Year three is when a lot of private fund investors start second-guessing the whole thing.
The quarterly statement still looks bad. Capital keeps getting called. Distributions are thin or nonexistent. IRR is negative or barely above zero. TVPI sits under 1.0x, or just above it if the manager has marked a couple of deals. It feels like you paid for a product that is not working.
That feeling has a name. It is the J-curve. And year three is often the worst place to judge the investment.
What the J-curve actually is
Plot a closed-end private equity fund’s cumulative net cash flow over time. Money goes out first: capital calls, management fees, deal costs. Almost nothing comes back. The line drops below zero. Then, if the fund works, exits start arriving, valuations rise, and the line bends up. The shape looks like a J.
The same pattern shows up in IRR. Early IRRs are ugly because the math is cruel when the denominator is small, fees are live, and realizations have not started. A vintage-year IRR only becomes useful after several years. Reading a year-two or year-three IRR as a verdict is how people abandon decent funds at the trough.
Why year three feels like a mistake
By year three, you have usually sent a large share of the commitment. Paid-in capital on a typical buyout or value-add fund often sits around 55% to 75% of the original pledge. Fees have been charged on the full commitment, not just the invested slice. Portfolio companies are still being fixed, not sold. DPI, the cash you have actually received, is often close to zero.
That combination is unpleasant. You have real cash out the door. You do not have real cash back. Marks on private companies are conservative early, and write-downs look like losses even when they are accounting, not exits. The statement does not say “this is the planned dip.” It just shows a negative number.
Cambridge Associates data puts the median time to peak drawdown for private equity funds around two to three years from inception. Preqin vintage data on buyouts has shown the cash-flow bottom arriving around year four, with funds turning cash-flow positive later, often around year seven on average. The exact year moves with strategy and market, but the middle of the investment period is where the pain concentrates.
The three forces that create the dip
First, fees start on day one. A 1.5% to 2% management fee on committed capital during the investment period is normal. If you committed $5 million and only $1.5 million has been called, that fee is a much larger percentage of capital actually at work than it looks in the pitch deck. Organizational expenses hit early too.
Second, investments go out before value shows up. Buyouts close with transaction costs. Value-add real estate spends on capex before rent rises. Venture checks sit at cost until a later round or exit resets the mark. The fund is doing work. The scoreboard is not.
Third, distributions lag. PitchBook has noted that about half of funds make a first distribution by roughly 1.5 years, but a chunk wait closer to 2.5 or even 3.5 years. Meaningful harvest usually clusters later, often around years six and seven. Until then, net cash flow stays negative.
None of that requires a bad manager. A bad manager makes the trough deeper and longer. A good one still has a trough.
Not every J looks the same
Buyout funds usually have a shallower, shorter curve than venture. Portfolio companies already produce earnings, and exits can arrive sooner. Venture dips deeper and stays down longer because winners take years and many checks go to zero. Private credit and some infrastructure strategies are flatter. Interest and contracted cash start earlier, so the letter J is less dramatic.
Subscription lines can hide the dip. If the fund borrows against uncalled commitments and delays calling your cash, reported IRR can look better than the underlying cash story. A surprisingly pretty year-two IRR is a reason to ask about the credit facility, not a reason to celebrate.
How to tell a J-curve from a real problem
Year three is not a free pass. It is a bad time to use IRR as a report card, and a decent time to watch operations.
Look at deployment pace. Is paid-in capital moving in line with the investment period, or is the fund sitting on commitments while still charging full fees? Look at the portfolio, not just the composite number: are companies hitting the operating plan, or is every mark a story? Watch TVPI before you obsess over DPI. Value has to appear on paper before cash comes back. Then watch whether DPI starts to move in years four to six. If nothing is being realized and marks keep slipping, that is no longer just the curve.
Vintage matters. A 2021 venture fund still looking rough in year three is a different conversation than a 2016 buyout fund still underwater. Markets, rates, and exit windows change the timing. Holding periods in buyouts have stretched in recent years, which can push the upswing later than older textbooks imply.
What year three is asking of you
Liquidity, mostly. Capital calls do not pause because the statement looks ugly. The J-curve and the call schedule stack: you keep wiring money into a position that has not paid you back yet. That is why uncalled commitments belong in a cash plan, not in the “maybe later” column.
It is also asking for patience with the metric set. Early IRR is a timing artifact. Final performance shows up in DPI and net multiple after exits, usually deep into years seven through twelve. The hook at the top of the J is the part that matters. Year three is the bottom of the letter.
The mistake is treating that bottom as the result. Sometimes it is. More often it is the fee-and-timing math doing exactly what the structure was built to do.