Is Private Equity Cooked?

Published November 4, 2026, the same day as Sequoia’s The Sequoia of Tomorrow. Commentators paired the two as rival visions of a venture firm: Sequoia narrowing, a16z making a case for a larger share of LPs’ private-markets allocation. See Two Ways to Build a Firm.

Summary

Kha opens with a puzzle: most people agree tech buyouts are in a tough spot, yet LP money keeps flowing into re-ups with the same managers. Her running analogy is housing. Buyout is the good house on the ideal street; venture was the undeveloped lot across town, until recently.

The cycle argument. Tech buyout works best once a technology wave has settled and business models are predictable enough to borrow against; venture shines when a new wave drops. Top-quartile venture beat top-quartile buyout through the internet build-out and again through mobile and cloud, but trailed for the roughly 1998–2005 vintages. She believes we are at the start of the next wave, and warns against trying to time it.

How buyout got here. Early buyouts were lumpy and venture-like (Wesray’s 1982 Gibson Greetings deal: ~$1M of equity on an ~$80M deal, public 16 months later at ~$290M). The second wave professionalised family businesses. Forty years on, the “fixer-uppers” are gone: in 2025 about half of US PE exits by count were sales to another sponsor. The 2010s were the golden years, with falling rates and rising multiples that Bain says powered over half of buyout returns, and buyout AUM growing from $1.6T (2015) to $3.9T (2021). SaaS was ideal for leverage, but venture had funded it, unlevered, first.

The numbers she leans on:

• First half of 2026: venture funds marked tech holdings up 16%; buyout funds marked theirs down 4% (MSCI).
• PE-backed middle-market companies carry 4.7–6.0x adjusted EBITDA in net debt versus ~2x for the median S&P 500 company.
• Ten-year returns: US buyout 14.5% a year, behind the Russell 3000; US venture 17.6%, 2.3 points ahead, unlevered.
• Exits now average ~7 years, up from 5–6 in 2010–21.
• Top 5% of 2007–16 vintages: venture 5.3x DPI versus buyout 3.5x; roughly 2.2x better TVPI for venture. Venture track records persist; buyout top-quartile has not reliably predicted the next fund since 2000.
• 2021–22 deals: loan-to-value for PIK borrowers went from 44% to 75% in a year; 70% of debt foreclosed in 2025–26 came from 2021 or 2022 deals; PE firms backed 54% of the largest US bankruptcies in 2025. Her case study is Medallia, bought by Thoma Bravo in 2021 for ~$6.4B and handed to lenders led by Blackstone this year, wiping out ~$5.1B of equity.
• Marks: venture marks rise when outside investors pay a higher price; PE owners appraise their own house. Continuation vehicles were 14% of PE exit value in H1 2026 (5% in 2020), and NAV loans add a second layer of debt.
• About $140B of risky PE-backed loans mature in 2028 and $160B in 2029, against a ~$60B average two years out in 2023–25; about a third is software.
• “12 is the new 5”: a typical 2015 buyout needed ~5% annual EBITDA growth to make 2.5x in five years; today the same deal needs 10–12%. Golub’s software borrowers slowed from 9% revenue growth to under 3% in two years, while venture companies grow a median 18–30%.

Her conclusion. PE is not cooked overnight; capital is sticky, good managers will adapt, and smaller buyouts outside levered software can do fine. But the conditions of the 2010s are gone. Her advice: ask your manager how much interest is being added to the loan (PIK) on 2021–22 vintages, and find an exit. Moving into private credit or secondaries just holds the same companies with less upside. That leaves venture, but only the top 5% of managers: “why are you in the median?” She notes Anthropic’s $65B round in May exceeded the largest buyout on record (the $56.6B Electronic Arts take-private), and closes on a quote from her partner Martin Casado. The piece ends with a16z’s standard legal disclaimer.

Notable quotes

“Tech buyout is a late-cycle asset class.”

“Every bet we make is essentially an out-of-the-money call option, and we’re okay with the zeroes.”

“As they say, more fiction has been written in Excel than in Word.”

“If you’re going to be in venture, you need to be in the top 5%.”

The response

The piece drew a large response on X; the full archive of 20 quote posts and 25 replies is on [[Untitled tweet [53]… (tweet)]]. The main lines, without adjudicating them:

• “The same flaws apply to venture.” Dan Gray’s widely shared rebuttal (This article makes absolutely no… (tweet)) argued the criticisms (performance, liquidity, markup practices) apply at least as well to modern venture, that comparing top-5% funds uses the Power Law to obscure very different return distributions, that multistage firms have changed too much for persistence data to carry over, and that PE is more rigorous on valuation. Others made related points: VC marks also rise on insider-led rounds and lag on the way down (“until it’s DPI, it’s an opinion”), and the top-5% comparison is apples to oranges when top VC funds are closed to new LPs.
• “Talking your book.” Several replies read it as a pitch for allocation share: Gray suggested a16z wants a bigger chunk of PE allocation as it moves away from pure venture, and others called it “inter-wallet competition” between VC and PE managers for institutional capital.
• Portfolio-construction pushback. A former buyout investor replied that LPs are not choosing either/or; buyout capital formation dwarfs venture, and the top 5% of VC lacks the capacity to absorb that capital. He also disputed the closing suggestion that PE is mostly about cutting jobs. A long-time investor said cash returned, not marks, is the real scoreboard.
• Agreement on the software math. Many agreed software buyouts are under real strain. The 12% EBITDA growth bar was the number readers kept returning to; Will Manidis said scaled PE platforms broadly understand that what is coming is fundamentally different and want help with it; others predicted the next PE vintage moves to main-street roll-ups, regulated moats, and AI-driven turnarounds. One framing that circulated: PE is short volatility, venture is long volatility.
• Context from the buyout side. The same week, Wall Street Rollup relayed Orlando Bravo’s comments that many PE sponsors bid on a company but are undifferentiated and want minority stakes, so Thoma Bravo is selling to public strategics instead.
• The firm-strategy reading. Ho Nam quoted Gray’s rebuttal to contrast the week’s two posts: Sequoia returning to a flat partnership, a16z “eyeing the AUM and fee streams of PE.” He did not pick a winner, saying only that he believes in simplicity.

Full text

Connections

• The Sequoia of Tomorrow: Sequoia’s memo the same day, read as the opposite bet.
• Two Ways to Build a Firm: Ishaan Prasad reads this essay as a16z making JPMorgan’s bet on breadth, aimed at a larger share of each LP’s allocation.
• Don’t Call It a Comeback: another post-midterm venture-strategy piece from the same week.
• Private Equity · Leveraged Buyout · Enterprise Software Buyouts · Private Credit · DPI · Power Law
• a16z · Jen Kha · Martin Casado · Thoma Bravo · Medallia