E114: Markets update: whipsaw macro picture, big tech, startup mass extinction event, VC reckoning

The prediction specified that in the 30–90 days after early February 2023, equity markets would experience a pronounced “pain trade” move up, similar to the strong, post‑Fed‑pivot rally of late‑2018/early‑2019.

In reality, major U.S. equity indexes were essentially flat to slightly down over that window when measured from the time of the call:

• On Feb 3, 2023 (the last close before the Feb 4 episode), the S&P 500 closed at 4,179.76 and the Nasdaq Composite at 12,200.82. (nasdaq.com)
• By May 5, 2023 (about 90 days later), the S&P 500 was 4,136.25 and the Nasdaq 12,235.41—roughly −1% for the S&P and +0.3% for the Nasdaq versus Feb 3. (wellergroupllc.com)

Looking at the month-by-month path confirms there was no large, sustained upward move akin to 2018–2019:

• The S&P 500 returned −2.5% in February, then +3.7% in March, +1.6% in April, and only +0.3% in May 2023; from Feb’s start through end‑May it effectively moved sideways around the 4,000–4,200 range rather than breaking notably higher. (statmuse.com)
• The Nasdaq Composite posted −1.0% in February, +6.76% in March, +0.66% in April, and +5.94% in May 2023; but from Feb 3 to early May it was still only marginally above its starting level, oscillating between roughly 11,500 and low‑12,000s. (statmuse.com)
• S&P Dow Jones’ own commentaries describe February as a decline and May as a small, range‑bound gain, not a sharp, capitulation‑style rally. (spglobal.com)

By contrast, the end‑2018/early‑2019 episode Chamath referenced featured a double‑digit rebound: the S&P 500 gained about 13% in Q1 2019, its best first quarter in over 20 years, and roughly 14% for the quarter overall, a clear, powerful rally off the prior lows. (indexologyblog.com) Nothing comparable in magnitude occurred in the 30–90 days after Feb 4, 2023; the large 2023 tech‑led surge really accelerated later (e.g., in June), outside his stated window. (statmuse.com)

Because broad equity markets did not move significantly higher over that specific 30–90 day period—and certainly not in a way that resembled the strong 2018–2019 head‑fake rally—the prediction that “the pain trade is to go up” in that window is best classified as wrong.

The prediction is framed as a relative probability update: that, as of early February 2023, a U.S. soft landing in calendar year 2023 was "quite a bit" more likely than it had seemed a few weeks earlier, based on the then-current labor data.

What actually happened in 2023:

• The U.S. did not enter a recession in 2023; the NBER’s Business Cycle Dating Committee still dates the last recession as ending in April 2020, with no new recession peak or trough after that. (nber.org)
• Real U.S. GDP grew about 2.5% in 2023, up from 1.9% in 2022, according to the BEA’s third estimate. (bea.gov)
• The labor market remained strong, with unemployment averaging about 3.6% in 2023 and employment at record levels. (en.wikipedia.org)
• Many retrospective analyses describe 2023 as a year in which the U.S. economy avoided the widely forecast recession and, in effect, achieved or approached a soft landing (growth continued while inflation fell). (cnbc.com)

However, the prediction we are scoring is not "there will be a soft landing in 2023"; it is the narrower statement that the odds of a 2023 soft landing were higher in early February 2023 than a month earlier. That is a claim about how underlying probabilities (or at least informed beliefs) changed over a one‑month window.

On that narrower question, the evidence is mixed and subjective:

• Late‑2022 economist surveys still put high odds on a 2023 recession (e.g., one Bankrate survey found an average estimate of a 64% chance of recession in 2023). (bankrate.com)
• Around February 2023, commentary was divided: some analysts argued a soft landing was still possible, while others said the "door to a soft landing is rapidly closing" as inflation data stayed high and the Fed signaled further tightening. (cnbc.com)

Because we have no direct, objective measure of how the true probability of a soft landing changed between early January and early February 2023—and expert views at the time were not uniformly shifting in one clear direction—the claim that the odds were "quite a bit better" than a month earlier cannot be cleanly classified as right or wrong, even though the eventual outcome (no 2023 recession with continued growth) is consistent with a soft landing.

Given that 2023 is over (so it’s not "too early"), but the prediction concerns an inherently unobservable probability shift over a short horizon, the fairest judgment is:

• Result: ambiguous – the statement about improved odds cannot be definitively verified or falsified from available data, despite the favorable eventual outcome.

Evaluating the three parts of the prediction against what actually happened:

1. Peak level and number of hikes. At the time of the podcast (early Feb 2023), the federal funds target range was 4.50–4.75%. (ycharts.com) Sacks said the curve implied “maybe one more” 25 bp hike and a peak around 4.75–5.0%. In reality, the Fed hiked three more times: to 4.75–5.00% in March 2023, 5.00–5.25% in May, and 5.25–5.50% in July, where it peaked. (ycharts.com) That is three additional 25 bp hikes and a peak 50 bps higher than his implied 4.75–5.0%, so this component is clearly wrong.

