IRR, MOIC, and the J-Curve: Why Do Early Private Fund Numbers Mislead?

A private fund can report a 25% IRR in year two and still return less money than investors put in. That is not fraud; it is how the metrics work. Private market performance figures — IRR, MOIC, TVPI, DPI, RVPI — measure different things, react differently to timing, and behave strangely in a fund's early years. If you already know what a fund-type RWA product is and want to read its performance section properly, the single most useful skill is knowing what each number can and cannot tell you.

This article explains the three concepts that cause the most confusion: IRR and its sensitivity to cash flow timing, multiples and the split between realized and unrealized value, and the J-curve pattern that makes early figures — good or bad — unreliable. All numbers below are labeled hypothetical examples, not any real product's results.

Why Private Fund Numbers Need Their Own Rulebook

Public market performance is simple to state. A stock has a price every trading day, so a return over any period is just the change in price plus dividends. There is nothing to interpret.

Private funds do not work that way. Money goes in gradually as the manager calls capital, comes back gradually as investments are sold, and everything in between is valued by estimate rather than by a market price. A private fund's "return" is therefore a construction: someone has to decide how to handle irregular cash flows and how to value positions that have not been sold. Different constructions give different numbers for the same fund, which is why performance figures for non-listed assets can only be understood alongside how they were measured.

Two families of metrics dominate. IRR answers "how fast did the money grow while it was actually invested?" Multiples answer "how much money came out relative to what went in?" Both are legitimate. Both can mislead if read alone — especially early in a fund's life, and especially before fees are deducted, a topic covered separately in fund fees explained.

What Does IRR Measure, and How Can Timing Bend It?

IRR stands for internal rate of return. It is the annualized rate that makes all of a fund's cash flows — money in, money out, plus the current value of what is still held — net out to zero. In plain terms, it asks: at what constant yearly growth rate would this exact sequence of payments make sense?

The key property is that IRR is money-weighted: it depends heavily on when cash moves, not just how much. Getting money back quickly produces a high IRR even if the total profit is small. A fund that doubles your capital in one year shows a 100% IRR; a fund that doubles it over seven years shows about 10%. Same multiple, very different IRR — and neither number alone tells you which outcome you would prefer.

This timing sensitivity creates room for management. The most discussed technique is the subscription line: in one sentence, the fund borrows against investors' committed capital to make investments first and calls investor money later, which shortens the measured holding period and mechanically lifts the reported IRR without changing the underlying investments at all. The Institutional Limited Partners Association (ILPA), an industry body for fund investors, published guidance in 2017 recommending that funds disclose IRR both with and without such credit lines, precisely because the two figures can differ meaningfully. When you see an IRR, it is fair to ask which version you are looking at.

Two more reading rules for IRR:

• Net or gross. Gross IRR is calculated before management fees, fund expenses, and the manager's performance share; net IRR is what investors actually experience. The gap is often several percentage points per year.
• Interim or final. Until a fund is fully wound down, its IRR includes an estimate — the current value of unsold holdings. An interim IRR is partly a forecast wearing the clothes of a result.

MOIC, TVPI, DPI, RVPI: What Do the Multiples Tell You?

Multiples ignore time entirely and just compare money out to money in.

• MOIC (multiple on invested capital) — total value created divided by capital invested. A MOIC of 1.8x means every unit invested is now worth 1.8 units, on paper or in cash.
• TVPI (total value to paid-in) — the fund-level version: total value, realized plus unrealized, divided by the capital investors have paid in.
• DPI (distributions to paid-in) — cash actually returned to investors, divided by paid-in capital. This is the realized part.
• RVPI (residual value to paid-in) — the estimated value of what the fund still holds, divided by paid-in capital. This is the unrealized part.

The arithmetic is simple: TVPI = DPI + RVPI. The reading rule that matters is the split between the two components. DPI is cash that has left the fund and reached investors; it cannot be revised. RVPI is a mark — a valuation estimate for assets that have not been sold, which can move up or down before the fund ends. A fund reporting a 1.8x TVPI built from 0.2x DPI and 1.6x RVPI has mostly paper gains; a fund reporting 1.8x TVPI from 1.5x DPI and 0.3x RVPI has mostly banked them. Same headline, very different certainty. For Pre-IPO and venture-style funds, those marks often trace back to the price of the latest funding round, which has its own limitations — see how Pre-IPO valuations work.

Multiples also say nothing about time or exit. A 2.0x multiple achieved in four years and a 2.0x multiple achieved in twelve years are the same MOIC but very different investments once you account for how long capital was locked up and what else it could have done. That is why multiples and IRR are read together, alongside the fund's term and its actual exit mechanics — never one metric in isolation.

The J-Curve: Why Do Early Numbers Look Bad — or Too Good?

