Advisory firms assess multi-stage private equity strategies on three dimensions:
- Realized return evidence, weighted toward DPI rather than TVPI or IRR.
- Manager diligence into how the manager decides on follow-on investments and values its own portfolio companies.
- Portfolio fit, meaning the strategy's role, duration, and sizing within the private markets sleeve.
Each dimension requires documentation that a generalist fund review does not produce.
What is a multi-stage private equity strategy, and why does it need its own assessment framework?
Why it matters: a blended track record hides the two separate skills a multi-stage manager is being paid for.
A multi-stage private equity strategy invests across multiple phases of a company's life cycle within a single program. The typical pattern is an initial investment in an early institutional round, followed by additional capital in later rounds as the company scales. For this article, the term refers to strategies that span venture capital and growth equity, the most common approach among managers who describe themselves as multi-stage.
A multi-stage strategy relies on two distinct return engines:
- Early-stage selection: identifying companies before the market has priced their potential.
- Later-stage capital allocation: deciding which existing holdings deserve more capital, at what price, and in what size.
These skills are related but not interchangeable. A manager can excel at one while being mediocre at the other.
A standard fund review evaluates a manager's aggregate track record. For a multi-stage strategy, that aggregate answers the wrong question. A strong blended multiple can reflect a handful of early-stage outcomes paired with weak follow-on decisions, or disciplined follow-on capital compensating for an average early-stage hit rate. The assessment must separate the engines before it can judge either one.
The risk runs in the opposite direction as well. Multi-stage construction concentrates more capital in fewer companies over time. That concentration can amplify good selection and equally amplify a manager's attachment to its own portfolio.
Which return metrics matter most when evaluating a multi-stage manager?
Why it matters: the metric a manager leads with often reveals which part of the story it would rather not tell.
Distributions to paid-in capital (DPI) is the first metric advisory firms should examine because it is the only standard multiple that measures cash returned. Carta's Q2 2026 VC Fund Performance report found median DPI of 0.37x for the 2017 vintage, 0.15x for the 2018 vintage, and 0.04x for the 2019 vintage, even as those funds approach a decade since inception. For any manager whose recent funds fall into those vintages, a DPI near these medians reflects the market environment as much as manager skill. Comparisons should therefore be made within vintage, never across vintages. The same shortfall appears in buyout, where five-year rolling DPI reached its lowest recorded level in 2025, as BIP Capital's analysis of degraded diligence signals shows.
Total value to paid-in capital (TVPI) and net IRR now tell markedly different stories for late-2010s funds. Carta reported that the 90th-percentile IRR for 2017-vintage funds fell to 25.7% as of Q2 2026, down from 28.7% two years earlier, while the 90th-percentile TVPI for the same vintage rose to 4.14x from 3.31x. Carta attributes the divergence to IRR's sensitivity to time. TVPI does not penalize a long hold; IRR does. Much of that value remains unrealized, and the timeline to realize it appears longer than many investors expected.
TVPI for a young fund has limited predictive value. Carta's analysis of roughly 2,900 U.S. venture funds found that TVPI becomes highly correlated with ultimate DPI only after six or seven years, possibly longer in the current environment, a window consistent with the six-year minimum cited in BIP Capital's manager selection analysis. A manager presenting a strong TVPI for a fund younger than that is presenting an estimate, not an outcome.
Subscription facilities distort the IRR of any fund that uses them, and advisory firms should request figures both with and without the facility. The Institutional Limited Partners Association's Performance Template standardizes reporting of IRR and TVPI, with designated breakouts for gross and net figures, both with and without the impact of fund-level subscription facilities. ILPA's guidance applies to funds launched on or after Q1 2026. It states that presenting fund-level net figures that include the facility alongside portfolio-level gross figures that exclude it violates the SEC Marketing Rule. A manager's willingness to report in the ILPA format is itself a diligence data point.
A public market equivalent (PME) comparison completes the picture by asking whether the same cash flows, invested in a public index, would have produced a similar outcome. PME matters most for growth-stage exposure, where valuations track public technology multiples more closely than early-stage valuations.
For a multi-stage strategy, advisory firms should request four decompositions that aggregate metrics do not reveal:
- By entry stage: returns are split by the stage at which each company first entered the portfolio.
- By capital type: returns on initial investments are separated from returns on follow-on investments.
- By concentration: the share of total fund value attributable to the three largest positions.
- By loss ratio: the share of invested capital held at a multiple below 1.0x, calculated separately for initial and follow-on dollars, using the definition in BIP Capital's manager selection analysis.
