DPI: The Only Private Equity Metric That Shows You Real Cash

Quick take: Private equity distributions as a share of NAV collapsed from 29% in 2021 to just 9% in the first half of 2024 -- the lowest level in over a decade, according to Goldman Sachs Asset Management. If you are an LP evaluating a fund right now, one metric tells you whether the manager actually delivered: DPI, or Distribution to Paid-In capital. Not IRR. Not TVPI. DPI. It is the only number that shows you real cash in your account.

What DPI Actually Measures

DPI answers one question: how many dollars has the fund returned to me for every dollar I put in?

The formula is blunt. DPI = Cumulative Distributions to LPs divided by Paid-In Capital from LPs. Distributions are net of management fees and carried interest. Paid-in capital means what you actually wired -- not your committed amount, but what the GP has called.

Here is the concrete example. You commit $10 million to a buyout fund. The GP has called $7 million so far. You have received $5 million in distributions from two portfolio exits. Your DPI is $5M / $7M = 0.71x. You have gotten back 71 cents for every dollar you paid in. You are not yet whole.

Scale that to a real fund. A $500 million mid-market buyout fund at year five has called $420 million and made $175 million in distributions. Remaining portfolio NAV is $490 million. DPI = $175M / $420M = 0.42x. TVPI = ($175M + $490M) / $420M = 1.58x. That 1.58x looks decent on a pitch deck. But 42 cents of actual cash per dollar invested tells a different story. The other 1.17x exists only on paper, inside portfolio companies you cannot spend.

DPI carries two aliases in fund documents: the Realization Multiple and the Cash-on-Cash Return. When a fund winds down completely, DPI should converge with TVPI as RVPI (Residual Value to Paid-In) approaches zero. A large gap between TVPI and DPI in a mature fund is a red flag, not a feature.

DPI, TVPI, and RVPI: The Triangle of Confusion

The fundamental identity: TVPI = DPI + RVPI. Every dollar of claimed fund value is either in your pocket (DPI) or sitting in unrealized portfolio companies (RVPI). Most GPs pitch TVPI. You should care about DPI.

Metric	What It Measures	Manipulation Risk	Timing Sensitive?	Best Used For
DPI	Actual cash and in-kind distributions returned to LPs as a multiple of paid-in capital	Low -- cash was wired or it was not. Exception: NAV loan distributions	No	Re-up decisions, fund maturity assessment, actual vs. promised performance
RVPI	Unrealized NAV of remaining portfolio as a multiple of paid-in capital	High -- fair value marks set by the GP, subject to optimistic assumptions	Yes -- marks compress when exit markets close	Estimating remaining potential. Sophisticated LPs apply a discount.
TVPI	Total claimed value (realized plus unrealized) as a multiple of paid-in capital	Moderate -- headline inflated by RVPI marks. GPs select whichever metric flatters most.	No	Early fund life when DPI is near zero. Less meaningful as fund matures.
IRR	Annualized internal rate of return on cash flows, accounting for timing	High -- subscription lines inflate reported IRR by 200-500 basis points. J-curve timing is gameable.	Yes -- timing is both a feature and a vulnerability	Comparing funds of different sizes. Not reliable as the primary metric for mature funds.

The industry maxim captures it: TVPI is a story. DPI is proof.

What "Good" DPI Looks Like by Vintage and Strategy

DPI does not mean much without context. A 0.3x DPI in year two is normal. A 0.3x DPI in year eight is a problem. Cambridge Associates benchmark data gives you the reference points.

For PE buyout funds, the 2010-2014 vintages -- fully mature funds -- show DPI of 1.4x to 1.8x. The 2015-2018 vintages show only 0.6x to 1.0x as of Q3 2024, and those funds should be well into harvest mode. The 2019-2021 vintages show 0.1x to 0.3x, still early. Top-quartile buyout funds eventually return 2.3x to 2.7x TVPI and 18% to 22% net IRR -- but a fund with 2.0x TVPI may carry a DPI of only 0.8x if most positions remain unsold.

