Conference run-up study
By year: the bubble, and the hangover
| Year | n | Median D-30 → D-1 |
|---|---|---|
| 2017 | 30 | +2.46% |
| 2018 | 36 | +2.41% |
| 2019 | 66 | +3.25% |
| 2020 ← bubble | 47 | +17.27% |
| 2021 | 71 | +2.19% |
| 2022 | 167 | −3.33% |
| 2023 | 357 | −2.86% |
| 2024 | 371 | −1.74% |
| 2025 | 177 | +1.22% |
| 2026 | 103 | +4.51% |
The whole path, and the fade after the event
Readers always ask the obvious next question: what if I held through the presentation? Historically, you gave it back.
| Window | n | Median | 25th pct | 75th pct | Mean |
|---|---|---|---|---|---|
| D-30 → D-1 | 1425 | −0.03% | −10.41% | +13.58% | +5.53% |
| D-20 → D-1 | 1425 | +0.35% | −8.11% | +11.52% | +4.07% |
| D-10 → D-1 | 1425 | +0.63% | −5.15% | +7.32% | +3.13% |
| D-5 → D-1 | 1425 | −0.06% | −3.63% | +4.64% | +1.14% |
| Event day | 1425 | −0.56% | −4.85% | +2.43% | −0.50% |
| D-1 → D+5 | 1425 | −1.59% | −9.03% | +3.20% | −1.72% |
| D-1 → D+10 | 1425 | −1.93% | −10.95% | +4.34% | −2.27% |
The move on the presentation day itself is −0.56%, and it keeps drifting down: −1.59% over the next five trading days and −1.93% over ten.
Why the average lies
The mean D-30 run-up is +5.53%. The median is -0.03%. That gap is the entire story, and it is why you should distrust any single conference “run-up number”; including a mean we could have quoted to flatter ourselves.
| Of all 1,425 presentations… | |
|---|---|
| were positive at all | 49.8% |
| ran up 25% or more | 15.7% |
| ran up 50% or more | 6.2% |
| fell 25% or more | 8.6% |
| standard deviation | 33.5% |
Barely half of presenters went up at all, a coin flip. A small tail of huge winners (6.2% ran 50%+) drags the mean 5.6 points above the median. The average is real; it is just not what happens to you.
By market-cap tier
| Cap tier | n | Median D-30 → D-1 | 25th pct | 75th pct |
|---|---|---|---|---|
| Nano | 108 | −7.86% | −17.70% | +14.86% |
| Micro | 260 | −1.95% | −14.76% | +17.86% |
| Small | 298 | +2.75% | −12.38% | +22.86% |
| Mid | 116 | +3.45% | −7.58% | +21.11% |
| Large | 323 | −0.19% | −4.88% | +6.02% |
Market capitalisation at the date of each presentation (point-in-time, not today) is available for 1,105 of the 1,425 presentations. The remaining 320 have no cap tier and are excluded from this table, so these rows sum to 1,105, not 1,425. Tier medians are computed only over the presentations where the tier is known.
Nano-caps are the worst cohort, not the best: a median of −7.86% (n=108). This inverts the folk wisdom that the smallest names run hardest. Cap tier is each company’s market cap at the time of that presentation, not today’s (see corrections). Tier resolves for 1,234 of 1,425 events; the rest are excluded from this table rather than guessed.
By conference
| Conference | n | Median D-30 → D-1 | 25th pct | 75th pct |
|---|---|---|---|---|
| ASCO | 276 | +1.42% | −9.05% | +12.44% |
| ASH | 201 | +3.13% | −7.12% | +20.03% |
| ESMO | 192 | −3.33% | −14.21% | +4.84% |
| AACR | 129 | +2.53% | −7.44% | +18.57% |
| EHA | 86 | +3.07% | −4.28% | +18.41% |
| SITC | 65 | −2.27% | −13.34% | +16.33% |
| AASLD | 41 | −2.22% | −12.37% | +13.03% |
| EASL | 34 | −2.68% | −9.24% | +3.51% |
| AAN | 28 | −3.54% | −10.38% | +13.20% |
| SABCS | 28 | +5.42% | −10.72% | +27.88% |
| AHA | 23 | −0.36% | −15.16% | +12.82% |
| ENA | 23 | +0.16% | −18.38% | +27.34% |
| ASCO-GI | 21 | +5.17% | −3.71% | +28.56% |
Only conferences with n ≥ 20 are shown. We previously published rows as thin as n=13; we have cut them. 24 events whose conference label we could not verify are excluded from this table; they are real presentations with real prices, so they stay in the headline sample, but we will not put them under a conference name we cannot stand behind.
