TVPI vs DPI: How to Read a Young Fund’s Track Record

A venture capital fund reports a TVPI of 1.8x. Another has already distributed some capital to investors. Which is performing better? The answer depends on how much capital has been paid in, how long it has been invested and what supports the value still held in the portfolio.

The difference between TVPI and DPI is easy to explain and easy to misuse. TVPI includes distributions and residual value; DPI measures distributions relative to paid-in capital. The former contains a valuation component. The latter records what has already been distributed, but does not automatically explain its source, economic merit or repeatability.

For a limited partner, the investor committing capital to the fund, the problem becomes more interesting when the track record is young. Demanding only liquidity can rule out good managers too early. Accepting only valuation markups can finance a story that never translates into returns.

A young fund should be assessed with evidence appropriate to its age, without turning that age into an exemption from scrutiny.

TVPI, DPI and RVPI: three ratios, one denominator

The starting point is paid-in capital: the capital investors have actually contributed under the report’s methodology. It does not necessarily equal committed capital or the amount the manager has invested in startups. Contributions may also fund management fees, expenses and cash still available to deploy.

Metric	Formula	What it measures
DPI — Distributions to Paid-In	Cumulative distributions / paid-in capital	How much has been distributed for each euro contributed.
RVPI — Residual Value to Paid-In	Residual NAV / paid-in capital	How much value remains in the fund for each euro contributed.
TVPI — Total Value to Paid-In	(Cumulative distributions + residual NAV) / paid-in capital	Total distributed and residual value.

Therefore, TVPI = DPI + RVPI, provided the period, scope and calculation conventions match. This relationship also appears in CFA Institute’s GIPS handbook. GIPS, explanation of performance multiples.

NAV, or Net Asset Value, is the residual net value attributable to investors: it includes the fund’s assets, including any cash, less liabilities and relevant accruals. It is not simply the sum of the startups’ valuations.

A TVPI of 1.8x does not mean an investor has received an 80% profit. It means distributions and reported residual value together equal 1.8 times paid-in capital. A DPI of 0.3x indicates distributions equal to 30% of paid-in capital, not a 30% annual return.

Even a DPI of 1x does not mean the investment has concluded at a profit: distributions equal contributions to date, but further capital calls, expenses and capital at risk may remain.

An example: DPI can rise while TVPI falls

Consider a hypothetical fund, with net amounts referring to the same group of LPs: €10 million paid in, €2 million already distributed and €12 million of residual NAV. TVPI is 1.4x, DPI is 0.2x and RVPI is 1.2x.

Now consider two alternative outcomes for a holding carried at €3 million in NAV. In the first, it is sold for €3 million; in the second, for €2 million. In both cases, all proceeds are distributed. To isolate the mechanism, there are no other cash flows, valuation changes or additional costs.

Situation	TVPI	DPI	RVPI
Before the sale	1.4x	0.2x	1.2x
Sale for €3 million and distribution	1.4x	0.5x	0.9x
Sale for €2 million and distribution	1.3x	0.4x	0.9x

In the first case, value converts into distributions without increasing: realisation confirms the reported value. In the second, liquidity arrives, but a loss emerges relative to the previous NAV. More DPI does not necessarily mean more value created.

If the fund instead receives and retains the proceeds, an exit can be realised at portfolio level without immediately increasing the LP’s DPI. The cash remains in NAV. This is another reason why “We completed an exit” and “We distributed proceeds to investors” convey different information.

How young is a young fund?

The year a fund was established is not enough. At least three timelines matter: the fund’s age, the age of its investments and the market conditions in which those investments will have to be sold.

Two funds formed in the same year may have started investing at different times and at different speeds. One may hold a portfolio built four years ago; the other may have acquired many of its positions in the past twelve months. A pre-seed fund and a growth fund also begin at different distances from a potential exit.

In the early years, management fees and expenses can weigh on performance before sufficient gains emerge to offset them. This is one mechanism behind the J-curve: initially weak performance that may improve as the portfolio matures. The shape of a curve, however, does not guarantee that the recovery will happen.