2. Net cuts over the following two years. He said the curve was implying about a 50 bp reduction in rates over the subsequent two years. The Fed held the peak range of 5.25–5.50% from July 2023 through mid‑2024, then began cutting in September 2024 (to 4.75–5.00%), November 2024 (4.50–4.75%), and December 2024 (4.25–4.50%). (cnbc.com) By early 2025, the target range was still 4.25–4.50%. (ycharts.com) Relative to the February 2023 level of 4.50–4.75%, that’s only about 25 bps lower after two years, not the ~50 bps Sacks described, and it required a much higher‑than‑expected peak and late‑starting, larger cuts. So the trajectory and net change both differ materially from his implied path.

3. Long‑term rates around 3.5% and no return to ZIRP. Long‑maturity Treasury yields in late 2025 are closer to 4% than 3.5%: the 10‑year Treasury yield is around 4.0–4.1%, with the 5‑year near 3.7%. (mintrates.com) That doesn’t match “stabilizing” at 3.5% so far, though it’s also not wildly off. His claim that we’re “not going back” to near‑zero policy rates is not yet testable as a long‑run statement, but to date has not been falsified: the Fed funds rate has stayed in the 3.75–5.50% range since 2023, far from 0%. (ycharts.com)

Because the central, time‑bounded elements of the prediction—how high and how quickly rates would peak, and the scale of cuts over the next two years—were significantly off relative to actual Fed policy, the overall prediction is best classified as wrong, even though the very long‑run “no ZIRP” idea remains inconclusive.

Available market data supports Sacks’s view that, in the years after early 2023, public SaaS valuations did not return to the 2021 bubble extremes and instead re‑normalized to roughly pre‑COVID / ~2017 ranges.

Bubble reference point

• During the 2020–2021 boom, Bessemer’s Cloud 100 shows average cloud valuation multiples around 34x ARR in 2021, far above historical norms.
• A detailed SaaS valuation study using S&P Capital IQ notes the highest observed public SaaS multiple was Asana at ~89x LTM revenue in November 2021, illustrating the kind of ~100x revenue outliers Sacks was talking about, and that these levels were confined to the 2020–2021 bubble period. (bvp.com)

Post‑2022 normalization & current levels

• Multiple independent sources now place median public SaaS EV/Revenue around 6–7x in 2024–2025. For example: a SaaS M&A / public‑market review cites BVP’s Cloud Index at ~7.5x revenue in early 2025, up from ~6x a year earlier but still far below the 2021 peak near 18–20x. (buttondown.com)
• A separate 2025 valuation summary similarly reports median EV/Revenue for large U.S. public SaaS at ~6.1x as of September 2025, versus a peak near 20x in 2021. (aventis-advisors.com)
• The SaaS Capital Index, tracking public SaaS names, finds median ARR multiples around 7x in early 2025, with even the top 10 highest‑valued companies mostly in the low‑ to mid‑teens (e.g., CrowdStrike ~20.8x, ServiceNow ~19.5x, Datadog ~17.5x)—an order of magnitude below 2021’s ~90x outliers and well below the 30x+ cohort common at the peak. (saas-capital.com)

Comparison to 2017‑era valuations

• Historical work on SaaS comps shows that in 2016–2017, leading horizontal/vertical public SaaS names typically traded in the mid‑single‑ to high‑single‑digit EV/Revenue range (roughly 6–9x depending on subsector). (meridianib.com)
• A 2024 analysis of BVP’s Cloud Index explicitly notes that the current 6–7x median revenue multiple was last seen in October 2017, and characterizes 2022–2024 levels as a return to the “pre‑pandemic normal,” with no signs of a reversion to the 2020–2021 bubble. (linkedin.com)

Assessment versus the prediction

• Sacks predicted: (1) we’re never going back to the 2021 bubble, where some SaaS names traded near 100x revenue, and (2) valuations would revert to a more normal environment similar to 2017.
• As of late 2025, public SaaS valuations have indeed reverted to ~6–7x median EV/Revenue, matching late‑2010s norms, and even the top decile of names trade at <25x ARR—far below the 2021 extremes.

Given nearly three years of post‑podcast data showing sustained normalization at 2017‑like multiples and no re‑emergence of bubble‑level (~100x) public SaaS valuations, the prediction has so far played out as described. Hence, it is reasonable to mark this forecast as right on the available evidence.

Reasoning summary

• What he said (paraphrased): because of record amounts of capital raised, there would be more cash that needs to be deployed in the following 12–36 months (2023–2026) than has ever been deployed in the history of venture – if current fundraising/mandate structures held.
• Your normalized version preserves that as: record dry powder would have to be deployed in 2023–2026 “unless fundraising/mandates change materially.”

