Phase 3 Search
DATA REPORT | 2026 EDITION · UPDATED JUNE 2026
A NOTE ABOUT THE AUTHOR

Phase 3 Search is a global life sciences executive search firm.

We place the CEOs, CTOs, and CMC and Quality leadership teams behind biotech companies taking platforms from Series A through clinical readout, BLA, and exit. CMC and Quality leadership is our specialty. Executive search is our craft.

We have made 112+ CMC and Quality leadership placements since 2018, working with biotech founders, boards, and the venture firms backing them.

We are powered by 3 core beliefs:

01
THE PRIME DIRECTIVE
The Prime Directive of biotech, pharma, and all those that support the space is to get medicines to the patients that need them in time to make a difference. This is the timeline that matters.
02
DATA
We believe in data. Evidence beats opinion when deciding about people, capital, and platforms.
03
LEADING THE RIGHT WAY
We believe in leading the right way. The right CTO at the right stage. The right CMC head before the first inspection. Leadership is what turns science into approved medicine.

The report follows.

DATA REPORT | 2026 EDITION · UPDATED JUNE 2026

The Biotech Capital Report 2026

What happens when you analyze 30,486 verified biotech venture deal events across 517 verified equity investors from 2015 to 2026? You get a clear picture of which firms come back for round two — and which write one check and stop. We sorted every investor with a meaningful biotech portfolio on a single metric: their follow-on rate, the percentage of their portfolio companies in which they participate in two or more rounds. Tier 1 sits at 50% or higher. Tier 2 covers 30 to 49%. Tier 3 sits below 30%. Grant agencies excluded. Minimum portfolio: five companies.

90%

of the 517 verified biotech equity investors meeting our minimum portfolio threshold follow on in fewer than half of their portfolio companies. Only 51 — about one in ten — re-invest in 50% or more of what they back.

Phase 3 Capital Intelligence dataset. 30,486 verified deal events across 517 investors meeting the n≥5 threshold, 2015-2026 window. Investor type verified by deterministic classifier with three independent verification gates. Computed by Phase 3 Search.
30,486
Verified Events
517
Investors Ranked
2015–26
Time Window
11 yrs
Observation Period
KEY FINDINGS — 2026 EDITION

What the data shows.

Source: The Biotech Capital Report 2026, Phase 3 Search.

THE FRAMEWORK

Three tiers by follow-on behavior.

Of the 517 verified equity investors meeting the minimum threshold, here is how the universe breaks down.

TIER 1
51
Long-term participants
50% OR HIGHER FOLLOW-ON
Firms whose participation pattern shows multi-round engagement across half or more of their portfolio. Median follow-on rate in this tier: 57%.
TIER 2
162
Selective re-uppers
30–49% FOLLOW-ON RATE
Firms with mixed follow-on behavior. Re-engagement in roughly one in three to one in two portfolio companies. Median: 36%.
TIER 3
304
Single-round participants
UNDER 30% FOLLOW-ON
Firms whose participation is largely a single round. Includes crossover funds, hedge funds, pharma corporate BDs, asset managers, and public-market vehicles. Median: 15%.

METHODOLOGY & CAVEATS

Follow-on rate = percentage of an investor's portfolio companies in which they participated in two or more rounds at distinct dates, computed across the 2015-2026 observation window. Grant-only agencies (NIH and sub-institutes, NSF, US Department of Defense, BARDA, ARPA-H, equivalent international agencies) excluded — grant funding cycles are not comparable to venture follow-on behavior. Minimum portfolio for inclusion: five companies. Sample-size confidence is tagged on every published row: Robust (n≥25), Standard (n=15-24), Adequate (n=10-14), Limited (n=5-9). Of the 517 firms, 157 sit at Robust or Standard, 88 at Adequate, 272 at Limited — about half the universe is in the noisier band, which reflects the long tail of specialist and emerging-manager funds. Tier boundaries: Tier 1 at 50% or higher, Tier 2 from 30 to 49%, Tier 3 below 30%. Follow-on is a slow-burning metric — it requires a multi-year observation window to register. The 11-year window is chosen for that reason. Round coverage is sensitive to source data completeness. Firms with documented multi-round patterns may read lower than their actual behavior where rounds are missing from coverage. The dataset is compiled by Phase 3 Search.

THE PATTERNS

What the data shows when you visualize it.

Four visualizations covering the full 517-firm universe. Aggregate views first; firm-level scatter where sample-size confidence is encoded visually.

Investor types within each tier

Where the money behaves like venture, and where it behaves like crossover or asset management.