Zero DPI in the second year of a pre-seed fund can be consistent with its strategy. The same figure requires a different explanation if the manager promised rapid realisations or if its portfolio companies are much more mature than the vehicle’s age suggests.

The useful question is: “Which results should already be observable today, given when you invested and what you promised?” These might include verifiable revenue, external financing, industrial progress or initial monetisations. They are not yet distributed returns, but they help test whether the thesis is progressing.

Before comparing: whose return, gross or net?

A multiple on portfolio holdings and a net fund multiple do not describe the same economic experience.

The former may compare the value of companies with the capital the fund invested. The latter looks at flows between the fund and its investors and the NAV attributable to them, accounting for management fees, expenses and carried interest under the methodology adopted. Carried interest is the manager’s share of profits under the fund’s terms.

MOIC, or Multiple on Invested Capital, is often used for investment multiples. It may coincide with TVPI in some reports, but the label does not guarantee identical denominators or scope. It is necessary to read how the number is constructed.

A successful investment at 5x does not automatically make the fund a 5x fund. The other positions, losses, costs and the investment’s weight in total capital still matter. A “net” figure for the whole fund may also differ from an individual LP’s return, for example because of different fee terms or entry dates.

ILPA’s Performance Template distinguishes fund performance from portfolio performance and separates the impact of subscription facilities. It is a useful reference for requesting reconcilable data, not a seal certifying the manager’s quality. ILPA, Performance Template.

The real test of TVPI is the quality of NAV

Calling TVPI “just paper” is a shortcut. Residual value is necessary to assess an investment that has not yet been liquidated. The issue is which evidence supports that estimate and how much it could change.

The December 2025 IPEV guidelines clarify that a recent financing round’s price does not automatically equal fair value at subsequent reporting dates. Events since the transaction and the different rights of share classes must be considered. IPEV, Valuation Guidelines 2025, section 3.10.

For example, a new investor may receive economic protections that an earlier fund does not hold. Multiplying the company’s new valuation by the old ownership percentage may therefore be an incomplete representation of the holding’s value. The prestigious name of the investor leading the round does not resolve the issue either.

An LP should ask how much NAV depends on the three largest positions, how old the market evidence is and which assumptions support valuations of struggling companies. A concentrated portfolio is not automatically weak: in venture capital, a major holding may generate a large share of the outcome. But it is necessary to understand how much of the assessment of the fund depends on that one thesis.

A further check compares net exit proceeds, where available, with the values reported before the transaction had become foreseeable. Looking only at the latest valuation, updated after an offer arrived, risks measuring the speed of the accounting update rather than the quality of earlier estimates.

DPI deserves weight, but its source matters

A distribution is more tangible than an unrealised valuation. That does not mean all distributions tell the same story.

It is necessary to distinguish money from sales and dividends from the return of unused capital or liquidity financed by borrowing. NAV facilities, loans backed by portfolio value, can bring distributions forward while leaving debt and its cost in the fund. ILPA explicitly analyses their effect on DPI. This does not mean the young fund under review uses them: their presence needs to be checked. ILPA, guidance on NAV facilities.

The form and terms also matter. Reporting may include shares distributed to LPs as well as cash, and distributions that can be recalled. Receiving securities is not equivalent to having sold them; receiving recallable cash is not equivalent to permanently concluding that commitment. GIPS, treatment of distributions.

Another mistake would be to encourage a manager to sell early simply to improve DPI before its next fundraise. A secondary sale can be a sound decision or sacrifice too much value. It should be assessed using the information available at the time: the achievable price, remaining risks, concentration, timing and alternatives. Observing years later whether the asset rose or fell is not enough.

The question is: “Does this distribution result from a good investment decision, and what remains to be financed or put at risk afterwards?”

The denominator and timing can change the interpretation

DPI does not have to rise every quarter. If a fund has distributed €2 million against €10 million paid in, it is 0.2x. If the fund calls another €5 million without new distributions, it falls to approximately 0.13x. No earlier distribution has disappeared: the denominator has changed.

TVPI can also decline when new capital arrives. If the contribution temporarily remains in cash, it increases both NAV and paid-in capital; for a fund above 1x, this tends to pull the multiple towards 1x without any loss in the value of its holdings.

