the loss of information production, not by anything the company did. The finding also explains a corporate behavior documented in follow-up studies, where firms that lose coverage compensate by becoming their own information source, pushing more frequent and more forward-looking disclosures through investor relations channels and pr distribution services rather than waiting for a sell-side desk to interpret them. This article assembles the evidence on what analyst coverage is actually worth, why the supply of it collapsed where it was needed most, and what the substitution toward company-produced disclosure looks like in practice.
Coverage Is Capital, Not Commentary
The intuition behind the numbers is information asymmetry. An equity analyst who builds a model, questions management quarterly, and publishes estimates is manufacturing public information about a firm. When that production stops, uncertainty about the firm’s value rises, and investors demand compensation for bearing it — through wider bid-ask spreads, lower prices, and higher required returns on both equity and debt.
François Derrien and Ambrus Kecskés turned this intuition into causal evidence. The methodological problem they solved is worth understanding, because it is what separates their result from correlation-hunting. A company that loses an analyst usually loses it for a reason — deteriorating fundamentals, shrinking trading volume — so comparing covered and uncovered firms directly tells you nothing. Broker closures and broker mergers break that circularity, because when a brokerage house shuts down, every stock its analysts covered loses coverage simultaneously, regardless of each company’s condition. Tracking those firms against matched controls produced the investment and financing declines cited above, with the effects concentrated exactly where theory predicts, namely in smaller firms, firms with thin existing coverage, and firms dependent on external financing. A companion study on debt markets found that the same quasi-random analyst disappearances raised borrowing costs by roughly 25 basis points and made defaults and other credit events 100 to 150 percent more likely relative to control firms. One departing analyst, in other words, is not a public-relations inconvenience. It is a financing shock.
Six Analysts Versus Thirty
The second piece of the picture is how unevenly the analyst workforce is distributed. Asset-manager data puts the average large-cap U.S. stock at around 30 covering analysts, the average mid-cap at 17, and the average small-cap at about six — and averages flatter the bottom of the market, where a substantial share of micro-cap companies trade with one analyst or none. The economics are straightforward. Sell-side research is funded, directly or indirectly, by trading and banking revenue, and a company with modest float and thin volume cannot generate enough of either to pay for a professional’s full-time attention.
This asymmetry inverts the usual relationship between information and need. The marginal thirtieth analyst on a mega-cap adds almost nothing to what the market already knows; the first and only analyst on a $400 million industrial firm may be the sole professional translating its filings into estimates anyone reads. Yet it is precisely that first analyst whose seat is economically fragile. The Derrien–Kecskés results confirm the asymmetry from the damage side, since the firms hurt most by a coverage loss are the ones that had the least coverage to begin with.
What Unbundling Did to the Supply Side
In January 2018, the European Union’s MiFID II directive forced asset managers to pay for research explicitly instead of folding it into trading commissions. The stated goal was transparency; the practical effect was to make every research subscription a visible line item that a fund’s management had to justify. Budgets contracted, and the industry braced for small-company coverage to be the casualty.
The empirical record turned out to be messier and more interesting. Several academic studies found the post-2018 decline in analyst numbers was concentrated in large caps, where the marginal analyst was redundant, while the UK’s Financial Conduct Authority reported limited evidence of small-cap coverage loss. France’s securities regulator, the AMF, concluded the opposite in 2020, arguing that unbundling had aggravated an already long-running erosion of small-company research. ESMA’s own study split the difference with a finding that should worry issuers more than any single trend line, observing that small and mid-sized companies remained structurally characterized by less research, lower research quality, and a persistently higher probability of losing coverage altogether — a condition unbundling neither caused nor cured. The European Commission found the situation troubling enough to partially reverse course, exempting research on companies below €1 billion in market capitalization from the unbundling rules. A regulator does not roll back its own flagship reform for a segment that is thriving. Whatever the precise attribution, the direction is not disputed by anyone: professional information production about smaller public companies has been declining for years, and no regulatory adjustment has restarted it.
The Substitution Toward Self-Produced Disclosure
The most practically useful finding in this literature is about what companies do next. Research on MiFID II’s aftermath by Lang, Pinto, and Sul documented that firms losing sell-side coverage responded by upgrading their own disclosure — publishing more often, adding forward-looking content, and making announcements informative enough to move prices without an analyst intermediating. This is the rational response to the arithmetic above. If the market prices what it can verify, and the verification workforce has left, the issuer must replace the lost information flow itself or accept the valuation discount.
Replacing an analyst’s function is a different discipline from ordinary corporate communication, because the audience is portfolio managers and screening algorithms rather than journalists. The academic evidence and disclosure practice point to a specific set of components that a self-produced information program needs in order to substitute for lost coverage:
- Estimate-ready quantitative detail — segment revenue, margins, and guidance presented in a structure an investor can drop into a model, since the absent analyst is no longer doing that translation.
- A predictable disclosure calendar — fixed announcement rhythms reduce perceived uncertainty on their own, because irregular silence from an uncovered company is read as bad news.
- Broad, simultaneous dissemination — announcements must reach terminals, databases, and screening tools at once, both to satisfy fair-disclosure rules and because an uncovered stock has no analyst note to carry the message afterward.
- Forward-looking framing — the studies show price reactions attach to statements about the future, which is exactly the content sell-side research used to supply and filings do not.
- Machine readability — a growing share of small-cap information intake is quantitative screening, so numbers buried in prose or image-based tables are effectively undisclosed.
None of this restores a human being who argues a stock’s case to clients, and companies below a certain size increasingly pay for sponsored research to fill that final gap. But the disclosure channel is the part the issuer fully controls, and the evidence says the market rewards firms that use it deliberately.
The research on analyst disappearances measures something companies rarely see itemized, which is the price of being unexamined — lower investment, costlier debt, and a doubled default hazard traced to nothing more than the loss of one professional’s attention. Since the supply of that attention keeps shrinking at the small end of the market and regulators have not reversed the trend, the burden of information production has shifted permanently onto issuers themselves, and the ones who treat disclosure as replacement infrastructure rather than paperwork are the ones the market can still price correctly.
