E73: Late-stage VC markdowns and mistakes, market strategy, Ukraine/Russia update with Brad Gerstner

Chamath said on March 26, 2022 that “the markets are mostly moving upwards for the short term” and then “volatility is going to come back.”

Looking at broad U.S. equity indices after that date:

• The S&P 500 weekly closes show 4,543 on March 25, 2022 (the day before the podcast release) and then 4,545 on April 1, followed by a steady slide to 4,488 (Apr 8), 4,393 (Apr 15), 4,272 (Apr 22), and 4,132 (Apr 29), continuing downward into May and June. This is a clear downtrend starting almost immediately after late March, not a period where markets were “mostly moving upwards.” (fedprimerate.com)
• On a monthly basis, the S&P 500 returned +3.83% in March 2022, but then -8.99% in April, roughly flat in May (+0.04%), and -8.78% in June. (statmuse.com) Over the “next few weeks to a few months” after late March, the index was predominantly down or flat.
• The tech‑heavy Nasdaq did even worse, with monthly returns of +3.67% in March but then -13.56% in April, -2.03% in May, and -9.43% in June 2022, reflecting a sharp and sustained selloff rather than a short‑term upward trend. (statmuse.com)
• A Business Insider summary of the March 2022 “death cross” notes that after that signal in March 2022, the S&P 500 went on to fall about 16% and the Nasdaq about 27%, underscoring that the subsequent move was decisively downward. (businessinsider.com)

On volatility, the Cboe VIX Index remained elevated as the market slid into a >20% drawdown by June 30, 2022, when the S&P 500 was down about 20.5% over six months and the VIX closed at 28.71, well above its long‑run average. (cboe.com) But this was continuation of high volatility rather than a calm “upward” period followed by volatility “coming back.”

Because the core, time‑bounded part of his prediction—that broad equity markets would mostly trend upward in the immediate short term (weeks to a few months) starting from late March 2022—was not borne out (they mostly went down and entered a bear market instead), the prediction is best classified as wrong, even though volatility indeed remained or became elevated.

Evidence from 2022–2023 shows that late‑stage private SaaS/software companies experienced exactly the valuation compression and financing difficulty Chamath anticipated.

• Sharp valuation markdowns in SaaS: TechCrunch reported that SaaS valuations fell ~75% from October 2021 to October 2022, creating a “perfect storm for companies needing follow‑on funding,” as exits collapsed and investors reassessed richly valued SaaS names. (techcrunch.com) An analysis of SaaS valuations similarly notes that companies which previously raised at 100x ARR might have to refinance near 20x ARR, with late‑stage investors “shunning the late-stage private financing market” because prior rounds were too expensive. (paddle.com)

• Late‑stage software/SaaS specifically flagged as most at risk: In April 2022, TechCrunch highlighted that late‑stage software/SaaS startups were “in the most valuation trouble,” as their revenue multiples had been bid up the most in 2020–2021 and were now retracing in line with public software stocks. (techcrunch.com) This directly matches Chamath’s concern that the repricing of public SaaS would spill over into late‑stage private SaaS.

• Severe late‑stage valuation resets in 2022: Carta’s 2022 data show that by Q4 2022, median valuations for late‑stage rounds had collapsed: Series D valuations were down 58% year‑over‑year and Series E+ down 72%. The median Series D round size was 79% smaller than a year earlier, and Carta describes late‑stage fundraising as “difficult,” with far fewer, much smaller deals at much lower valuations. (carta.com)

• Ongoing late‑stage funding trouble in 2023: Carta’s Q1 2023 update shows median Series D valuations down ~70% year‑over‑year and Series E+ down 82%, calling it “a difficult start to 2023 for late-stage startups hunting for new venture capital.” (carta.com) TechCrunch likewise described the “late‑stage venture market” as “crumbling,” noting median valuations for Series D+ rounds had fallen about 60% and that late‑stage deal counts were plunging. (techcrunch.com)

• Financing environment clearly worse for late‑stage SaaS: A TechCrunch piece focused on SaaS follow‑on rounds observed that venture investment fell 70% in Q4 2022 vs. Q4 2021 and that overvalued SaaS names, combined with a collapse in exits, were particularly hurting companies trying to raise extension and follow‑on rounds. (techcrunch.com) That is precisely the “trouble” Chamath was pointing to: not just paper markdowns, but real difficulty raising capital at prior valuations.