Fund Year	Buyout DPI (Typical Range)	VC DPI (Typical Range)	Red Flag Signal
Years 1-3	0.0x-0.1x	0.0x	None -- J-curve trough is normal
Years 4-6	0.1x-0.5x	0.0x-0.2x	Buyout at zero DPI by year 6 warrants direct questions
Years 7-9	0.5x-1.2x	0.2x-0.8x	Buyout below 0.8x at year 7 needs a credible exit roadmap
Year 10+	1.5x-2.0x+ (target)	1.0x-3.0x (wide range)	Any strategy below 1.0x DPI at year 10 is a serious warning

Venture capital lags buyout by two to four years on every milestone. The VC power law means two or three winners in a 30-company portfolio drive almost all returns, and those winners take ten to fifteen years to reach exit-ready stage. Carta's analysis of 1,803 venture funds found that at year three, only 9% of 2021-vintage funds had achieved any DPI above zero, compared to 25% for the 2017 vintage. By year five, only 39% of 2019-vintage VC funds had any DPI, versus 59% for the 2017 cohort. On average, funds now reach 1.0x DPI by year eight. Historically that milestone arrived at year six or seven, but the exit drought has pushed the timeline to years seven through nine or beyond.

How Managers Inflate TVPI While DPI Stays Low

TVPI can be moved without a single exit. DPI cannot. That asymmetry is where you need to focus.

Optimistic fair value marks. GPs set the NAV of unrealized portfolio companies using fair value methodologies they control. A PE fund claiming an HVAC services company is worth 20x EBITDA when public comparables trade at 15x inflates RVPI -- and therefore TVPI -- without any cash changing hands. DPI does not move one cent.

NAV loans and synthetic distributions. Facing LP pressure to show distributions when portfolio exits were scarce, GPs began borrowing against portfolio NAV and distributing the proceeds to LPs. The NAV loan market grew at a 30% CAGR from 2019 to 2023, reaching an estimated $100 billion to $150 billion in 2024. These distributions technically increase DPI. But the distribution is secured by leverage against unrealized assets. The loan must eventually be repaid, reducing future distributions. You are getting your money early with balance-sheet risk buried in the portfolio.

Dividend recapitalizations. The GP loads a portfolio company with new debt and routes the proceeds back to the fund as a dividend. The company's balance sheet deteriorates and future exit proceeds may be lower. In just the first seven weeks of 2025, US managers executed $22.4 billion in dividend recaps. That pushed DPI for participating funds, but the underlying companies absorbed the risk.

Subscription lines inflating IRR. GPs use short-term credit facilities to delay LP capital calls. This boosts reported IRR by 200 to 500 basis points with no change to actual realized returns. DPI stays completely unaffected -- which is exactly why IRR and DPI tell different stories about the same fund.

The SoftBank Vision Fund: A Case Study in Paper Gains

SoftBank Vision Fund 1 raised $98.6 billion -- the largest technology VC fund ever. As of 2025, it shows a TVPI of 1.4x and an IRR of 7%. Both figures underperform median PE buyout benchmarks. Vision Fund 2 at $56 billion performed worse: TVPI of 1.03x, IRR of 0.2%. The fund barely breaks even on paper.

SVF1 distributed $38.7 billion by March 2022 -- a real sum. But a significant portion of the remaining portfolio was carried at valuations from the 2019-2021 era when capital was cheap and startup multiples were at historic highs. When the team marked WeWork, DoorDash, and dozens of other holdings at peak-cycle valuations, TVPI looked compelling. Then the rate environment changed. The fund lost a record $32 billion in fiscal year 2023, and those RVPI marks came down hard.