Plot a typical private fund's net IRR over its life and you get a J shape: negative in the early years, then climbing. The mechanics are mundane:

1. Fees start on day one. Management fees are typically charged on committed capital from the start, while investments need years to mature. Early on, costs are real and gains are not.
2. Deployment is slow. Capital is called over several years, so early performance is measured on a small, immature portfolio.
3. Losses show up before wins. Failing investments tend to get written down early; successful ones take years to exit. Accounting conservatism front-loads the bad news.

So a negative IRR in years one to three usually tells you almost nothing about final outcomes. But the trap runs both ways: early numbers can also look too good. If a young fund marks up a position after a new funding round, RVPI and interim IRR jump on the strength of a valuation event, not a sale. Early paper gains can overstate just as easily as early fees understate.

Hypothetical example (illustrative only, not any real product): a fund raises 100 in commitments. By the end of year 2 it has called 40, paid 4 in fees and expenses, and its investments are still held near cost — net IRR is roughly -10% and TVPI about 0.9x. In year 3, one holding raises a new funding round at double the prior valuation; the fund marks it up, and interim net IRR swings to +20% with TVPI at 1.3x — all of it RVPI, none of it distributed, and dependent on that mark holding through an actual exit. By year 8, after real exits, the fund lands at 1.6x DPI and a 12% net IRR. Every intermediate number was "true" at the time; none was a reliable preview of the end result — which depended on the underlying companies' performance, the fund's term, and whether exits actually materialized.

Why Comparing an Interim IRR to a Stock Return Is Meaningless

It is tempting to line up a fund's since-inception IRR against, say, an index's annual return and declare a winner. The comparison fails on several counts.

First, the definitions differ. A stock return assumes your money was fully invested for the whole period. An IRR is computed only on capital while it was deployed; committed money sitting in your account waiting to be called earns the fund's IRR in the fund's marketing, but not in your bank balance. Second, an interim IRR leans on unrealized marks, while an index return is fully realized pricing every day. Third, "since inception" quietly does a lot of work: it anchors the figure to the fund's chosen starting point, blends early subscription-line effects into the base, and can keep a strong early exit propping up the headline number for years after performance has flattened. A fund can show an attractive since-inception IRR while its recent vintage of investments has done little — the average hides the trajectory.

Vintage year matters too, and deserves one paragraph. A fund's vintage is the year it started investing, and it largely sets the price environment the fund bought into and the exit environment it sold into. Funds that deployed capital into cheap markets and exited into expensive ones look brilliant; the reverse looks poor, with the same manager skill. Serious comparisons are therefore made against funds of the same vintage and strategy, not against a stock index or a fund from a different era. A standalone IRR with no vintage context is a number without a denominator.

What to Ask Before Trusting Any Performance Figure

None of this means private fund metrics are useless — it means they are answers to specific questions, and you need to know which question was asked. Before treating any performance figure as information, check:

The general habit: any return figure — IRR or multiple, past or projected — is only readable together with its source (what the underlying assets did), its term (how long capital was committed), its exit path (how value actually converts to cash), and its risks (valuation, liquidity, credit, and manager risk among them). Past performance, interim or final, does not indicate future results, and none of this is investment advice.

When you review a fund-type RWA product on Bifu's RWA page, apply the same checklist to the product documents: look for whether figures are realized or marked, net or gross, and measured over what period, and read them next to the stated term, exit arrangements, and risk disclosures before forming any view.

FAQ

What counts as a "good" IRR for a private fund?

There is no universal number, because IRR only means something next to its vintage year, strategy, and how much of it is realized versus marked. A figure that looks strong for one vintage and strategy can be mediocre for another, so serious comparisons are made against funds of the same vintage and strategy, not against a fixed benchmark.

How can I tell if a fund's reported IRR includes a subscription line?

Check whether the fund discloses IRR both with and without credit-line financing, which is the practice ILPA has recommended since 2017 specifically because the two figures can differ meaningfully. If a product only shows one IRR with no mention of financing effects, that missing disclosure is itself worth asking about before trusting the number.

Is DPI more important than TVPI when evaluating a fund?

DPI carries more weight once a fund is well into its life, because it is realized cash that cannot be revised, while TVPI still includes RVPI, a valuation estimate that can move before exit. Early in a fund's term, TVPI is often the only figure available, so it is only meaningful alongside how much of it is DPI versus RVPI.

Can a fund have a high MOIC but a low IRR?

Yes. A high MOIC reached slowly still produces a modest IRR, because IRR is money-weighted and penalizes capital that took a long time to return, while MOIC only counts the multiple achieved, not how long it took. That is why the two metrics need to be read together rather than treating either one alone as "the" return.

Related Reading

• New to this? Start with what RWA is.
• In the same area: how to read a fund-type RWA.