Data sourced from Joe, the private fund performance platform powered by Dakota. Learn More | Request Access
Five acronyms carry almost every private fund performance conversation: Net IRR, TVPI, DPI, RVPI, and MOIC. Joe, Powered by Dakota, tracks the first four across 18,500+ funds and 159,600+ performance records in seven asset classes, and the fifth shows up in nearly every investment committee memo and fundraising deck built on top of that data.
The math behind each one is simple. The trouble is that two funds can report the same TVPI and mean completely different things by it, MOIC gets quoted gross in one deck and net in the next, and IRR can be flattered or punished by timing that has nothing to do with investment skill. Most disagreements about a fund's performance turn out to be disagreements about definitions.
In this article, we'll define each of the five metrics with its formula, work every one against the same fund so the numbers tie together, name what each metric hides, and flag the specific ways each gets misread.
The Reference Fund
Every example below uses one fund, so the metrics reconcile against each other rather than floating independently.
Fund XI, L.P. Middle market buyout, vintage 2019.
|
Input
|
Value
|
|
Committed capital
|
$500M
|
|
Paid-in (called) capital
|
$400M
|
|
Cumulative distributions
|
$360M
|
|
Remaining value (NAV)
|
$280M
|
|
Capital invested into portfolio companies
|
$380M
|
|
Realized + unrealized value of those investments
|
$722M
|
1. Net IRR (Internal Rate of Return)
Definition: The annualized, money-weighted rate of return delivered to investors after management fees, fund expenses, and carried interest.
Formula: The discount rate at which the net present value of all fund cash flows equals zero.
0 = Σ [ Cash Flowt ÷ (1 + IRR)^t ]
Capital calls enter as negative cash flows, distributions as positive, and the ending NAV is treated as a final positive flow on the measurement date.
Fund XI: 18.4% Net IRR.
What it tells you: How fast capital compounded for the investor, with the timing of every call and distribution factored in. IRR is the only one of the five metrics that accounts for when money moved.
What it hides: Scale. A $5M early win can carry a fund's IRR for years. IRR also implicitly assumes interim distributions get reinvested at the same rate, which rarely happens, and it can be manipulated with subscription lines of credit that delay capital calls and shorten the measured holding period.
Common misread: Comparing IRR across vintages. A 2019 fund and a 2023 fund are measuring different points in their own life cycles, and a young fund's IRR is barely a signal at all.
Gross IRR vs. Net IRR
Gross IRR measures the return on the underlying investments before fund-level fees and carry. Net IRR measures what the investor actually earned.
For Fund XI, a 24.6% gross IRR nets down to 18.4% after management fees, fund expenses, and carried interest. That 6.2 point gap is the fee load, and it is the single most common apples-to-oranges error in performance comparison. Confirm which one you are looking at before comparing anything.
2. TVPI (Total Value to Paid-In Capital)
Definition: Total value created per dollar called, combining cash already returned and value still held in the portfolio. Also reported as "total value multiple" or, loosely, "net multiple."
Formula:
TVPI = (Cumulative Distributions + Remaining Value) ÷ Paid-In Capital
Fund XI: ($360M + $280M) ÷ $400M = 1.60x
What it tells you: The fund's total value creation to date, insensitive to the cash flow timing that swings IRR. TVPI is the cleanest single number for "how much has this manager made per dollar."
What it hides: Everything about liquidity, and everything about how the NAV was struck. A 1.60x TVPI made up of 1.50x cash and 0.10x NAV is a very different fund from one made up of 0.20x cash and 1.40x NAV, and TVPI reports both as 1.60x.
Common misread: Treating TVPI as realized performance. It is not. The unrealized half rests on the manager's own valuation marks.
The other half is the peer group. A metric only means something next to funds of the same vintage, strategy, geography, and fund size as the one in front of you. Joe, powered by Dakota, tracks Net IRR, TVPI, DPI, and RVPI across 18,500+ funds so you can run that comparison in minutes instead of assembling it by hand. Request access.
3. DPI (Distributed to Paid-In Capital)
Definition: Cash and stock actually returned to investors per dollar called. Often called the "realization multiple" or "cash-on-cash multiple."
Formula:
DPI = Cumulative Distributions ÷ Paid-In Capital
Fund XI: $360M ÷ $400M = 0.90x
What it tells you: What has genuinely left the fund and reached investors. DPI is the one metric a manager cannot mark their way into. A fund at DPI above 1.0x has returned more than it called.
What it hides: Remaining upside. A fund can post a strong DPI by selling its best assets early and holding a weak residual portfolio, and DPI will look identical to a fund that realized evenly across a strong book.
Common misread: Judging DPI without fund age. A 2023 vintage at 0.05x DPI is behaving normally. A 2014 vintage at 0.60x DPI is a problem. DPI is only interpretable against same-vintage peers.