Top Tier 1 firms by follow-on rate

All 13 Tier 1 firms at Standard or Robust sample (n≥15).

Portfolio breadth vs follow-on conviction

Every firm in the dataset. Portfolio breadth on the X axis, follow-on rate on the Y. Bubble size scales with portfolio size. Color encodes tier. Limited-sample firms (n=5-9) appear with reduced opacity.

The largest bubbles in the top-right are absent — the law of large numbers caps follow-on rate at scale. The largest portfolios in the dataset (RA Capital at n=188, OrbiMed at n=139, Cormorant at n=111) sit in Tier 2 and Tier 3.

Follow-on rate vs exit rate

Conviction (X) and outcome (Y) are different metrics. Each tier occupies a different region — and the two metrics do not correlate as strongly as a simple "good investor / bad investor" frame would suggest.

Tier 3 firms (red) cluster low-left on follow-on but spread vertically on exit rate. They enter at the exit stage by design. Tier 1 firms (green) cluster mid-to-high on follow-on with mid-range exit rates. Tier 2 firms (amber) sit in between.

THE STRUCTURAL DIFFERENCE

Long-term participants vs single-round capital.

The numbers behind the two ends of the tier list.

BehaviorTier 1 · Long-termTier 3 · Single-round
Firms in tier51304
Median follow-on rate57%15%
Pooled exit rate (acquired + public)37%39%
Typical structuresVenture vehicles, corporate VCs with long mandatesCrossover funds, hedge funds, asset managers, pharma BDs, public-market vehicles, late-stage PE
THE CONVERGENCE

Follow-on rate and exit rate measure different things.

Pooled exit rates across the three tiers sit at 37%, 34%, and 39%. Conviction and outcome are decoupled. A firm with high follow-on does not produce more exits per portfolio company than one with low follow-on; they produce exits through different paths and at different stages. The structural difference is what you take their check for, not whether you take it.

NOTABLE PATTERNS

Five observations from the dataset.

Purely observational — what the numbers show.

01

90% of investors with five or more biotech portfolio companies follow on in fewer than half.

Across the 517 firms meeting the minimum threshold, 466 have follow-on rates below 50%. Only 51 are at 50% or higher.

02

Fourteen firms with 10+ portfolio companies show 0% follow-on across the 11-year window.

These names span asset managers, crossover funds, healthcare systems, government and non-profit vehicles, and several pharma corporate venture arms. The structural commonality is single-shot capital by mandate, not a single sector. Specific firms are available in the Follow-On Explorer companion page, suppressed here pending independent verification.

03

Corporate venture arms re-invest more like VCs than their parent BD desks.

SR One (51%) clears Tier 1; Illumina Ventures (48%) and MRL Ventures Fund (31%) sit in Tier 2 — all re-investing well above their parent-company BD desks. Direct pharma activity from Merck, AbbVie, Sanofi, and others shows materially lower follow-on, reflecting strategic optionality rather than financial commitment.

04

Portfolio breadth and follow-on rate are inversely correlated at scale.

The largest portfolios in the full universe sit in Tier 2 or Tier 3 — RA Capital Management (188 companies, 30% FO), OrbiMed (139, 46%), Cormorant Asset Management (111, 27%). The largest Tier 1 portfolios — ARCH (109) and Atlas (66) — sit right at the Tier 1 boundary. Funds with the broadest deal footprints find it mathematically harder to sustain high follow-on rates; the law of large numbers caps the percentage.

05

The 11-year window matters. Follow-on is a slow-burning metric.

A firm investing in Series A in 2024 has limited opportunity to participate in Series B by 2026. The window must be wide enough for multi-round behavior to register. Shorter windows systematically undercount follow-on, regardless of source-data completeness. This report uses a 2015-2026 window for that reason.

THE FULL LIST

The 517-firm tier list, filterable and searchable.

This page shows the framework, the structural analysis, and the named callouts that meet the report's defensibility standard. The full universe — all 517 firms across all three tiers, with sample-size tags, exit rates, and structural detail — lives in the Follow-On Explorer companion page. Search by name. Filter by tier or type. Click any firm for the underlying numbers.

Open the Follow-On Explorer →
CUSTOM ANALYSIS

Custom Reports.

The tier list above is the standardized view. On request, we produce custom reports sliced by modality (cell therapy, gene therapy, ADC, mRNA, antibody, oligonucleotide, radiopharmaceutical), technology platform (AI drug discovery, CRISPR, base editing, protein degradation), therapeutic area, and stage activity (Seed through Series F lead patterns).

Email [email protected] to request a custom report scoped to your company.