Together, these reveal which engine produced the track record and how dependent that record is on a small number of outcomes.
What due diligence criteria separate one multi-stage manager from another?
Why it matters: the follow-on decision is where a multi-stage manager's incentives and investors' incentives are most likely to diverge.
Track-record attribution and persistence
Track-record attribution identifies which individuals made which decisions and whether they remain at the firm. Research on persistence gives this step greater weight in venture capital than in buyout. Harris, Jenkinson, Kaplan, and Stucke, in the version published in the Journal of Corporate Finance in 2023, used Burgiss cash-flow data and found little evidence of buyout persistence when limited to information available at the time of fundraising, but continued persistence for venture capital funds. The authors' 2020 working paper reported that venture persistence declined for post-2000 funds and attributed part of the gap to interim performance at fundraising being only moderately correlated with results.
Final-return data tells a stronger story than fundraising-time data, and the distinction matters when evaluating persistence claims. BIP Capital's analysis of CalSTRS and CalPERS commitments found a 0.47 correlation between consecutive fund IRRs across 230 fund pairs, measured using mature, largely final returns. Harris and coauthors also find persistence in final returns; the gap appears only when an allocator must rely on interim figures available when a successor fund is raised. For a multi-stage manager, the early-stage record is therefore the more informative evidence, with one caveat: a track record built during a low-rate, rising-multiple period reflects conditions that may not return. The interim marks on the most recent fund are the least reliable.
Follow-on reserve governance
Reserve policy determines how much of a fund is reserved for follow-on investments and how that capital is deployed. Advisory firms should document three things:
- The target reserve ratio and how it has compared with actual reserve use in prior funds.
- Who decides on follow-ons, and whether that decision requires separate investment committee approval?
- Pricing discipline, including whether the manager has declined follow-ons in its own portfolio companies and at what rate.
A manager who follows through in nearly every company is not exercising selection at the later stage. It is averaging down its conviction.
Valuation of insider-led rounds
An insider-led round occurs when an existing investor, often the manager itself, sets the price for a new financing round. The resulting price becomes the mark for every earlier investment in that company. Advisory firms should ask:
- What share of follow-on dollars went into rounds led by an independent outside investor?
- How are insider-led rounds valued for reporting?
- Does a third-party valuation provider or the fund's auditor review those marks?
This question bears directly on the reliability of the TVPI figures discussed above. The position-level look-through questions in BIP Capital's valuation diligence analysis apply here as well.
Cross-vehicle allocation
Some multi-stage managers deploy later-stage capital through vehicles separate from the fund that made the initial investment. That structure can give investors targeted exposure to maturing companies. It also places investors in different vehicles on potentially different sides of the price. Advisory firms should review:
- The manager's written allocation policy.
- Whether an advisory committee or independent party reviews cross-vehicle transactions.
- How the arrangement is disclosed in the manager's Form ADV Part 2A.
Fees, structure, and operational infrastructure
Fee analysis should account for every layer of cost an investor bears, including management fees, carried interest or incentive allocations, fund expenses, and any fees charged at the follow-on vehicle level. The terms governing carried interest and the sequencing of distributions are covered in BIP Capital's analyses of carried interest, hurdle terms, and distribution waterfalls. Operational diligence should confirm the identities of the fund administrator, auditor, and valuation provider. It should also confirm that reporting can flow into the advisory firm's performance reporting system without manual rekeying.
How does a multi-stage strategy fit within a private markets allocation?
Why it matters: a multi-stage strategy is the longest-duration, lowest-distribution component most advisory firms will hold, and sizing errors surface late.
Within a private markets sleeve, multi-stage venture and growth equity typically serve as the capital appreciation component. They produce no current income and carry the longest expected duration. Pairing them with private credit or other income-oriented strategies can offset part of that profile at the sleeve level. The pairing does not change the equity position’s liquidity characteristics.
Duration has lengthened, and planning assumptions should reflect that. Carta has observed that venture-backed companies remain private longer, extending the expected lifespan of many venture vehicles. The median DPI figures cited above, for funds approaching their tenth year, are the practical consequence. Advisory firms should model client liquidity assuming capital committed to a drawdown fund will not return in meaningful amounts for most of a decade, and should set client expectations for the early-period J-curve accordingly.