4. RVPI (Residual Value to Paid-In Capital)
Definition: The value still sitting in the portfolio per dollar called, at the manager's current carrying value.
Formula:
RVPI = Remaining Value (NAV) ÷ Paid-In Capital
Fund XI: $280M ÷ $400M = 0.70x
What it tells you: How much of the story is still unwritten. High RVPI late in a fund's life means concentrated dependence on future exits landing at or above current marks.
What it hides: Whether the marks are right. RVPI is the most valuation-dependent of the five metrics, and in slower exit environments it is where optimism accumulates.
Common misread: Reading high RVPI as upside. It is exposure, not upside. It becomes upside only if the exits clear the marks.
5. MOIC (Multiple on Invested Capital)
Definition: Total value generated per dollar of capital put to work in investments. Unlike the four metrics above, MOIC's denominator is capital invested, not capital called, which means it excludes fees, expenses, and uninvested capital.
Formula:
MOIC = (Realized Value + Unrealized Value of Investments) ÷ Invested Capital
Fund XI: $722M ÷ $380M = 1.90x gross MOIC
What it tells you: Investment selection ability, stripped of fund structure. MOIC is the deal-level scorecard, which is why it appears at the individual portfolio company level far more often than at the fund level.
What it hides: Time and cost. MOIC has no time dimension at all, so a 1.90x held for three years and a 1.90x held for eleven are identical under this metric. It also ignores the fee load, which is why Fund XI's 1.90x gross MOIC becomes a 1.60x net TVPI.
Common misread: Comparing a gross MOIC to a net TVPI and concluding the manager outperformed. This is the most consequential definitional trap of the five. MOIC is quoted gross by default in most decks and net only when explicitly labeled, so the label matters more than the number.
How the Five Metrics Relate
Three relationships do most of the work in reading a fund report.
TVPI = DPI + RVPI. This always holds, by construction. For Fund XI: 0.90x + 0.70x = 1.60x. The split is more informative than the total, because it tells you what share of the reported value has actually been converted to cash. Fund XI has realized 56% of its total value (0.90 ÷ 1.60), which is reasonable for a 2019 vintage.
IRR and multiples answer different questions. A multiple tells you how much, an IRR tells you how fast. They diverge sharply with holding period:
|
Scenario
|
Multiple
|
Hold Period
|
Approx. IRR
|
|
Fast, modest return
|
1.50x
|
2 years
|
22.5%
|
|
Balanced
|
2.00x
|
4 years
|
18.9%
|
|
Slow, strong return
|
3.00x
|
10 years
|
11.6%
|
|
Slow, modest return
|
2.00x
|
8 years
|
9.1%
|
The 1.50x fund beats the 3.00x fund on IRR and loses badly on total dollars returned. Neither metric is wrong. They are measuring different things, and a fund reporting only one of them has made a choice about which story to tell.
Gross and net are two different scales. Gross IRR and gross MOIC describe investment performance. Net IRR, TVPI, DPI, and RVPI describe investor experience. Never compare across the line.
Two Definitional Details That Change the Numbers
Paid-in capital is not always defined the same way. Some reports use capital called for investments only. Others include management fees and fund expenses in the denominator. The second definition produces a lower TVPI and DPI for an identical fund. When multiples from two sources disagree by a few basis points of a turn, the denominator is usually why.
Uncalled capital sits outside all five metrics. A fund that has called 40% of commitments and one that has called 95% can post the same TVPI while representing very different amounts of committed-but-unexposed capital. Fund XI has called $400M of $500M, leaving $100M uncalled, and no metric above reflects that.
Reading the Five Together
A workable order for a fund report: start with DPI to establish what is real, add RVPI to size what is still at risk, confirm TVPI as the sum, use Net IRR to judge pace, and treat gross MOIC as a read on deal selection rather than investor outcome. Then discard any comparison that crosses vintages, strategies, or the gross-net line.
The failure mode is not misunderstanding any single formula. It is anchoring on whichever metric looks strongest and comparing it against a peer group that was never constructed to be comparable.
See Where a Fund Actually Stands
Joe, powered by Dakota, tracks Net IRR, TVPI, DPI, and RVPI across 18,500+ funds and 159,600+ performance records in seven asset classes, with every record verified by Joe's research team before publication. Filter by vintage year, asset class, sub-strategy, geography, and fund size to build the peer group your comparison actually requires, then export to Excel or CSV for the diligence file.
Coverage includes 5,200+ private equity funds, 4,600+ private real estate funds, 1,900+ private credit funds, 1,300+ real assets funds, 1,100+ venture capital funds, and 590+ private infrastructure funds, plus evergreen and interval funds.
Define the metrics once. Then benchmark them against funds that are actually comparable. Request access to Joe, powered by Dakota.