ABOUT PHASE 3 SEARCH

We build the leadership teams behind biotech platforms.

Phase 3 Search is a global life sciences executive search firm. We are not investors. We do not advise on capital. Our craft is search. This report is part of the capital intelligence work we publish for the operator and board community we serve.

RELATED PHASE 3 ASSET

The CTO Mandate Framework.

A 31-page board governance framework for evaluating technical leadership alignment in biopharma. Maps how CTO emphasis shifts across five stages of development. Covers CRL valuation data, board governance gaps, KPIs, and a board-ready calibration tool.

Built with an advisory panel of seventeen CxO-level technical operations leaders.

Download the framework →
ABOUT THE DATA OR ABOUT YOUR SEARCH

If a firm in this report relates to your cap table, or if you have a leadership search coming up, we welcome the conversation.

We can run any firm not in the shown universe, pull deeper analysis on a specific syndicate, or talk about an upcoming CMC, CTO, CEO, or VP-level search. Send us a note.

No spam. One email when we publish the next report.

FREQUENTLY ASKED QUESTIONS

About the data and the methodology.

What is the source dataset?+
A compiled record of 30,486 biotech venture deal events within the 11-year analytical window (2015–2026), drawn from a source dataset of 37,308 events reaching back to 1995. Covers 517 verified equity investors meeting the n≥5 portfolio threshold. Investor type is verified by a deterministic rules-based classifier with three independent verification gates (top-50 manual audit, ground-truth check against twelve known prolific VCs, and zero-FO defensibility review). Compiled and analyzed by Phase 3 Search as part of our executive search practice. Maintained and refreshed by Phase 3 Search.
Why an 11-year window?+
Follow-on rate is a slow-burning metric — it requires multi-year observation to register multi-round investment patterns. A 3- or 5-year window cuts off the back end of follow-on opportunity for recent investments and the front end for legacy ones. The 11-year window (2015-2026) captures the modern biotech VC era and gives the metric the runway it needs to read accurately. Shorter windows systematically undercount.
Why a five-company minimum?+
Follow-on rate is a percentage — with very small portfolios, it becomes statistically noisy. We set the inclusion threshold at five companies to balance two concerns: keeping the list informative by including specialist and emerging-manager funds, and filtering out the highest-noise entries. Smaller portfolios produce less stable rates. Sample-size confidence is tagged on each row.
Why are NIH and grant agencies excluded?+
Grant funding behavior is structurally different from venture follow-on behavior. NIH and its sub-institutes repeatedly re-fund the same companies through SBIR/STTR cycles — that pattern would distort the tier list if treated as comparable to a Series A VC re-investing at the B. We exclude NIH and sub-institutes, NSF, the US Department of Defense, BARDA, ARPA-H, the California Institute for Regenerative Medicine, and similar grant-only sources.
Why does Tier 3 have a comparable pooled exit rate to Tier 1?+
Tier 3 includes crossover funds, asset managers, public-market vehicles, and late-stage PE — capital that enters at or near liquidity by design. They are selecting companies already tracking to acquisition or IPO. Tier 1 firms hold across the lifecycle, including through failures and prolonged development. Follow-on rate measures conviction over time. Exit rate measures outcome at the company level. Neither is "better." They reflect different roles in the capital stack.
Can this report be sliced by modality, therapeutic area, or technology platform?+
Not in this published view. For company-specific analysis — for example, which Tier 1 firms have followed on in cell therapy specifically, or which investors have backed AI drug discovery platforms through Series C — Phase 3 produces custom reports on request. These custom reports slice the same dataset against modality, therapeutic area, technology platform, and stage profile. Email [email protected] to request a custom report scoped to your company.
Does this report connect investor tier to CMC or quality outcomes?+
No. This dataset measures venture behavior — not regulatory or operational outcomes. The Phase 3 CTO Mandate Framework integrates FDA Complete Response Letter data (where 74% of CRLs cite manufacturing or quality deficiencies) with the leadership and governance question. Connecting investor tier to CMC outcome at the company level is a separate analysis we produce on request.
What is different about v2.2 of this report?+
Version 2.2 applies a corrected methodology with stricter investor verification: entity-resolution merges (Frazier, Mirae, Venrock, and others) applied at the canonical layer; an extended 11-year window so the follow-on metric has time to register; updated type taxonomy with crossover, hedge fund, and asset manager classifications properly distinguished from venture capital; sample-size confidence enforced on every published row; and a defensibility gate run against a ground-truth list of prolific company-creation VCs before publication. Named callouts at the tail of the distribution have been suppressed where the data does not survive verification.