1. What actually happened to deployment?

Global VC investment (actual dollars deployed into deals) has not set new records in the 2023–2025 period:

• 2021 remains the clear peak year, with about $751.5B in global VC investments. (starthub.asia)
• After the downturn, global VC investment fell to roughly $349–346B in 2023 and only recovered modestly to about $368B in 2024—still roughly 50% below 2021’s peak. (starthub.asia)
• As of late 2025, industry reports describe 2024 as a “slight recovery” from a five‑year low, not a new high, and 2025 deal value is expected to grow from that subdued base, not surpass the 2021 boom on a rolling 12–36 month basis. (rothschildandco.com)

So if you interpret the prediction as “we will actually see record VC capital deployed in that 12–36 month window,” evidence to date points strongly against it: deployment has been far below the 2021–2022 peak levels.

2. What happened to dry powder and mandates?

On the dry‑powder side, his premise of “a lot more cash” was correct:

• Combined private equity + venture capital dry powder hit record highs in 2023 (~$2.59T) and again in mid‑2024 (~$2.62T). (spglobal.com)
• VC‑specific dry powder was about $317B at the end of Q1 2024—over 5× the ~$60B level in 2015—and still above $310B in 2025, i.e., not yet drawn down aggressively. (eisneramper.com)
• A growing share of this capital is “aging”: funds four years and older accounted for a larger share of uncalled capital, indicating capital not being forced out quickly. (forbes.com)

Crucially, fundraising and mandates did change materially, which is exactly the escape clause baked into the normalized prediction:

• Fundraising collapsed after 2021’s mega‑boom. Global VC fundraising dropped from about $404B in 2021 to roughly $214B in 2023 and $170B in 2024, and the number of VC funds fell by about a quarter. (rothschildandco.com)
• LPs became far more cautious, often pressuring GPs to slow the pace of capital calls and deployment rather than rush to deploy dry powder. (eu.vc)
• In response to the downturn, GPs have more frequently extended investment periods and fund terms, or at least drafted new funds with longer deployment windows; LPs have been more willing to grant such extensions in bad markets. (loyensloeff.com)
  • Articles explicitly note GPs “holding on to this capital a bit longer and really squeezing that investment period” rather than deploying on the old timetable. (spglobal.com)

Those changes mean that a large chunk of the record dry powder was not required to be pushed into deals within his 12–36 month window; instead, timelines were stretched and deployment stayed depressed even while dry powder remained at records.

3. Why this is classified as ambiguous

There are two overlapping interpretations:

1. Unconditional forecast about realized deployment
   If you read him as saying “VC will deploy record amounts in 2023–2026,” that is likely wrong, given clearly lower deal values vs. 2021–2022 and persistent record dry powder.

2. Conditional statement about what would need to happen if mandates didn’t change
   Your normalized version includes: “unless fundraising/mandates change materially.” In reality, fundraising, fund sizes, and deployment mandates did change in a major way (fundraising collapse, end of the megafund era, widespread extensions of investment periods), exactly the carve‑out he allowed. (rothschildandco.com)
   In that sense, we ended up in the exception case he explicitly set aside, so we can’t directly observe whether his statement about record deployment in the no‑change world would have been correct.

Because the prediction is explicitly conditional and the condition (no material change in fundraising/mandates) clearly failed, the realized world doesn’t provide a clean test of his “if that holds true” scenario. At the same time, the unconditional reading conflicts with observed deployment data.

Given that tension, and the difficulty of cleanly measuring “needs to be deployed” versus “actually deployed,” the fairest classification is “ambiguous” rather than definitively right or wrong.

Baseline: Global VC investment was about $294B in 2020 and $621B in the record year 2021, more than doubling year over year. (finbold.com) In the U.S., VC deal value was roughly $172.3B in 2020 and $329.9B in 2021. (nvca.org)

Post‑2023 reality: Global VC funding fell to roughly $248–262B in 2023 (about 40–60% below 2021, i.e., around 2–2.5x lower, not 3–4x), then partially recovered to about $275–315B in 2024, still well below the 2021 peak but clearly not at one‑third or one‑quarter of that level. (spglobal.com) H1 2025 already saw about $205B of global funding, the strongest half-year since H1 2022 and on pace to approach 2022 levels again, which is inconsistent with a sustained 3–4x deployment slowdown. (news.crunchbase.com)

U.S. data tell the same story: VC investment dropped from $329.9B in 2021 to $170.6B in 2023 (roughly half, i.e., about 2x slower than the peak, not 3–4x) and then rebounded to $215.4B in 2024. (nvca.org) In the first half of 2025 alone, U.S. startups raised $162.8B, up 75.6% year over year and already the strongest funding period since 2021, implying that full‑year 2025 will likely be back near or above pre‑slowdown levels rather than at one‑third or one‑quarter of 2020–2021. (reuters.com)

Qualitatively, Sacks was directionally right that funding would drop and deployment become more cautious compared with the 2021 boom, but the specific claim that deployment over the 3–4 years after early 2023 would run at roughly three to four times slower than in the prior few years is clearly not borne out by the actual multiples observed so far, and 2025’s AI‑driven rebound makes that magnitude of slowdown increasingly implausible for the remaining part of the window. Therefore, taken as a quantitative prediction about pace (3–4x slower), it is best judged as wrong.

