The Exit Illusion

The same regime shift that destabilized public market correlations has also broken the private market’s exit architecture.

The 2025 IPO narrative was compelling: pent-up supply would meet eager demand, and the great exit engine of venture capital would roar back to life. Approximately one-third of 2025’s IPOs fell below their issue price within months ¹². The median new listing declined 9% while the S&P 500 gained 18%. The market was punishing new listings, not rewarding them.

Figma became the exit illusion’s defining case. Its June 2025 IPO was pitched as validation that the market would pay premium multiples for high-quality software businesses. The stock surged 2.5x on its first day. Investors had marked Figma against Adobe’s $20 billion acquisition price, blocked by regulators in late 2023. Even that surge was a partial recovery of marks, not a realized gain.

Then came earnings. Figma’s stock crashed 60% ¹³, erasing nearly all its post-IPO gains. Operational performance was fine. The threat was structural: AI-native design tools were commoditizing the collaborative workflows that had underpinned Figma’s valuation premium. This was not narrow-float price discovery. It was the market re-rating a business whose moat had been narrowed by the same technology wave its IPO tried to ride. By April 2026, Figma was trading more than 80% below its IPO price ²⁹.

Figma was not alone. Navan, the travel and expense management company, priced at $6.2 billion, roughly one-third below its peak private valuation of $9.2 billion. It promptly fell below issue price, trading at $2.7 billion as of early March ¹⁴. Gemini, the cryptocurrency exchange, dropped 58% from its IPO price ¹⁵. The “open” window delivered volume without value.

The traditional venture model assumed a predictable progression: raise capital, deploy into startups, scale the winners, exit via IPO or M&A, return capital to LPs, repeat. Each stage depended on the next. When exits freeze, the entire system backs up. New investments slow because old investments haven’t returned capital. Valuations become fictional because there is no market to test them.

This is the industry’s structural problem. For the disciplined investor who has not over-deployed into the cycle, it is also an asymmetric opportunity: every dollar of institutional capital frozen in illiquid positions is a dollar unavailable to compete at the next dislocation.

Global private equity and venture capital dry powder stands at $4.6 trillion, a record according to PitchBook ¹⁶. Most sits in two-to-five-year-old funds, unable to find productive deployment. Abundance and paralysis coexist. That is the liquidity paradox.

Into the vacuum steps the secondary market. In 2024, secondary transaction volume reached $160 billion. In 2025, it hit $240 billion (up 48% year-over-year), and Jefferies forecasts annual volume will approach $300 billion over the next 12 to 24 months ²². Wellington Capital predicts that 2026 will see secondaries “increasingly become a core liquidity tool” rather than an exceptional measure ¹⁸.

The shift is structural, not cyclical. Stripe, arguably one of the most successful private fintech companies in history, chose a $91.5 billion tender offer over an IPO ¹⁹, and a year later came back with another tender at $159 billion ²³, a 74% jump in valuation without ever touching the public markets. Ramp, Revolut and others followed suit. The public markets are not offering terms that the best companies will accept.

Yet secondaries have their own limits. Only 2% of unicorn market value trades on secondary markets, massive underpenetration. Prices often reflect desperation rather than value, with sellers accepting steep discounts for any liquidity at all. The aesthetic has shifted: buyers are hunting bargains, not growth stories.

The exit illusion exposes a deeper problem: the venture capital model depends on a small number of outsized exits to cover losses across the portfolio. When exits are delayed, diminished, or denied, the math breaks.

A venture fund typically has a 10-year life. Extensions are possible but increasingly common, and costly to LP returns. A company funded in Year 2 needs to exit by Year 8 or 9 to leave time for distributions. If IPO windows are unpredictable and M&A is constrained by regulatory scrutiny, what is the exit path?

There is, increasingly, none. Venture capital will not disappear. Capital will always chase high-risk, high-reward opportunities. But the current model of massive fund sizes and assumed IPO exits is under stress.

Yale and peer endowments sold private equity holdings at steep discounts in 2024, not because the investments had failed but because illiquidity created a cash crunch at the worst possible moment ²⁰. The average university endowment now holds 56% of its portfolio in alternatives ²¹.

Founders must choose between accepting terms that would once have been unacceptable and waiting indefinitely for a window that may not open. LPs absorb extended hold periods and compressed IRRs. For GPs, the hardest conversations are now routine: write-downs, fund extensions, the quiet revision of what success looks like.

The manager who treats current dysfunction as a temporary weather event, to be waited out until rates fall and the IPO window reopens, may be waiting for a climate event that will not arrive anytime soon.

The illusion got worse, not better, in the weeks after publication. Q1 2026 produced $9.4 billion of IPO proceeds across 22 deals, the strongest first quarter in five years by volume. Seven of the top ten deals traded below their issue price within a month, with a median decline of 28% ²⁴. The iShares Expanded Tech-Software ETF fell 25% ³³ on a single thesis: agentic AI tools are breaking the per-seat SaaS licensing model that underwrites most software valuations. The “window delivered volume without value” pattern widened from a handful of names to the category.

The endowment pressure named in the piece sharpened. Yale moved to sell up to $6 billion of private equity interests, roughly 15% of its $41.4 billion portfolio and its first-ever sale of private assets, advised by Evercore ²⁵. The university simultaneously announced a “far more constrained” FY2026 budget. Harvard sold roughly $1 billion of PE stakes and issued $750 million of bonds as a liquidity cushion ²⁶. Both institutions accepted discounts of 10 to 20% below net asset value. Illiquidity forced the timing; the math did not improve.

Stripe came back. In February 2026, the company announced a second tender offer at a $159 billion valuation ²³, up 74% from the $91.5 billion cited above, and still without an IPO. Payment volume had reached $1.9 trillion in 2025 ³¹. Meanwhile, OpenAI raised a single $110 billion venture round in the same month, roughly one-third of all 2025 venture capital by itself ²⁷. The top 25 VC firms now control $645 billion in AUM and over $130 billion in dry powder ³². The liquidity paradox has become a concentration paradox: capital exists in record volume, and it exists in fewer hands, with fewer destinations.

The AI commoditization thesis around Figma played out in public view. On April 14, 2026, news of a new Anthropic design tool sent Figma down sharply ²⁸; the stock closed the YTD window 45.9% lower and more than 80% below its IPO. Chime, which had listed in June 2025, traded 28% below issue ³⁴ as concerns about digital banks without clear profit paths spread. Platform biotech IPOs, Eikon and Generate, priced below their peak private valuations and continue to trade below issue ³⁰. The pattern from Navan and Gemini is now the category. Distributions remain sluggish. The climate has not changed.