Data and methods
How to read the research.
Good research should let you follow the reasoning. This page explains the samples, definitions and measures behind our first two white papers, along with the limits to keep in mind when using the findings.
Colibrí Institute, 2026 • Moncada, 2026
2,142 U.S.-based venture capital funds, 2000–2024.
Definition: what makes a manager emerging
Our definition focuses on characteristics associated with barriers to institutional capital. It builds on NexTier’s Why Not the Best? framework and includes four routes into the emerging category. A fund qualifies if it meets any one of the criteria below.
- Scale
- The firm oversees less than $100 million in total assets under management across all funds.
- Sequence
- The fund is Fund I, II, or III, regardless of the prior investing experience of the general partners.
- Organizational age
- The management firm has a vintage of less than five years, reflecting organizational youth rather than fund age alone.
- Access
- The founding team includes women or otherwise underrepresented general partners. This criterion recognizes a dimension of access to institutional capital; it does not imply a uniform performance effect.
Both papers use this main definition and examine alternatives. Read each result with the definition and sample specified in that paper.
Source
Fund-level performance data sourced from PitchBook and analyzed as part of doctoral research (Moncada, 2026).
Version
White Paper I studies performance in 2,471 funds. White Paper II studies risk efficiency in 2,142 funds. These are distinct analytical samples, not interchangeable versions of one dataset. The number of observations also varies by metric and model.
Reuse
Definitions, measures and exhibits are licensed under CC BY 4.0, so you may reuse them, including in commercial work, with credit to the Institute. The PitchBook fund data behind the papers is not published or licensed.
Read the licenseThe database
Our VC database is a separate source from these papers. It is built from SEC filings and public statements and uses no licensed vendor data.
Read the database methodologyThe Fund Benchmark
Fundraising and structure from SEC filings and net returns as limited partners publish them, by fund cohort, free and under the same license. It uses no manager-reported return and no licensed data. What it publishes, and when a figure is withheld, is set out in the database methodology.
Read how the benchmark is builtThe measures
Two ways to put fund performance in context.
White Paper II uses reported fund data from PitchBook. Each measure compares a fund with others raised in the same year, known as its vintage. That makes age comparisons more meaningful, but it does not remove the uncertainty in unrealized values or substitute for cash-flow analysis.
Return per unit of risk
A fund’s internal rate of return (IRR) divided by the standard deviation of IRRs in its vintage cohort. In plain language, it compares annualized return with how widely returns varied among the funds raised that year. The denominator describes the cohort, not the fund’s own risk.
The average score is 0.65 among the 1,773 funds with the required data.
Upside capture
A fund’s total value to paid-in capital (TVPI) divided by the median TVPI of its vintage cohort. It compares the fund’s cash returned plus remaining reported value with the middle fund raised that year.
A score of 1.00 matches the vintage’s median TVPI. A score of 1.40 is 40% above it. This measure is available for 1,916 funds.
Illustration · reading two measures together
| Measure | Fund A | Fund B |
|---|---|---|
| Net IRR | 13.9% | 13.6% |
| TVPI | 3.25x | 2.00x |
| Return per unit of risk | 0.64 | 0.62 |
| Upside capture | 2.48 | 1.53 |
These hypothetical funds have similar annualized returns but different total-value multiples. Reading both measures gives a fuller picture. The illustration does not provide the underlying cash-flow timing or establish either fund’s probability of loss.
Limitations
Stated here and repeated in the body of each paper beside the finding they qualify.
- Association, not cause
- The design compares funds at a point in time. Every result is an association between manager characteristics and performance.
- Coverage is not neutral
- Commercial databases do not capture every fund, and available observations differ across metrics and models. These limits may affect the results; the analysis does not establish whether the reported differences are understated or overstated.
- Fund level, not deal level
- The study uses reported fund-level performance. It does not use the complete cash-flow and deal-level information needed to measure every dimension of risk.
- Uneven generalization
- Results vary across stage groups. The later-stage sample is small, so its findings need particular care.
- U.S. only
- Both papers cover U.S.-based funds. Findings should not be assumed to apply to other markets.
Corrections policy
We publish our errata. When a figure, a citation, or a unit is wrong, we correct it, date the correction, and leave the record of what changed. Corrections that affect a headline finding are announced to anyone on the research list.
Corrections to date
Website clarification, September 18, 2026: We clarified that White Papers I and II use distinct samples; explained the constructed risk-efficiency measures and their limits; corrected the website’s IRR unit and dollar interpretation; and narrowed language about portfolio choices, screening and subgroup results. The underlying papers and data have not been changed.
Questions about method
hello@colibri.institute