There is blood on the streets, and Goldman just bought the mop company.
Its deal to acquire Industry Ventures is an acknowledgment that the crisis is here: late-stage tech and AI portfolios are stuffed with assets that won't clear at book, exits are anemic, and continuation funds have become the market's operating system for triage. Goldman bought flow control over the cleanup.
The acquisition, $665 million at closing plus up to $300 million in earn-outs through 2030 for a $7 billion AUM platform, closing in Q1 2026, represents something more fundamental than a strategic expansion. It's the institutional validation of venture capital's transformation from an equity appreciation game to a duration management business. When the bulge bracket buys the secondary specialist, it's about to monetize the restructuring.
The Continuation Fund Machine: How Venture Built Its Own Toxic Bank
Continuation vehicles started as legitimate portfolio management tools. A GP with one or two breakout companies trapped in a fund nearing expiration could roll them into a new structure, providing LPs with a liquidity choice while preserving upside. The mechanism was elegant: existing investors could elect to cash out or roll, new capital financed the transaction, and the GP reset the economics and timeline. Everyone wins, in theory.
In practice, continuation funds have metastasized into the market's primary mechanism for avoiding price discovery. The numbers tell the story: GP-led secondaries scaled from approximately $9 billion in 2016 to the low-$60 billion range by 2023, then set a new record at $72 billion to $ 75 billion in 2024. Total secondaries reached $152-162 billion in 2024 and a record $103 billion in just the first half of 2025, with GP-led volume at $46-47 billion, up 60% year-over-year. Continuation funds now comprise 78-79% of GP-led secondary volume; they've become the dominant liquidity mechanism.
What started as year 10 exceptions became year 7 playbooks. Single-asset quality plays became multi-asset bundles. Modest liquidity discounts became routine 20-40% haircuts dressed up as "fairness opinions."
The mechanics reveal the game: A continuation fund creates three layers of opacity between current marks and market reality.
First, the GP selects which assets to roll, invariably the ones that can't clear traditional exits at carrying value. Second, the "independent" fairness opinion anchors on the GP's own marks, creating circularity. Third, the secondary buyer's discount is attributed to an "illiquidity premium" rather than overvaluation, thereby preserving the fiction upstream.
Consider the typical continuation fund sequence: A 2019 vintage fund sits at 0.4× DPI in year six. The flagship Series B investment, valued at 15x its entry, has no path to an IPO at current multiples. The GP proposes a continuation vehicle at a 25% discount to the last mark, still 11× the original basis. Legacy LPs, exhausted by capital calls and desperate for distributions, take the cash. New investors, seeking "discounted" exposure to a branded asset, provide fresh capital. The GP resets carry, extends duration, and maintains fee streams on both the old fund and the new vehicle.
The result? The same impaired asset, repackaged with fresh documentation and a new clock, generating fees for another seven years. The valuation problem hasn't been solved; it's been deferred and monetized.
This deferral infrastructure now processes hundreds of billions in venture NAV. This is the operating system. And Goldman has just acquired the largest operating system company.
Industry Ventures: The Infrastructure Play Goldman Actually Bought
Industry Ventures positions itself as a "complete lifecycle" venture platform, but it's not a venture capital firm in any classical sense. It doesn't pick founders or build companies. It's a liquidity engineering platform that turns duration mismatches into fee streams. The firm operates three interconnected businesses that, together, form the plumbing through which distressed venture assets will flow for the next decade.
The primary business is secondary fund management, buying LP positions at discounts and holding to maturity. This requires deep coverage of hundreds of venture funds, pricing models for illiquid assets, and relationships with sellers ranging from university endowments to family offices. Industry Ventures has built proprietary datasets on fund performance, portfolio company trajectories, and LP behavior patterns that allow it to price complexity others won't touch.
The second business is GP-led solutions, structuring and leading continuation vehicles for venture funds. This means originating transactions, running valuation processes, securing fairness opinions, syndicating to co-investors, and managing documentation. It's operationally intensive, relationship-dependent, and generates fees at multiple touchpoints: advisory on structuring, placement for syndication, and management on the resulting vehicles. The firm's public materials and research emphasize GP-led secondaries as a core liquidity tool.
The third business, often overlooked, is the data and distribution network. By touching hundreds of venture funds and thousands of LPs, Industry Ventures maintains real-time intelligence on where assets are stuck, which LPs need liquidity, and what prices actually clear. This information asymmetry is worth more than the returns of any single fund.
Goldman is buying the machine that processes venture's distressed inventory. Every continuation fund needs structuring, fairness opinions, syndication, and ongoing administration, often involving seven-figure advisory and process costs. Every secondary transaction needs pricing, documentation, and settlement. Industry Ventures has industrialized these processes at scale.