Sizing should be based on median outcomes rather than upper-percentile outcomes. Carta's Q1 2026 report found that for every vintage from 2017 through 2024, except 2021, the 90th-percentile net IRR exceeded 20%, while no vintage in that range had a 75th-percentile IRR above 15.5%. The distribution is skewed: a small minority of funds produce the results that shape industry expectations. J.P. Morgan Asset Management's Guide to Alternatives similarly identifies venture capital and private equity as asset classes with pronounced manager dispersion, and BIP Capital's manager selection analysis compiles quartile and decile spreads from several providers. An allocation appropriate only if the manager lands in the top decile is not appropriately sized.
Vintage diversification and pacing reduce the risk of overcommitting at a single valuation peak. For drawdown funds, achieving a target exposure requires a commitment schedule spread over several years. Evergreen vehicles provide immediate exposure but inherit the vintage mix of their existing portfolio. That mix should be reviewed before subscribing.
Overlap is a less visible risk. A multi-stage strategy's growth-stage holdings may be correlated with public growth-equity exposure elsewhere in the client portfolio, and quarterly private marks can lag public repricing, a measurement issue examined in BIP Capital's analysis of smoothed private-market marks. Advisory firms should also check for company-level overlap when a client holds more than one venture or growth manager.
Structural choices carry suitability implications. Drawdown funds require investors to meet capital calls over several years. Semi-liquid vehicles differ: interval funds commit in advance to periodic repurchase offers under Rule 23c-3, while tender offer funds and non-traded business development companies repurchase at their boards' discretion. In both cases, offers are capped and prorated when requests exceed the cap, as BIP Capital's guide to interval and evergreen structures documents. Funds that charge performance-based fees are limited to qualified clients. Under the SEC's inflation adjustment, effective June 29, 2026, that threshold is $1.4 million in assets under management with the adviser or $2.7 million in net worth.
What does a documented assessment look like in practice?
Why it matters: the documentation, not the conclusion, is what an examiner or investment committee reviews.
A defensible assessment of a multi-stage strategy proceeds in six sequential steps.
- Define the role. The advisory firm records the role the strategy is expected to play in the sleeve, the target allocation range, and the client segments for which it is suitable before reviewing any manager materials.
- Establish the return evidence. The firm calculates DPI, TVPI, and net IRR for each prior fund, compares them within each vintage, and requests ILPA-format figures with and without subscription facility effects.
- Decompose the track record. The firm requests returns by entry stage, by initial versus follow-on capital, and by concentration in the largest positions, and notes which engine produced the result.
- Test governance. The firm documents its reserve policy, follow-on decision rights, insider-round valuation practices, and any cross-vehicle allocation policy, with the manager's Form ADV Part 2A disclosures attached.
- Model the fit. The firm runs sizing and liquidity scenarios under median-outcome assumptions, records overlap with existing holdings, and confirms that the vehicle structure aligns with client eligibility and liquidity needs.
- Record the decision and the dissent. The firm documents the conclusion, the principal risks identified, and any unresolved questions, along with a review date tied to the manager's next fundraise or annual report.
What are the limits of this assessment framework?
Why it matters: every benchmark cited in a manager review carries its own sample bias.
Benchmark data for venture and growth funds is incomplete, and each provider's sample is skewed. Carta's Q1 2026 report drew on 2,775 funds, about 89% of which were smaller than $100 million. Figures from Cambridge Associates, Burgiss, or PitchBook will differ because the underlying samples differ. Figures from different providers, reported on different dates, are not directly comparable.
The persistence research is the most rigorous available, but its cash-flow data predate the 2021 vintage cycle. Its findings may not carry forward unchanged. Unrealized valuations for any fund younger than about seven years remain estimates, and the framework above reduces reliance on them rather than eliminating it.
Glossary
Additional definitions appear in the BIP Capital venture capital glossary.
DPI (distributions to paid-in capital): Cumulative cash distributed to investors divided by paid-in capital. A DPI of 1.0x means investors have received their paid-in capital back.
TVPI (total value to paid-in capital): Cumulative distributions plus remaining net asset value, divided by paid-in capital. TVPI includes unrealized value.
RVPI (residual value to paid-in capital): the remaining net asset value divided by paid-in capital. TVPI equals DPI plus RVPI.
Net IRR: The annualized, time-weighted return to investors after management fees, carried interest, and fund expenses are deducted.
PME (public market equivalent): A method that compares a private fund's cash flows with the returns those same cash flows would have earned in a public index.
Subscription facility: A credit line a fund uses to bridge capital calls. It can shorten the measured holding period and increase reported IRR.
Reserves: Capital a fund sets aside at inception for follow-on investments in its existing portfolio companies.
Insider-led round: A financing round priced and led by one or more existing investors rather than by an independent new investor.
Vintage year: The year a fund begins deploying capital, used to group funds for performance comparison.