Venture capital had a remarkable 2025—at least on paper.
Global deal value reached about $512 billion, the second-highest total on record, as AI startups raised historic sums and Andreessen Horowitz closed a $15 billion fund in January.
The headlines were relentless, and the optimism was loud.
Beneath that activity, something in the system is fundamentally broken.
Venture capital depends on exits—IPOs and acquisitions that turn paper valuations into cash for the limited partners who finance the funds.
Right now, that cash is not flowing back.
Since 2022, investors have received almost $200 billion less than they put in, and fundraising for new VC funds has fallen to its lowest level in a decade.
More than 1,500 private companies, collectively valued at around $6 trillion, have no realistic path to converting those marks into money.
This is what a liquidity crisis looks like when the industry has every incentive not to call it one.
The exit market has effectively closed for most companies. In 2021, 311 venture-backed firms went public, cash flowed freely, and the model worked as designed—until it did not. In 2022, just 38 companies were listed.
In 2024, the number was 72.
Last year, despite a modest recovery in total exit values, only 62 companies completed IPOs out of a queue of more than 1,500 waiting for their moment.
At the current pace, it would take roughly 49 years to work through the existing backlog of US venture-backed unicorns.
The acquisition route is not much better
The major technology buyers that historically absorbed smaller startups have spent years under antitrust scrutiny, making large deals slower and less certain.
The buyers who remain are disciplined in their pricing: they are not paying 2021 valuations for 2025 assets.
So companies wait—operating, sometimes profitably—but their value stays locked up and unreachable.
The clearest way to see the strain is in DPI, or distributions to paid-in capital, which measures how much actual cash a fund has returned relative to what investors contributed.
It cannot be smoothed or estimated; either the money came back, or it did not.
The data are damning. Of all venture funds launched in 2019, more than three in five had not returned a single dollar after five years.
The median fund from that vintage had returned just 22 cents for every dollar invested by 2024, compared with 47 cents for 2016 funds at the same point.
Each successive wave of funds is performing worse than the one before it.
Across private markets, distributions fell to about 6% of assets under management in the first half of 2025, less than half the roughly 14% ten-year average.
The consequences are direct and compounding. Pension funds, endowments and family offices are not receiving the cash they planned on.
Without it, they cannot commit to new funds, which is why US venture fundraising fell to its lowest level since 2018 last year, with new fund closings at only about 30% of their 2021 peak.
$6 trillion on paper, very little in cash
More than 1,500 private companies are currently valued at $1 billion or more, collectively worth $6 trillion according to Crunchbase.
Over 60% have not raised at a disclosed valuation in more than three years. Nearly half of US unicorns have been in investor portfolios for over nine years.
Many are real businesses generating real revenue. The problem is that they were last valued when investors were paying 15 to 20 times revenue for fast-growing software companies.
Public markets now value comparable businesses at 3 to 5 times revenue. A company that raised at a $5 billion valuation in 2021 cannot realistically go public at $2 billion in 2025 without inflicting losses on its most recent investors. So it does not.
The mark stays on the books, and every quarterly letter describes a portfolio that looks significantly more valuable than any real buyer would confirm.
The AI numbers are hiding everything else
The obvious counterargument is AI, and it is fair up to a point. AI deals captured 65% of all US venture deal value in 2025. OpenAI raised $40 billion in a single round.
A group of the seven most valuable private tech companies is now worth $1.3 trillion combined.
But half of all venture dollars last year went into just 0.05% of deals. The top 10 funds captured over 42% of all LP commitments.
Andreessen Horowitz’s $15 billion raise alone represented 18% of every dollar committed to US venture over the prior year.
The AI boom is producing real value, but in an extraordinarily narrow band at the very top of the market, while the rest of the portfolio waits in a market with no obvious mechanism to absorb it.
The median time for a company valued at over $500 million to reach an IPO has now exceeded 11 years, the longest on record.
What investors should understand?
The venture industry is undergoing a significant transformation, but it will not change overnight.
The strongest funds will still generate real returns, and the clear AI winners will eventually go public and create substantial wealth for their early backers.
What is in dispute, or should be, is the aggregate picture.
The thousands of portfolio companies still marked at 2021 valuations, and the hundreds of funds from the 2018 to 2022 vintages running low on time, are likely to look very different once cash reality catches up with the paperwork.
That adjustment is already visible in the DPI numbers, in secondary-market deals where portfolios traded at 20 to 40 cents on the dollar versus reported value through 2024, and in a fundraising drought that is unlikely to break until exits resume at scale.
The reckoning is not on the horizon; it is already underway.
It is simply moving slowly enough—and the incentives to obscure it are strong enough—that many people with money in the system have yet to fully absorb what the data already shows.
The post VC investors’ $200 billion hole: Where did all the money go? appeared first on Invezz