By late 2025, the available data and reporting strongly support Sacks’s directional claim that launching new VC funds—especially first‑time, solo, seed and micro‑VC vehicles—has become significantly harder, with many such players effectively washed out.

Evidence that it’s much harder for new / emerging funds to raise:

• PitchBook data show that in the U.S., capital committed to first‑time VC managers fell from about $14.7B across 318 funds in 2021 to just $6.5B across 97 funds in 2023, and analysis suggests more than a third of the 2019–2021 first‑time managers may never raise a second fund. (crowdfundinsider.com)
• PitchBook further reports that 2024 set a record low for first‑time fundraising activity, with only 56 debut funds raising $3.7B and emerging managers capturing just 23% of fund value—a 10‑year low as LP capital concentrates into larger, established firms. (linkedin.com)
• Q1 2023 data cited by TechCrunch show emerging managers raised only ~13–14% of U.S. VC fundraising, down from roughly 26–29% in the prior boom years, indicating a sharp relative pullback. (techcrunch.com)
• A 2023 survey of emerging fund managers found 91% described fundraising as difficult or very difficult, with 85% taking more than six months to close and a large share delaying future raises—clear signs of a much tougher market for new/small funds. (signatureblock.co)
• A 2024 survey by Citrin Cooperman on emerging PE/VC managers reports that fundraising remains a major challenge; institutional LPs (pensions, insurers) are relatively scarce backers and prefer established managers, and 63% of respondents expect capital raising to remain difficult through 2024. (dakota.com)

Evidence of consolidation and “washing out” of smaller firms:

• The number of active U.S. VC firms has dropped by more than 25% since 2021 (from roughly 8,315 to 6,175), while over half of all U.S. VC capital raised in 2024 went to just nine top firms. This concentration of LP commitments has weakened smaller VCs and narrowed funding opportunities for early‑stage/startup‑focused firms. (ft.com)
• In the U.S., first‑time managers’ capital and fund counts have more than halved compared with 2021, and analysis of vintages 2019–2021 suggests hundreds of first‑time managers are unlikely ever to raise a second fund—effectively a culling of many new/small platforms that launched in the boom. (crowdfundinsider.com)
• European data tell a similar story: the number of VCs announcing new first‑time funds fell from 66 in 2022 to 42 in 2023 and 34 in 2024 (~50% decline in two years), and McKinsey data cited in the same analysis show sub‑$1B fund closures and new-manager formation in 2023 at their lowest levels since 2012. (blog.joinodin.com)
• Commentary from investors and Cambridge Associates in 2023–2024 underscores that this period is a “tough, tough environment” for emerging and solo managers; LPs are scrutinizing younger funds more intensely, and solo GPs without a very distinct edge are described as having a “really difficult” time raising in this market. (techcrunch.com)

Nuance: solo/micro funds still exist, but the easy-hype era is over:

• Not all solo or micro‑VCs have disappeared. Carta’s 2025 analysis finds that 69% of new VC funds on its platform in 2024 were under $25M and notes “a surge in the number of solo GPs,” especially at the very small end of the market. However, it also emphasizes that mid‑sized funds ($25M–$100M) have been “the hardest hit,” stuck between friends‑and‑family money and institutional mandates, and many of the 2020–2021 micro‑funds were self‑funded or reliant on HNW capital rather than traditional LPs. (carta.com)
• A 2024 report on emerging managers shows that over half of new emerging funds globally are led by solo GPs and most target very small fund sizes (<$10M), but it also documents substantial delays, cancellations, and fundraising difficulties tied to limited LP access and macro uncertainty—again indicating that, while some new solo/micro funds still launch, doing so has become meaningfully harder. (everythingstartups.com)

Assessment vs. Sacks’s prediction:

• Sacks predicted that “over the next several years” after early 2023, it would become much harder for new funds—especially solo capitalists, seed funds and micro‑VCs—to get started, and that a lot of that cohort would be “washed away.”
• By late 2025, the data show: (1) sharply lower fundraising volume and share for first‑time and emerging managers, (2) record‑low counts of debut funds, (3) a material decline in the number of active VC firms, and (4) clear evidence that many boom‑era emerging/solo managers cannot raise follow‑on funds.

Even though solo and micro‑VC funds have not vanished and some new ones still form, the environment is decisively more difficult and many hype‑driven or marginal funds have effectively been culled or frozen out of the market. On balance, that matches the core thrust of Sacks’s prediction, so it is best judged right.

Evidence strongly indicates that by late 2025 Apple had already exceeded $1 trillion in cumulative cash returned to shareholders (dividends + buybacks), so Chamath’s prediction came true.

