Key Takeaways

  • The tech industry generated roughly 20 decacorns per decade over the past 50 years, but produced 42 decacorns in 2026 alone.
  • Late-stage private valuations are propped up by liquidation preferences, allowing investors to price deals as 1x downside protection plus high-volatility call options.
  • Jason Lemkin argues that paying extreme prices is necessary venture math, pointing to Andreessen Horowitz's strategy of absorbing overpriced rounds to own the ultimate winners.
  • While liquidation preferences in 2021 SaaS unicorns proved worthless when growth stalled, Lemkin contends that top-tier AI liquidation preferences hold durable value.

The Disagreement

Rory O'Driscoll looked at the historical record and saw impossible math. Over the past half-century, public and private markets produced about 20 companies worth ten billion dollars every ten years. In 2026, the AI sector minted 42 in under twelve months.

“We roughly generated 20 plus or minus decacorns per decade for the last five decades,” O'Driscoll said. “And we generated what did you say, 42 decacorns in one, in less than a full year of this decade.”

To O'Driscoll, the driver behind these numbers is financial engineering, not sudden enterprise productivity. Late-stage venture investors do not purchase common stock on an open exchange where assets mark to market daily. They buy preferred shares protected by senior liquidation preferences. “Once you have a lick pref and you believe it exists, you can say to yourself, unlike a common stock where it trades every day, you can say it's a 1x plus option value,” O'Driscoll said. When market volatility is high, call options become expensive, creating inflated headline valuations that mask the investor's downside floor.

Jason Lemkin pushed back, arguing that venture capital math requires backing the outliers regardless of the entry price. “That's what Andre got so right,” Lemkin said, pointing to the early strategy of Andreessen Horowitz. “People mocked Andre in the early days for overpaying and doing this wrong and saying we have to be in all the winners, and they just did the math. Yes, you have to be in all the winners, and even a lot of the winners you won't make money in because you'll overpay, but the super winners will make up for all of it.”

Lemkin also rejected the idea that today's preferences mirror the tech runup of five years ago. “Today it feels like a liquidation preference in Jev is much more valuable than a thousand liquidation preferences in a thousand 2021 unicorns,” Lemkin noted. “Those are worthless. Many of those will end up being worthless.”

Who's Right (and When They're Wrong)

O'Driscoll is right on the macroeconomic scale. Public market demand cannot absorb 42 new ten-billion-dollar companies every year. Corporate software budgets are not expanding fast enough to justify dozens of simultaneous decacorns. When dozens of private startups claim ten-figure marks within months, most of those valuations are paper constructs built on liquidation preference floors. The preference protects the investor's principal, but common shareholders absorb the damage when the option expires worthless.

Lemkin is right about fund-level venture mechanics. For a multi-billion-dollar fund, failing to own the single company that captures the market destroys the partnership. Overpaying for four failed decacorns does not matter if the fifth company hits a two-hundred-billion-dollar market capitalization. The super-winner pays for every round where the firm misjudged the price.

Where Lemkin's logic breaks down is for founders and employees. A venture capitalist can justify paying a ten-billion-dollar valuation because their 1x preference protects their capital if the company sells for two billion. The founder and the early team do not have that protection. When late investors stack hundreds of millions of dollars in preferred claims above common equity, the team's shares sit underwater until an exit clears the entire preference stack.

What to Do With This

Pull your cap table and model your liquidation waterfall this week. Add up the total dollar value of senior preferred stock sitting ahead of your common shares. Calculate the exact exit price your company must achieve before common options receive a single dollar, and share that hurdle number with every executive holding incentive equity.