The strategic logic is compelling and impressively deft.
Goldman gets fee durability through locked-up continuation vehicles, distribution leverage through its private wealth channels hungry for "discounted" tech exposure, and information advantage through visibility into venture's entire secondary market.
The bank doesn't need venture capital to recover; it needs venture to keep restructuring.
The AI Unwind: When Hundreds of Billions Meet Reality
The mathematics of the venture capital AI bubble is non-negotiable. As I detailed in Capturing Dislocation, cumulative AI and AI-adjacent funding since 2020 exceeds $400 billion, with H1 2025 alone seeing $116.1 billion in AI funding. These investments were made at 25-30× revenue multiples, with many companies pre-revenue entirely. Public market throughput remains far short of what would be required to clear this backlog, making the system's reliance on GP-led restructurings inevitable.
The deployment-to-exit imbalance creates structural insolvency. Fund lives are capped at 10-12 years. The 2020-2021 vintage funds that deployed most aggressively into AI are already in year five. By year seven, LPs expect meaningful distributions. By year ten, everything must be liquidated. The clock is non-negotiable, and the math doesn't work.
The concentration makes it worse. OpenAI, Anthropic, Databricks, and xAI absorbed a disproportionate share of capital at valuations that assume TAMs larger than global GDP. The long tail of AI startups, the computer vision companies, the vertical SaaS plays with "AI" appended, the infrastructure picks-and-shovels, are marked at multiples that require these giants to succeed at scale.
Secondary markets are already pricing in the correction. Marquee AI names see varying marks in private transactions, some tenders at step-down valuations, others at step-ups, depending on share class and timing. Smaller names don't trade at all, there are no bids. The markdown cascade hasn't begun in earnest because GPs still control marks, but the secondary market is the truth-telling mechanism.
According to a University of Chicago study, 80-90% of LPs offered the chance to roll into continuation vehicles choose to cash out instead. They're accepting significant discounts, typically 20-40% from last round valuations, to escape. When the sellers outnumber the believers by such margins, it suggests something beyond normal market dynamics.
The unwind will follow a predictable pattern. First, the marked-to-market investors (crossover funds, mutual funds) will recognize losses, creating precedent prices. Second, venture funds will begin to offer selective markdowns on clearly impaired assets while defending their "core" positions. Third, as fund expiration approaches, continuation vehicles will proliferate to defer recognition. Fourth, LPs will revolt, demanding cash at any price. Fifth, mass liquidation through secondaries, creating a buyer's market for patient capital.
Goldman's timing is perfect. It's like buying the toll booth just as traffic is about to explode.
The Ten-Year Fee Extraction Machine
Goldman's acquisition economics are elegant and shrewd in their cynicism. The bank is betting on process fees from a decade of restructuring. The model is credit-like in its predictability, with equity optionality in the event of miracles.
Start with base management fees. Continuation vehicles typically run for 7-10 years, with annual management fees of 1.5-2% on committed capital. A $1 billion continuation fund generates $15-20 million annually in fees regardless of performance. Industry Ventures' $7 billion AUM generates over $ 100 million in annual management fees at its current scale. As continuation volume grows, and it must, given the exit mathematics, AUM compounds.
Layer on transaction fees. Every continuation vehicle requires extensive advisory and process work, often seven-figure costs for fairness opinions, structuring advice, and syndication placement. A $500 million continuation fund can generate $10-20 million in transaction fees across these services. Industry Ventures touches dozens of these deals annually. Goldman's infrastructure can further industrialize this.
Add distribution margin. Goldman's private wealth and family office channels desperately seek tech exposure but fear venture's J-curve and illiquidity. Continuation funds offer "discounted" access to branded assets with shorter duration. Goldman can package these positions into structured products, capturing another 2-3% margin on distribution.
Include NAV lending. Continuation funds often use leverage to juice returns. Goldman can provide the debt at high-single to low-double-digit coupons on loans secured by "discounted" venture assets. It's picks-and-shovels economics: lend against the gold mine, don't dig for gold.
Consider the information arbitrage. Seeing every continuation fund, every secondary trade, every distressed seller provides unparalleled market intelligence. Goldman can position its own balance sheet, advise corporate clients on acquisitions, and guide LPs on portfolio construction with perfect information asymmetry.
The beauty is that this model works regardless of venture outcomes. If AI recovers and exits reopen, Goldman captures performance fees and early liquidity. If valuations crater, the restructuring volume and fee velocity accelerate. The mop gets paid by the gallon, not by the stain's ultimate fate.