Key points:

• A detailed April 9, 2025 analysis from Moods Investments states that since resuming dividends in 2012, Apple had returned $958 billion to shareholders (dividends plus buybacks) as of early April 2025. (moodsinvestments.ca)
• After that snapshot, Apple’s own Q2 FY2025 earnings call (covering the quarter ended March 29, 2025, reported May 1) disclosed that the company returned $29 billion to shareholders in Q2 alone ($25B in buybacks + $3.8B in dividends). (insidermonkey.com)
• In the subsequent Q3 FY2025 call (quarter ended June 28, 2025), Apple reported it returned over $27 billion to shareholders in Q3 ($21B in buybacks + $3.9B in dividends; the rest from other equity-settlement effects). (fool.com)
• Combining these: starting from $958B as of early April 2025, adding Q2 and Q3 returns (~$29B + >$27B ≈ $56B+) brings the cumulative total to about $1.014 trillion by the end of Q3 2025, i.e., well before the end of calendar 2025.
• Independent commentary is consistent with this math. A July 7, 2025 Motley Fool piece notes that since fiscal 2012 Apple had returned “nearly $1 trillion” to shareholders through dividends and buybacks as of Q2 2025, with an additional $100B buyback authorization just approved, implying the $1T mark would imminently be crossed. (fool.com)
• An October 4, 2025 analysis (citing Bloomberg estimates) goes further, stating that Apple has carried out over $1 trillion in share buybacks alone since 2015, not even counting dividends—so total cash distributions are clearly well above $1T by that point. (ainvest.com)

Taken together, these sources show that Apple’s cumulative shareholder distributions surpassed $1 trillion sometime between mid‑2025 and early Q4 2025, comfortably satisfying Chamath’s prediction “By 2025, Apple will have exceeded $1 trillion of cash distributions.”

Meta has clearly started returning large amounts of cash since early 2023, but the 4–5 year window Chamath referred to (roughly 2023–2027/28) is not over yet, so we cannot say whether "several hundred billion" will actually be reached.

Key facts so far:

• Share repurchases 2023–2025 (so far)
  • In 2023 Meta repurchased and retired 92 million shares for about $20.0B. (sec.gov)
  • In the first nine months of 2024 it repurchased 65 million shares for about $29.8B. (sec.gov)
  • In the first nine months of 2025 it repurchased 40 million shares for about $26.3B. (sec.gov)
  → That’s at least ~$76B of buybacks from Jan 1, 2023 through Sept 30, 2025, ignoring any Q4 2024 repurchases not broken out yet.

• Dividends added on top of buybacks
  • Meta initiated its first regular cash dividend in early 2024 at $0.50 per share and paid roughly $5.1B of dividends in 2024. (sec.gov)
  • In 2025 it raised the quarterly dividend to $0.525 and paid about $4.0B of dividends in the first nine months of 2025. (sec.gov)
  → Combined 2024–9M 2025 dividends are a bit over $9B.

• Total capital returned since 2023
  • Adding the known numbers, Meta has already returned ≈$85B+ to shareholders from Jan 1, 2023 through Sept 30, 2025 (≈$76B buybacks + ≈$9B dividends), with additional Q4 2025 activity still to come. (sec.gov) • Over the last decade as a whole, Trefis estimates Meta has returned about $178B via buybacks and dividends, underlining that "hundreds of billions" over multi‑year periods is plausible for this company, but that figure is for 10 years, not just 2023–2028. (trefis.com)

• Run‑rate and remaining authorization
  • Recent annualized repurchase pace is on the order of $25–35B per year, plus several billion per year in dividends, and Meta’s board has repeatedly topped up repurchase authorizations (e.g., +$40B in Jan 2023; +$50B in Jan 2024). (sec.gov)
  • As of Sept 30, 2025, about $25B remained authorized for future buybacks under the existing program, and the company explicitly notes that the timing and amount of future repurchases and dividends are discretionary and not guaranteed. (sec.gov)

Why this is inconclusive rather than right or wrong:

• Chamath’s normalized claim is that between roughly 2023 and 2028, Meta would return "on the order of several hundred billion" dollars to shareholders. By late 2025, the company has returned on the order of $100B or a bit less for that specific window—large, but still well short of the $200B+ most people would read as “several hundred billion.”
• However, there are still around 2–3 years left in the forecast horizon. Given Meta’s current cash generation, existing authorization, and history of large top‑ups to its buyback program, it is plausible that cumulative returns for 2023–2028 could eventually exceed $200B. At the same time, Meta’s filings stress that neither buybacks nor dividends are obligated, and capital‑spending or regulatory shocks could reduce future payouts. (sec.gov)

Because the relevant period is still ongoing and future capital‑return decisions are inherently uncertain, the prediction cannot yet be judged as clearly right or clearly wrong; the evidence to date only shows that the company is on a path that could reach that scale, not that it definitively will.

Chamath’s prediction is about the eventual mortality rate of the “bubble” startup cohort (roughly the 2020–2021 funding vintages), i.e., what percentage will ultimately fail over their lifetimes.