The lesson is not that SoftBank was uniquely reckless. The lesson is that RVPI is a hypothesis. DPI told the more honest story: $38.7 billion distributed on a $98.6 billion fund is a 0.39x DPI on committed capital. You handed over $100 and got back $39 in real cash. The rest is still on paper. The 2018-2020 VC vintage shows the same pattern across the industry: median TVPI of 1.8x, median DPI of just 0.4x per Qubit Capital analysis. That 1.4x gap sits in RVPI -- marks set during a period of cheap capital that have not been tested by an actual exit.

What to Demand From Your Fund Manager

You are the LP. You have real power before you sign. Use it.

Ask for DPI by vintage year across all funds the GP has managed. Not aggregate TVPI. Fund-by-fund, vintage-by-vintage DPI. A manager who has run four funds should show you DPI for all four. If any mature fund (year eight or later) is below 1.0x DPI, you need a specific explanation, not a narrative about market conditions.

Demand ILPA-compliant reporting from day one. The ILPA January 2025 Performance Template, effective January 1, 2026, standardizes DPI, TVPI, RVPI, and both gross and net IRR in a locked format. GPs cannot reorder, remove, or modify line items. Every quarterly report must show all four metrics with and without subscription line impact. Ninety-two percent of institutional LPs say reporting quality directly influences their re-up decision.

Understand the waterfall structure. Under a European (whole-fund) waterfall, the GP receives no carried interest until the entire fund reaches DPI of 1.0x plus the preferred return -- typically 8%. That structure aligns GP and LP interests around actual cash returns. Under an American (deal-by-deal) waterfall, the GP can earn carry on individual profitable exits even while overall fund DPI sits below 1.0x. In a deal-by-deal structure, monitoring DPI at the fund level becomes even more critical.

Ask how the GP defines a distribution. A cash distribution from a genuine portfolio exit differs from a NAV loan-funded distribution or a dividend recap payment. Ask directly: "Of your total distributions to date, what percentage came from full or partial portfolio exits versus debt-funded distributions?" A GP who cannot answer this question clearly is one you should study harder before committing.

Look at the exit backlog. Approximately 28,000 to 29,000 PE-backed companies remain unsold globally as of 2024, representing $3.2 trillion to $3.6 trillion in unrealized value -- the largest exit backlog in PE history, according to Bain & Company. Median hold periods stretched from 4.2 years in 2010 to 6.8 years in 2023. If your manager's portfolio is heavily weighted toward 2015-2018 vintage companies, ask directly: what does DPI look like if the IPO market stays closed for another two years?

There is also a positive signal worth noting. Cambridge Associates data shows that in H1 2025, US PE distributions reached $78.9 billion against $67.6 billion in capital calls -- the first meaningful positive net cash flow period in several years. Compare that to 2022, when PE firms called $64 billion more than they distributed, and 2023, when they called $23 billion more. Demand the data that shows where your specific manager stands in that context.

The Honest Risk Caveat

DPI is the most honest metric in private equity. That does not make private equity a safe investment.

A high DPI means cash was returned, but it does not tell you what IRR was achieved, whether the GP took excessive risk to generate those returns, or whether the strategy repeats in a different rate environment. DPI also has no time dimension. A fund that returned 1.5x DPI over twelve years is not as good as one that returned 1.5x DPI over seven years. That is why you use DPI alongside IRR, not instead of it. The ILPA template requires both for exactly this reason.

Private equity is illiquid by design. Commitments typically run ten to twelve years. You cannot pull out when DPI disappoints. Diligence before you sign is the primary risk control -- and DPI history across a manager's vintage track record is the sharpest tool you have.

The 74% of institutional LPs who now rank DPI as their primary re-up criterion per ILPA's 2024 survey are not doing so because DPI guarantees good outcomes. They do it because DPI is the one metric a GP cannot dress up, smooth out, or mark to a model. Either the cash arrived in the account or it did not.

Author Disclosure: Jeff Barnes, MBA has no personal position in any company, fund, or platform named in this article. Angel Investors Network has no current commercial relationship with any party mentioned. AIN provides marketing and education services, not investment advice. Past performance does not guarantee future results. All investments involve risk, including loss of principal.