Market Memos · Where To Look
Biotech Catalysts
A dated calendar, a registry nobody reads, and a financing cycle that eats correct theses.
Biotech breaks the rules that apply everywhere else. A small biotech has no product, no revenue and no customers — so there is nothing to search for, no app to rank, no reviews to count. The entire behavioural layer that works on ordinary companies is simply absent.
What it has instead is something no other sector offers: a public calendar of dated events, and a registry that records changes nobody announces. The work is not finding sentiment. It is knowing what is scheduled and reading the filings more carefully than the people talking about them.
The registry — the real edge
Every trial on ClinicalTrials.gov has a record history: a numbered, date-stamped list of every version of that record since registration. The URL pattern is clinicaltrials.gov/study/<NCT ID>?tab=history, and appending &a=<n> opens a specific version.
Sponsors amend these records quietly. There is no press release. The edits that matter:
| The change | What it means |
|---|---|
| Recruiting → Active, not recruiting | Enrollment is complete. The readout clock has started and you can now count forwards. |
| Primary endpoint reworded | Rarely good news. Changing the measure late usually means the original one wasn't going to be met. |
| Enrollment target cut | Either the trial is struggling to recruit, or the statistics have been revisited. Neither is neutral. |
| Completion date moved | Matters in both directions. Earlier is not automatically good — it can mean a truncated trial. |
| New sites or countries | Enrollment is being accelerated. Sites quietly disappearing is the opposite signal. |
There is a free API at /api/v2/studies with no key required, but it has no version-history endpoint — it returns current values only. To get alerts, either poll it for a watchlist and diff the results yourself, or point a page-change watcher at the history URL.
The calendar
The FDA advisory committee calendar is free and authoritative, though not early — meetings appear only once formally noticed. BioPharmCatalyst carries PDUFA dates, free for the first 150 events in date order. MarketBeat and RTTNews run free alternatives.
The FDA publishes no list of PDUFA dates. Every PDUFA calendar anywhere — all three above included — is a scrape of company press releases and 8-Ks. They carry stale and wrong entries as a matter of course. Confirm any date against the company's own filing before acting on it.
Conference abstract titles appear in the online planner before the abstracts are released, and a title reading "met primary endpoint" moves a stock weeks early. But the biggest data is deliberately routed into late-breaking slots precisely so it stays embargoed until presentation. Trawling titles finds the mid-tier news, not the blockbuster.
Designations — one real database, three that don't exist
Orphan Drug has a genuine searchable FDA database with Excel export. A designation letter can predate the company's press release, which makes it the one designation that can actually be front-run.
Fast Track, Breakthrough Therapy and Priority Review have no FDA database at all. That absence is exactly why they are tradeable press-release events — the company controls when the news appears. Friends of Cancer Research maintains a free database of publicly announced Breakthrough designations with CSV download, and the gap it exposes is instructive: far more have been granted than announced. Companies disclose selectively.
The underrated ones
- FDA warning letters — free, filterable, Excel export, posted within days. The one nobody watches: a warning letter to a contract manufacturer can sink an approval for a small company that is never named in the letter. Form 483 inspection observations come earlier still, though posting is lagged and incomplete.
- EMA CHMP monthly agendas — the European regulator publishes a pre-meeting agenda naming the products up for opinion, before the opinion exists. Free, monthly, and largely ignored by US retail investors holding US-listed companies with European filings.
- EDGAR Atom feeds — live filings by company or form type, free. For a small biotech the news is the 8-K, and it hits EDGAR and the wires before any journalist writes it up.
The financing cycle — this is the one that costs money
A clinical-stage company has no revenue. It runs on cash it raised, and it will run out. That single fact drives more of what happens to the share price than most of the science does.
Good news is when a company can raise money. Which is precisely why good news is so often followed, within hours, by an offering priced into the spike — frequently at a discount, frequently with warrants attached.
| Check | Where | What it tells you |
|---|---|---|
| Effective S-3 shelf | EDGAR, form type S-3 | Permission to issue, already in place. Normal, and worth knowing. |
| At-the-market agreement | Full-text search for at-the-market | Continuous selling with no announcement until the next quarterly. |
| 424B5 filed | EDGAR | Not a warning — the offering itself. |
| Cash runway | 10-Q, going-concern language | Stated in plain English. Compare it to the date of the next catalyst. |
Check for a live shelf before buying a catalyst, not after. And if the cash runway is shorter than the time to the readout, the financing is coming whatever the data says.
Five expensive traps
- Dilution timed to the catalyst. Covered above, and it is the most common way a correct thesis loses money.
- The binary. A Phase 3 miss in a single-asset company routinely takes 70–90% off in one print. Position size is the only defence; conviction is not a defence.
- Options are priced for it. Implied volatility is enormous into a known readout and collapses the moment it passes. A correct directional call can still lose money. Buying premium into a scheduled binary is usually paying full price for the move you predicted.
- Press releases engineered to omit statistics. "Encouraging trend." "Clinically meaningful" without "statistically significant." No p-value, no confidence interval, no control-arm number. If the primary endpoint result isn't stated plainly in the first two paragraphs, assume it missed.
- Reverse splits. Nasdaq closed the reset loophole: fail the $1 minimum bid after a reverse split inside the prior year and there is no compliance period at all. A second split inside a year is close to terminal, not a fresh start.
And one that is a misunderstanding rather than a trap: an approval is not a revenue event. Launch, reimbursement and payer coverage take quarters, and a company that is approved but cannot fund its launch ends up doing another financing.
Filed transactions — insider buying, activist stakes, institutional flows — are on Money Flows. Everything else is indexed on Where To Look.