As of late 2025, that cohort is only ~3–5 years old. Typical analyses of startups suggest that 65–80% of startups fail within five years, and venture-backed companies and funds often take 7–10+ years to fully play out, with exits and write‑offs stretching over a decade.【2search12】 Carta- and PitchBook-based commentary on recent vintages explicitly notes that we are still "years out from seeing the performance data" for 2021–2022 funds, and that only a small minority of 2020–2021 funds have returned any capital to LPs yet—meaning many portfolio companies are still active (or at least not formally written off).【2search0】【2search1】

We do see strong signs of stress consistent with higher mortality: Carta data show that startup shutdowns have surged since 2022, with Q1 2024 recording the highest number of venture-backed closures this decade, and big percentage increases in closures at seed, Series A, and Series B stages.【2search10】【1search6】 However, these figures describe current flow of failures, not the final, lifetime failure percentage of the 2020–2021 funded cohort, and no credible source yet estimates that cohort’s ultimate mortality rate.

Because the relevant cohort has not matured and industry analysts themselves emphasize that final outcomes and performance for 2021–2022 vintages won’t be knowable for several more years,【2search0】【2search8】 there is not enough data yet to say whether the eventual failure rate will land around Chamath’s 50–60% range or higher/lower.

Therefore, the prediction’s accuracy is inconclusive (too early to tell) given the available evidence as of November 30, 2025.

Evidence from 2023–2024 shows that Sacks’ forecast of a major startup funding crunch in the second half of 2023 and through 2024 largely played out.

• Sharp drop in available capital: Global startup funding in 2023 fell to about $285B, down 38% from 2022 and the lowest level since 2018, with deep cutbacks across seed, early, and late stage rounds. (news.crunchbase.com) Multiple regions (North America, Asia, Africa, India) described this as a prolonged “funding winter” that continued into at least the first half of 2024. (brecorder.com)
• Down rounds and tougher terms: Data from Carta and other sources show that in 2023, roughly 19–20% of all VC rounds each quarter were down rounds, the highest rates since at least 2018. (carta.com) Preqin similarly reported that by Q2 2023, down rounds had doubled to about 16% of deals, up from a typical ~8%. (businesstimes.com.sg) In 2024, around **20% of all rounds remained down rounds—about double the long‑run average—indicating continued valuation pressure and harsher funding conditions. (forbes.com)
• Startups forced back to market and failing: Carta’s data show record startup shutdowns in 2023 (770 vs 467 in 2022), the highest ever tracked, driven by an inhospitable fundraising climate. (forbes.com) Carta’s 2024 analysis reports that even more startups shut down in 2024 than in 2023, with failures concentrated among companies funded in 2021–2022 that hit the end of their runways and couldn’t raise on acceptable terms. (forbes.com) This is exactly the dynamic Sacks described: companies that had waited were eventually forced to raise in much tougher conditions, leading to down rounds, recapitalizations, and shutdowns.

While there were pockets of strength (especially AI) and the crunch actually began earlier than H2 2023, the core prediction—that a major funding crunch in late 2023 and 2024 would hit startups that had delayed fundraising, resulting in widespread down rounds and failures—matches what happened in the data.

Sacks predicted that in 2H 2023 and through 2024 many startups that had delayed fundraising would hit low cash levels and be forced to raise into a much tougher funding environment than in prior years.

Global and regional data show a pronounced, prolonged funding crunch over exactly that window. Crunchbase data indicate global startup funding in 2023 fell to about $285B, down 38% from 2022 and the lowest since 2018, with early‑stage, late‑stage, and seed all dropping 30–40%; Q4 2023 was the weakest quarter in years, confirming a harsh fundraising backdrop as the second half of 2023 unfolded. (news.crunchbase.com) A CB Insights–based review similarly notes 2023 venture funding at its lowest since 2017, with late‑stage deal sizes halved vs. 2021 and IPOs at decade‑low levels, setting up a weak environment going into 2024. (foley.com)

Conditions did not normalize in 2024; they remained tight outside a small number of AI winners. Barron’s reports global VC investments in 2024 at $314B—only ~3% above 2023 and still about 55% below the 2021 peak, with a disproportionate share going to a handful of AI companies rather than the broader startup base. (barrons.com) In multiple ecosystems, this was explicitly described as a continuing “funding winter” and cash crunch: Indian startup funding fell 62% in 2023 to a six‑year low, helping drive tens of thousands of shutdowns and a very subdued 2024; analyses speak of a “relentless and prolonged funding winter” and deep cash shortages. (business-standard.com) Pakistan saw 2024 Q1 startup funding drop to zero deals, with commentary tying this to the weakest global tech funding since at least 2018. (brecorder.com) Reuters and others describe Europe and Indonesia facing similar 2023–24 funding collapses and explicit “funding crunch” conditions. (reuters.com)