The Venture Capital 2.0 Playbook: Control of Our Destiny
The Goldman acquisition clarifies the battlefield. The old venture model of 10-year blind pools betting on 100× outcomes through equity appreciation is broken. The new reality presents two paths: becoming part of the restructuring infrastructure (Goldman's choice) or building the next asset class.
I am choosing the latter.
The playbook starts with instrument innovation. Traditional venture equity is a binary bet: worthless or wonderful. In a world where exits are constrained and duration matters, this binary structure destroys value. Revenue-linked instruments like Safer (Simple Agreement for Future Equity or Revenue) create continuous return paths. A Safer investment might take 2-3× return through revenue share over 5 years, with residual equity upside if exits materialize. This transforms venture investing from a lottery ticket to a yield asset with optionality.
Capital structure matters more than entry price. A $10 million investment at a $100 million valuation that never exits is worthless. A $10 million investment at a $200 million valuation with 15% revenue share and buyback rights compounds value regardless of exit. The higher entry price with structured terms beats the lower price with binary outcomes.
Permanent Capital Vehicles (PCVs) align with this reality. Instead of 10-year funds racing against expiration, PCVs operate indefinitely with dividend policies. Investors receive a yield from revenue-linked instruments while maintaining exposure to the upside. The vehicle can hold positions for decades if necessary, eliminating the need for forced selling into continuation funds. This patience arbitrage is worth 20-30% of returns.
Operating leverage changes unit economics. Studios that amortize fixed costs, such as GPU clusters, data pipelines, compliance infrastructure, and go-to-market systems, across a portfolio can build companies at 30-40% of the standalone cost. When the same technical team, infrastructure, and playbooks get reused, the equity capital required drops dramatically. This efficiency allows profitable building even at compressed valuations.
The measurement framework must shift from marks to cash. Report DPI quarterly, not annual TVPI. Show revenue coverage of invested capital, not paper multiples. Track cash yield on deployed capital, not unrealized appreciation. LPs are learning that 1.5× DPI beats 3× TVPI every time.
Geographic arbitrage accelerates this transition. While Silicon Valley funds continue to pile into continuation vehicles to maintain Bay Area positions, emerging ecosystems offer more favorable economics. The same AI engineer costs 70% less in Toronto, 60% less in Austin, 50% less in Prague. The same revenue-linked deal, at 40% of Silicon Valley valuations, pencils out to higher cash returns.
There is nothing we can do to change the fate of venture capital. But there is plenty we can do to transform innovation investing. While banks extract fees from restructuring the old guard, Venture Capital 2.0 operators construct the new asset class: innovation finance that behaves like credit, compounds like private equity, and maintains venture optionality.
The Market We Have, Not the Market We Want
Two truths will define venture capital in the next decade. First, continuation funds are now essential infrastructure, not temporary bridges. The math of deployment versus exit capacity makes this non-negotiable. Second, continuation funds don't create value; they redistribute it from LPs to GPs and banks while deferring recognition.
Goldman's acquisition validates both truths. The bank witnessed the transformation of the venture capital industry from equity appreciation to duration management and acquired the core infrastructure. It's a brilliant trade for Goldman—fees are the only reliable source of return when underlying assets are impaired.
The evidence is compelling. NEA's $540 million continuation vehicle from July 2024 rolled eleven portfolio companies, led by Databricks at a $46 billion internal valuation, at approximately 19% discount to NEA's marks. Trinity Ventures' $435 million CV gave fifteen assets from a 2012 vintage fund, 13 years old, a second life. General Catalyst explored an $800 million to $1 billion CV, including Stripe, Gusto, and Circle. These aren't distressed assets or pandemic casualties. They're among venture's most successful companies, yet even they're being routed through continuation vehicles rather than traditional exits.
For everyone else, the message is clear. The venture capital industry has entered its structured products phase. The winners will be those who intermediate duration and distribute paper, the banks, the secondary funds, the continuation specialists. The losers will be founders, employees, and LPs holding marked-up equity waiting for exits that won't come at carrying value.
The blood is on the street. Goldman now owns the mop company. In the structured product era of venture capital, cash flow beats cap tables, duration management beats growth multiples, and the mop company always gets paid.
Upgrade to Venture Capital 2.0 Premium
This article is part of the free public release series from Venture Capital 2.0, examining how the next decade of innovation finance will be rebuilt from the failure of the power law era of venture capital.
Premium members receive:
- 📘 A paperback and digital copy of my book Venture Capital 2.0, sent directly to you
- 📊 Proprietary, in-depth research available only to premium subscribers
- ⚙️ Exclusive tools, frameworks, and guides for building, funding, and scaling venture studios and PCVs
- 💡 Actionable investment strategies grounded in the Venture Capital 2.0 methodology
If you’re serious about mastering what replaces the power-law model—upgrade to Venture Capital 2.0 Premium!