Across multiple independent datasets (Carta, Crunchbase, SVB) and SaaS‑specific analyses, late‑stage private SaaS/software companies saw large valuation markdowns, a spike in down/flat rounds, and a significant pullback in late‑stage capital in 2022–2023. This matches the substance and timing of Chamath’s prediction, so it is best scored as right.

The core claim was that among the big late‑stage, VC‑backed companies waiting to go public in early 2022, some would ultimately IPO at valuations below their last private round (down‑round IPOs) during the next IPO cycle (2022–2023).

That happened clearly:

• Instacart (Maplebear) was one of the marquee late‑stage unicorns in the 2021–2022 IPO backlog; CNBC in March 2021 noted its valuation had just doubled to $39B and that it was preparing for a “highly anticipated public debut,” with Goldman Sachs expected to lead the IPO. (cnbc.com) When Instacart finally went public in September 2023, its IPO priced at $30 per share, implying a fully diluted valuation of about $9.9B, a steep drop from the $39B private round two years earlier. (fortune.com) Axios explicitly labeled this a “down‑round IPO,” noting Instacart raised $660M at a $9.9B valuation, sharply below the $39B valuation of its March 2021 funding round. (axios.com) This single, very prominent example already satisfies “some of them are probably going to have to IPO at down rounds.”

• The broader pattern shows a meaningful subset, not just one outlier. For example, Arm’s 2023 IPO valued it at $54.5B, below an internal $64B valuation SoftBank had assigned shortly before, which market commentary described as a down‑round IPO alongside Instacart’s sharply reduced range of $8.6–$9.3B vs. its prior $39B private valuation. (nasdaq.com) Klaviyo, another late‑stage startup in that cohort, went public in September 2023 at about $9.2B, roughly in line with or slightly below its prior late‑stage valuation around $9.5B, and was reported as initially targeting an IPO valuation below its last private round. (cnbc.com)

• More generally, venture commentators summarized the 2023–2024 IPO window as being dominated by down‑round IPOs: a widely cited analysis by investor Aman Verjee (shared by Dave McClure) noted that “just about every VC‑backed IPO in 2023 and 2024 was a ‘down round’” relative to the prior private financing, explicitly listing Instacart, Reddit, Klaviyo, ServiceTitan, Circle, and Hinge Health as examples. (linkedin.com) This confirms that down‑round IPOs became common among the very cohort of late‑stage, VC‑backed companies Sacks was talking about.

Because at least one flagship backlog name (Instacart) and then a broader set of peer companies did IPO at valuations below their last private rounds within the ensuing IPO cycle (by late 2023), Sacks’s prediction that “some of them are probably going to have to IPO at down rounds” is borne out by subsequent events.

Chamath’s claim had two parts: (1) late‑stage private tech names like Gopuff, Canva, Klarna, Discord, Ripple, Grammarly “have to” be marked down ~15–40% immediately as rates rise, and (2) if they keep growing strongly, they can recover to prior peak valuations within about 18 months of that markdown.

What actually happened with the concrete examples he cited:

• Klarna – Peak valuation was about $45.6B in June 2021. By July 2022 it raised at $6.7B, an ~85% markdown, far beyond the 15–40% range. Even by 2025 its IPO valuation talk/targets are only around $12.5–15B, still far below the $45.6B peak, years after the markdown. (cnbc.com)
• Gopuff – Valued around $15B in 2021. Subsequent financing in 2025 values it at about $8.5B, a drop of roughly 40–45%, and it has not returned to the prior peak more than three years later. (ft.com)
• Canva – Valued at $40B in 2021. Blackbird Ventures marked its position down ~36% to $25.6B in mid‑2022, and T. Rowe Price later marked its stake down roughly 67.6%, again deeper than Chamath’s 15–40% band. Canva only re‑attained and then exceeded the prior peak via secondary/staff share sales at around $42B in 2025—roughly three to four years after the 2021 peak and about three years after the first markdown, not within 18 months. (startupdaily.net)
• Discord – Raised at a $15B valuation in 2021. By mid‑2023, Fidelity had marked its holdings in Discord down about 47% from cost, with no public evidence that valuations had returned to the 2021 peak within 18 months of the markdown. (techcrunch.com)
• Grammarly – Was valued at $13B in 2021; later coverage still references that 2021 figure, and recent financings are structured as non‑dilutive debt rather than new equity rounds, suggesting no clear equity repricing back to a meaningfully higher valuation. There is also no public indication of a clean 15–40% markdown then full recovery inside 18 months. (reuters.com)
• Ripple – Valued at about $15B in an early‑2022 buyback, then referenced around $11.3B in early 2024, implying a markdown of roughly 25%. Ripple does eventually raise at a $40B valuation in 2025, but that recovery occurs several years after the 2022 valuation event, not within an 18‑month window. (research.contrary.com)