Evidence also supports the mechanics Sacks described—companies that had extended runway after 2021 finding themselves forced back to market in 2H 2023–2024 and confronting harsher terms. A Silicon Valley Bank H2 2023 outlook estimated that U.S. VC‑backed tech startups collectively burned tens of billions per month and modeled a large share running out of runway by 2024, predicting outcomes like valuation capitulations, soft‑landing M&A, and outright failures if they couldn’t raise. (scribd.com) A separate fintech analysis forecast roughly a quarter of U.S. fintechs would be out of cash by Q3 2024 without new funding. (scribd.com) Empirical markers of this crunch show up in the surge of down rounds and alternative structures: Indian data show down rounds nearly quadrupling by early 2023 vs. a year prior, with commentators saying startups can no longer demand 2021‑style valuations. (hindustantimes.com) In Israel, a 2025 report finds that 43% of 1,530 startups surveyed needed to raise soon and 26% of 2024 later‑stage rounds were down rounds, reflecting widespread forced fundraising into reset valuations. (israeliinvestments.com) Across Europe, VC funding fell by more than half from 2021 to 2023, and startups increasingly relied on complex convertible‑debt structures to obtain cash while avoiding explicit valuation cuts, another sign of limited equity appetite on investor‑friendly terms. (reuters.com) Meanwhile, venture‑debt volumes hit record levels in 2024 as companies sought nondilutive capital in a “constrained equity market,” per the Wall Street Journal. (wsj.com)

Finally, industry observers explicitly anchored the timing of the reckoning to the same 2H 2023–2024 window. Elad Gil’s widely read "Startup Decoupling & Reckoning" essay argued that many companies which raised large 2021 rounds with 2–4 years of runway would hit a “hard reckoning” starting at the end of 2023 and accelerating through the end of 2024 as their cash ran out, closely matching Sacks’ timeline and logic. (blog.eladgil.com) Taken together—sharp and sustained global funding declines, a very tough 2H 2023 and 2024 environment outside a narrow AI boom, widespread down rounds and valuation resets, large numbers of startups facing or hitting cash‑out, and contemporaneous analyses pegging the crunch to exactly that period—the prediction that there would be a major funding crunch for startups in the second half of 2023 and throughout 2024, forcing many delayed fundraisers to raise under much tougher conditions, is well supported by the data.

Evidence from 2023–2024 shows the pattern Sacks described did in fact emerge for many startups with venture debt overhang during the funding crunch.

1. Funding crunch and heavy use of venture debt. In 2023, equity funding into startups fell sharply (e.g., Indian startup equity funding down ~70% vs 2022), while venture debt volumes rose or stayed high as founders tried to bridge rounds and extend runway without dilution.(business-standard.com) This matches the “upcoming funding crunch” backdrop Sacks was talking about.

2. Runway shorter than expected due to dependence on venture debt. When Silicon Valley Bank collapsed in March 2023, a large provider of venture credit, founders suddenly lost access to expected debt facilities. A GeekWire report noted that “many startups are waking up today to find that their runway is much shorter than they expected without that venture debt,” explicitly tying venture-debt assumptions to overestimated runway.(geekwire.com)

3. Lenders moving to protect themselves before companies run out of cash. Venture debt agreements commonly include covenants such as minimum cash balances and performance thresholds; breaching them can trigger default and allow lenders to accelerate repayment or impose stricter terms—i.e., move to collect while the company still has money. Multiple practitioner guides and analyses from 2023–2025 emphasize these covenants and note that default can lead to accelerated repayment or forced restructuring.(insightfulcfo.blog) This is structurally identical to Sacks’s point that banks “are going to try and collect the debt before they start running out of money.”

4. Concrete cases of debt shortening effective runway in the crunch. Reporting from 2024 shows several startups that leaned on venture debt during the funding winter—such as Good Glamm Group, Reshamandi, Arzooo and Waycool—later struggling to meet repayments and asking lenders for extensions as cash got tight. Lenders are described as reluctant to grant relief and having to decide which borrowers to support, underscoring that debt service and lender behavior were constraining these companies’ usable runway sooner than hoped.(livemint.com)

5. Lenders pushing for exits before total failure. By early 2025, TechCrunch reported that many lenders, facing challenged loans originated in the 2020–2022 boom, were “increasingly pushing startups to sell themselves to minimize potential losses,” with many companies “at the end of the rope.”(techcrunch.com) That is consistent with lenders acting to realize value while there is still cash or enterprise value left, not waiting for companies to hit zero.

Taken together, these points show that during the 2023–2024 crunch, startups that had layered on venture debt often discovered their real, usable runway was shorter than the headline cash balance implied, and that lenders’ covenants and repayment demands effectively forced earlier reckoning. While not every startup with venture debt experienced this, the mechanism and timing Sacks predicted are well supported by the subsequent funding environment and case reporting, so his prediction is best judged as right.

Available data show that Sacks’ timing and the qualitative description of a funding crunch with widespread down rounds and restructurings in H2 2023 and 2024 were broadly accurate.

