J-curve in private equity and VC: what it is, how long it lasts, and a simulator

What is the J-curve?

Key facts

• Hamilton Lane: the negative period "generally spans three to four years following the fund's inception". (Hamilton Lane)
• On cash alone, it took "almost until year 8" for the average and median fund IRR to turn positive in Ljungqvist and Richardson's data. (Ljungqvist & Richardson, 2003)
• The median 2021-vintage US VC fund on Carta still had a negative net IRR (−1.5%) after three years, against 19.4% for the 2019 vintage at the same age. (Carta, 2024)
• A 2% annual fee on commitments for ten years costs 20% of a fund's committed capital, which is why returns start negative. (Metrick & Yasuda, 2010)
• A subscription line "shortens the J-curve", ILPA notes; in a study it cites, the median IRR lift was "206bps by year 3, falling to 35-45 bps by the end of the fund life". (ILPA, 2017)

J-curve simulator

Set a fund's fees, investment pace, holding period and returns, and see the cash J-curve: what LPs have paid in minus what they've got back, year by year. It is a deliberately simple model: fees on commitments for ten years, equal yearly investments, every investment exiting after the same holding period at the same multiple, and carry with no hurdle. Real funds recycle capital, step fees down and exit unevenly.

How to measure the J-curve

Cumulative net cash in year t = Σ (Distributions − Capital calls), years 1 to t

In words: add up, year by year, what the fund has paid back to LPs minus what it has called from them. The low point is the bottom of the J; the year the total turns positive is break-even.

• Capital calls: Money LPs pay in, for investments and for management fees and expenses.
• Distributions: Cash (or stock) the fund pays back to LPs, after the manager's carried interest.
• t: Years since the fund's first close.

Worked example

A $100M fund with a 2% fee on commitments, investing over four years, each investment sold after five years at 3x, with 20% carry. The results below are computed by the simulator's tested code.

LPs get their money back: Year 7 Deepest point (cumulative cash, % of fund): -90.0%Net multiple to LPs: 2.08xNet IRR to LPs: 17.3%

Fees 2% a year for 10 years; investments spread over 4 years, each held 5 years at 3x, 20% carry on profits.

LPs are 90% of the fund out of pocket by year 5, before the first exit, and have their money back in year 7. Hold each investment seven years instead of five and they end with the same 2.08x, but break-even moves to year 9 and the net IRR falls from 17.3% to about 12% (our runs of the simulator).

In a spreadsheet

• Net cash flow for a year, with calls in B2 and distributions in C2: =C2-B2
• Cumulative net cash, copied down column E: =SUM($D$2:D2)
• The bottom of the J, for years in rows 2 to 13: =MIN(E2:E13)
• Break-even year (first year the cumulative total is zero or more): =MATCH(TRUE, INDEX(E2:E13>=0, 0), 0)
• Net IRR of the yearly flows: =IRR(D2:D13)

How founders and investors read the J-curve

How deep is the J-curve, and how long does it last?

How deep and how long the J-curve runs depends on what you measure. On reported IRR, most VC funds sit near zero for a year or two; on cash, LPs wait about seven to eight years to get their money back.

Benchmarks

The dip was deeper for recent vintages: after a year, Carta's 2022 funds had a median net IRR of −12.0%, against −1.2% for 2019 funds.

Venture waits longer than buyout. Ljungqvist and Richardson found that "venture funds take about a year longer to break even".

Compiled Oct 2, 2026. Next review: December 2026, after Carta's Q3 2026 release. Left out: Cambridge Associates (Its licence bars reuse of its benchmark figures without written permission); PitchBook, "Behind the J-Curve" (2024) (The analyst note is gated; we read only its abstract). Download these benchmarks as CSV.

The J-curve in one LP's book: CalPERS by vintage

All of CalPERS's active private equity partnerships, grouped by vintage, as CalPERS reported them as of 31 March 2026. The totals, DPI and TVPI are our arithmetic from CalPERS's fund-by-fund figures; across all 485 partnerships they come to 1.52x, matching CalPERS's reported net multiple of 1.5x. (CalPERS, Mar 2026)

Cash tells the J-curve story. CalPERS has had about 1.5 times its money back from its 2014 and 2015 commitments, a third of it from 2021's and almost nothing yet from 2023 on. The marks don't dip: its 2023 and 2024 commitments are already valued at about 1.5x. In this book the J shows up in cash, not in reported value.

CalPERS lists "all active PE partnership investments", so funds that have fully wound down drop out of the older vintages. It also warns that analysis of funds from 2021 on "would not generate meaningful results as private equity funds are understood to be long-term investments."

Cash J-curve vs IRR, TVPI and DPI over time

The cash J-curve's low point is the most LPs are ever out of pocket, and the year it crosses zero is the year DPI reaches 1x: distributions have caught up with everything paid in.

Common mistakes when reading the J-curve

1. Judging a young fund by its IRR. Early IRRs are mostly fees and marks. CalPERS treats performance analysis of its funds from 2021 on as not meaningful, and so should anyone comparing a young fund with a mature one.
2. Mixing up the two J-curves. A reported-IRR curve can turn positive in two or three years; a cash curve took "almost until year 8" in Ljungqvist and Richardson's data. Check which one a chart shows.
3. Missing the credit line. Delaying capital calls makes the dip look shallower without changing the money made. Albertus and Denes found the effect "amplified for young funds, inflating IRRs by an average of 9.7 percentage points" (2019).
4. Ignoring the fee terms. Fees on commitments run whether or not capital is invested; Metrick and Yasuda show 2% a year for ten years is "20% of committed capital" (2010). Fees that step down after the investment period make the curve shallower.
5. Expecting every vintage to follow the same path. Carta: "The median IRR for vintage 2021, by contrast, failed to break into positive territory after the same amount of time" (Carta, 2024). Entry valuations shape the curve as much as fees do.

Where the J-curve came from and how it's regarded today

How it's regarded today

• Recent vintages climbed out late. "The median net IRRs for both the 2021 and 2022 vintages of venture funds on Carta have climbed out of negative territory, currently sitting at 1.4% for the 2021 vintage and 0.7% for 2022" (Carta, Mar 2026).
• Credit lines reported both ways. ILPA's 2025 template asks GPs to show IRRs and TVPI with and without the subscription line, so LPs can see how much of a shallow J is borrowing.
• Cash is still slow. PitchBook: "Halfway through their standard 10-year fund term, 2021 vintage US VC funds are at the lowest DPI multiple of any vintage since at least 1997" (PitchBook, Aug 2026).

What experts say about the J-curve

Where they disagree

A CFA Institute author and an LP on whether the J-curve still describes private funds.

J-curve FAQ

Related metrics and guides

• IRRThe rate the J-curve is usually drawn in.
• DPICash back; the cash J-curve breaks even when DPI hits 1x.
• TVPIValue per dollar paid in, at any age.
• RVPIThe value still waiting to be sold.
• MOICThe deal-level multiple behind the curve.
• VC fund metrics: the guideHow the fund metrics read together.
• Reserves and follow-on planningHow follow-on checks stretch a fund's investment period.