In the broader late‑stage market, mutual‑fund and VC markdowns in 2022–2023 were often significantly steeper than 40%, and many unicorns remained well below 2021 peak valuations through at least late 2023, despite continued revenue growth at a number of them. (business2community.com)

So:

• The size of the required markdowns was generally larger than his 15–40% range for the emblematic names he mentioned.
• The timeline for recovery was materially off: even strong performers like Canva and Ripple only got back to or above their prior peaks years later, while others (Klarna, Gopuff, Discord) still haven’t as of late 2025.

Because both the magnitude and the 18‑month recovery path were meaningfully wrong for the very companies cited as examples, the prediction overall is best classified as wrong.

Evidence since 2022 shows exactly the kind of prolonged, complex unwinding of rate‑ and liquidity‑driven distortions that Chamath predicted.

From 2022 onward, major central banks reversed more than a decade of ultra‑low rates and QE. The Fed raised its policy rate from near zero in early 2022 to about 5.25–5.50% by July 2023 and began quantitative tightening in June 2022, shrinking its balance sheet by over $2 trillion from a ~$9T peak. A UN analysis explicitly describes this as an unwinding of QE that injected trillions in liquidity and notes that this process is proving particularly challenging.(twn.my)

This policy shift precipitated a broad 2022 bear market: global equity indices fell sharply, with the S&P 500 down 19% and the Nasdaq down 33% for the year, driven by inflation and aggressive rate hikes. Commentators widely framed this as the popping of an “everything bubble” inflated by years of easy money.(en.wikipedia.org)

The adjustment has clearly been multi‑year rather than a brief shock. Late‑stage venture and private tech have gone through sustained markdowns and funding contraction: late‑stage VC deal volume and capital in 2023 were far below the 2021 peak and even below 2019 levels. Global venture funding in 2023 fell about 42% year‑on‑year, with growth‑stage valuations described as “down massively” from 2021; major crossover investors like Tiger Global and SoftBank slashed activity and marked down portfolios (Tiger cutting startup values by roughly a third, SoftBank reporting multi‑billion‑dollar write‑downs).(carta.com)

Related markets that epitomized the bubble period also went through extended unwinds. The SPAC boom of 2020–2021 collapsed in 2022–2023, with SPAC IPO counts and proceeds dropping by an order of magnitude, while IPO activity and exits overall fell to their lowest levels since the financial crisis and, even by 2024–2025, remained far below 2021 despite a partial recovery.(redbud.vc)

Real‑asset markets show a similarly drawn‑out adjustment: rapid Fed hikes doubled U.S. mortgage rates and severely distorted the housing market, while commercial property in Europe has logged multiple consecutive years of price declines linked to higher rates and financing stress, still falling in 2024.(forbes.com)

By late 2025, the Fed is only just preparing to stop QT after reducing its balance sheet to about $6.6–$6.7T, with analysts noting that the pandemic‑era “excess liquidity” in facilities like the reverse repo has largely been drained.(reuters.com) This ongoing policy normalization and the multi‑year repricing in public, private, and real‑asset markets fit Chamath’s description of “the beginning of a very complicated process of unwinding the distortion” created in the prior few years of ultra‑easy money.

Given this evidence, the prediction that 2022 would mark the start of a prolonged, complex unwinding of those distortions is best judged as right.