Corrections
We now compute market cap as of the day of each presentation, from SEC-reported shares outstanding × the split-unadjusted close. (Splits matter: this universe contains 208 reverse-splits, and multiplying split-adjusted prices by as-reported share counts overstates historical caps by the split factor.) 17% of events changed tier. The headline finding survives; nano is still the worst cohort, but the magnitude moved from −9.84% (n=42) to −7.86% (n=108), and two tiers changed sign: micro-caps went from +2.14% on the hindsight tiers to −1.95% on point-in-time tiers.
“ANE” is a garbled letter-order of ENA: the AACR-NCI-EORTC International Conference on Molecular Targets and Cancer Therapeutics, a genuine annual meeting held each October. What actually happened is smaller and duller than “fake conference”: the meeting was real, its code was mangled, and two of its six stored dates were corrupt (a March date and a June date on an October conference), which dragged unrelated presentations into the bucket.
So we have split it honestly. 23 events whose dates fall in the October window are genuine ENA and are restored to the table under the right name. 24 events whose labels we cannot verify are now marked unknown: they keep their prices and stay in the headline sample, but they no longer sit under any conference heading. The corrupt dates are deleted rather than guessed, so those years fall back to month precision.
The lesson we are keeping: “this data looks like garbage” is a hypothesis, not a finding. We deleted first and checked second, and we owe you the check.
An SEC-sourced crawler cannot see the whole field. A conference poster is material to a $200M biotech: it files an 8-K. It is immaterial to Merck, no filing exists. Measured recall against the real AACR 2026 presenter list: 57% for companies under $2B, but only 38% for companies over $2B. Roche, Merck and Jazz were simply invisible to it.
So adding those events did not enlarge the sample evenly: it re-weighted it. Large-cap share of the study fell from 29.2% to 25.0%, and nano/micro rose to fill the gap. Nano is the widest-dispersion cohort, so the nano median duly “moved” from −7.86% to −10.95%. That was not a finding. That was the sampling frame changing underneath the number.
We have reverted to the curated 1,425-event sample. The crawler stays, but it does not feed this study until it is multi-source (filings + press releases + the ASCO/AACR abstract databases, which are the only ground truth) and we can publish a measured recall figure per conference and cap tier.
The rule we broke, now written down: never change the sampling frame and the published number in the same step. If the universe changes, publish old and new side by side with the recall figure, and say why they differ.
Method, and its limits
Anchor. The last daily close before the conference start date. Windows. D-30 means 30 trading days (not calendar days). Universe. 1,425 presentations by US-listed biotechs at 45 major medical conferences, 2017-2026, 383 distinct tickers. Statistics. We lead with the median and the interquartile range, and show the mean only to demonstrate how misleading it is. We never report a “win rate.”
2. Single anchor. For many conferences the abstract or late-breaker title drops weeks before the meeting, and that drop is itself a price event. Our windows are anchored on the conference start date only, so a move that happened on abstract release is partly captured and partly missed. A dual-anchor version is the correct way to do this, and we would rather tell you it is missing than quietly publish a number that pretends otherwise.
3. Regime, not law. The year-by-year table is the clearest warning on this page: the same event type produced +17% in one year and −3% in another. Whatever you take from this study, do not take it as a constant.
“pdufa.bio, Biotech conference run-up study (2017-2026), n=1,425 presentations, https://www.pdufa.bio/research/conference-runup”
Not affiliated with or endorsed by the FDA or any medical society. pdufa.bio is an independent publication with no affiliation with, endorsement by, sponsorship by, or connection to the U.S. Food and Drug Administration or any other government agency. “FDA”, “PDUFA” and all company, drug and ticker names are used descriptively and remain the property of their respective owners.
Informational and educational purposes only. Not investment advice. Nothing on this page is investment, legal, tax or medical advice, or an offer or solicitation to buy or sell any security. pdufa.bio is not a registered investment adviser or broker-dealer and does not recommend trades or publish individual-drug approval probabilities. Verify every date and outcome against primary FDA, SEC or company filings. Data is provided as is, without warranty of any kind, and past behaviour does not predict future outcomes.
© 2026 pdufa.bio. All rights reserved.