Understanding these movements therefore requires capital calls, distributions and a bridge between opening and closing NAV that separates new investments, realisations, upward and downward revaluations, foreign exchange movements and costs.

TVPI and DPI also do not measure speed. Doubling capital in five years and in ten years produces the same multiple. With a single initial contribution and a single final receipt, the annualised returns would be approximately 14.9% and 7.2%, respectively: a mathematical example, not a forecast.

IRR, the Internal Rate of Return, considers cash-flow dates. During the fund’s life, however, it also incorporates the estimated closing NAV. A high IRR driven mainly by early valuation markups is not a measure of profit already received.

If the fund uses a subscription line to finance investments before calling capital from LPs, the timing of their contributions also changes. Viewing performance with and without the facility helps separate investment results from financing effects. The conventions adopted for reinvested capital and recallable distributions must also be clear. ILPA, performance methodology guidance.

“Top quartile” against which sample?

A meaningful comparison specifies at least the metric, gross or net treatment, reporting date, vintage, strategy, geography, currency and benchmark source. Vintage is the year assigned to a fund; its definition also needs to be checked.

A fund’s TVPI quartile may differ from its DPI or IRR quartile. The median fund is also not the same as a portfolio aggregating all those funds: large funds can carry very different weights. Cambridge Associates’ methodology explicitly distinguishes rankings by different multiples. Cambridge Associates, venture capital benchmark methodology.

Recent data also needs boundaries. Carta’s Q1 2026 report draws on 2,775 funds on its platform. It reports recovering TVPIs and still-limited distributions; fewer than 20% of funds in the 2017 and 2018 vintages had reached 1x DPI. This is evidence about that sample, not a rule for the timeline of every European or pre-seed fund. Carta, VC Fund Performance Q1 2026.

For very young funds, small valuation changes can materially alter relative rankings. Being above the median today does not demonstrate that the advantage will survive through exits.

Finally, outperforming other funds does not exhaust the LP’s question. There is an opportunity cost. A public-market comparison can be useful if it considers the same investment and distribution timing and a suitable index. Simply placing a fund’s IRR beside an index return measured over different dates is not enough.

From the fund’s track record to the manager’s ability

When results are still developing, it is necessary to ask what can already be attributed to the team. Which investments did it originate? Who decided the price and position size? Who worked with the company, managed a problem or chose to sell?

Attribution is also a central theme in the 27 May 2025 episode of How I Invest with Hunter Somerville, covering manager assessment before liquidity and spinouts. The episode notes distinguish access to opportunities, performance and individual contribution. How I Invest, Hunter Somerville.

A personal investment, a deal managed at a previous employer and a position in the current fund can all provide useful evidence, but should not be blended into a single undifferentiated return. The sample should also include losses and poor decisions.

For Fund II, another question matters: does the method remain credible at the new size? A good outcome achieved with small cheques, occasional access or a particular balance between investments and reserves does not automatically transfer to a larger fund.

These checks do not replace missing DPI. They support a more informed judgement while the final outcome remains unavailable.

What to request before deciding

To make the conversation with the manager useful, I would request a coherent set of evidence:

• TVPI, DPI, RVPI and IRR at the same date, with explicit scope and cost treatment.

• Cash-flow history and reconciliation to NAV, distinguishing distributions, retained proceeds and fund debt.

• Valuations of major holdings, material assumptions and concentration of value.

• Portfolio development against original milestones, including delays and losses.

• An explanation of completed sales and liquidity opportunities declined.

• Comparable benchmarks and documented attribution of results to the team.

I would not look for a number that closes the discussion. I would look for consistency between the accounts, the reality of the portfolio companies, the manager’s decisions and the timelines promised to investors.

TVPI describes total reported value. DPI records how much has been distributed. The track record becomes convincing when the manager can explain how it moves from one to the other, with what risks and over what period.

To discuss how to assess a venture capital fund’s track record, contact Vitantonio Santoro on LinkedIn or email vitantonio.santoro@atlassgr.com.