1. Major funding crunch (capital into VC and startups)
   • In the U.S., VC investment fell from about $242B in 2022 to ~$171B in 2023 (≈35% drop), with deal count down ~23%, according to PitchBook/NVCA; NVCA describes 2023 as a pronounced continuation of the 2022 slowdown. (axios.com)
   • VC fundraising itself also seized up: PitchBook data show U.S. VC funds raised just under $70B in 2023, a ~60% decline from 2022 and a six‑year low, limiting available capital for startups in 2023–24. (pennmutualam.com)
   • Globally, Q1 2024 VC investment was near a five‑year low, with total funding and deal count both sharply reduced versus prior years, and analysts not expecting a rapid rebound—indicating the crunch persisted into 2024. (reuters.com)

2. Widespread down rounds in 2023–24
   • U.S. data compiled in the Big Book of Venture Capital 2023 show that 19% of venture rounds in 2023 were down rounds, up from only 5% in 2021 and the highest level in a decade; Q3 2023 alone had 17% down rounds. (slideshare.net)
   • A separate analysis using Carta data reports that on its platform, down rounds accounted for roughly 19–20% of all primary financings in each quarter of 2023, four consecutive record highs since 2018, and notes that down rounds had become “routine,” especially at later stages. (linkedin.com)
   • A 2024 Morgan Lewis survey likewise notes that 2023 produced the four highest quarterly down‑round rates since 2018, and that Q1 2024 had the highest share of down rounds in five years (≈23%), confirming that the wave of repricings continued into 2024. (jdsupra.com)
   • 2024 aggregate data show down rounds remaining elevated: about 20% of all priced U.S. rounds on Carta in 2024 were down rounds (roughly double the historical ~10%), with late‑stage companies seeing down‑round shares rising from 28% to 39% in early 2024; nearly half of late‑stage deals were flat or down. (scribd.com)
   • DocSend’s deal‑flow analysis finds that 33% of VC deals in Q1 2024 and 22% in Q2 2024 were down rounds, and nearly 30% of all deals were flat or down, still materially worse than pre‑2023 norms. (docsend.com)
   Together, these sources indicate that from H2 2023 through at least mid‑/late‑2024, down rounds were common across the ecosystem and well above normal levels—consistent with “widespread down rounds.”

3. Restructurings, recapitalizations, and harsher terms
   • Legal and market commentary aimed at venture‑backed startups notes a surge in complex down‑round financings, explicitly tying them to recapitalizations (“recaps”) and pay‑to‑play structures. Morgan Lewis describes recaps as often being done in conjunction with a down round and frames them as increasingly necessary tools to get new capital in during this period. (morganlewis.com)
   • Axios, citing Cooley data, reports that by Q2 2024, 8.7% of all VC deals included “pay‑to‑play” (cram‑down) provisions, a record share and a clear sign of aggressive recap-style structures that heavily dilute non‑participating existing investors. These provisions historically appeared mostly in stressed late‑stage deals, but by 2024 they were spreading across stages. (axios.com)
   • Standard VC glossaries and practitioner write‑ups emphasize that recaps are used when cap tables and prior preferences make fresh funding difficult—exactly the situation more startups faced after inflated 2020–21 valuations collided with the 2022–24 downturn—supporting that these recapitalization tools were indeed being deployed. (nexitventures.com)

4. Broader signs of ecosystem stress (“mass extinction” dynamics)
   • Morgan Lewis notes that startup shutdowns jumped from 467 in 2022 to 770 in 2023, and that Q1 2024 shutdowns rose another 58% year‑over‑year, highlighting elevated mortality as funding dried up. (jdsupra.com)
   • Carta and FT data show 2023 as the first year in at least five years with a net contraction in startup employment, including a wave of layoffs and sharply reduced hiring, reflecting widespread restructuring and belt‑tightening. (ft.com)

Taken together, these independent datasets show that in H2 2023 and throughout 2024 the startup ecosystem experienced: (a) a pronounced funding crunch, (b) historically high and sustained levels of down rounds, especially at later stages, and (c) increased use of harsh recap / pay‑to‑play structures alongside shutdowns and layoffs. That combination matches Sacks’ forecast of a major crunch characterized by widespread down rounds, restructurings, and recapitalizations, so the prediction is best judged as right.

As of late 2025, the broad public SaaS market has not clearly reverted to an ~8x EV/NTM revenue “long‑term mean,” but the outcome of Sacks’s open‑ended “eventually” prediction still depends on future market moves.

Evidence:

• Jamin Ball’s Clouded Judgement tracks a large universe of public software/SaaS names. In December 2023, the median EV/NTM revenue multiple was about 6.4x, which he notes is roughly 17% below a long‑term average of 7.8x.【turn0search6】
• Around year‑end 2023, his data still showed medians in the mid‑6x range (about 6.5x), i.e., re-rated upward from the 2022 lows but still materially under 8x.【turn0search4】
• Through 2024 and into 2025, subsequent Clouded Judgement posts show median EV/NTM revenue sliding back to roughly 6.0x–5.9x in early 2024 and then nearer 5.2x–5.4x and even ~4.8x by various points in 2025, indicating that the typical SaaS name has not sustained a return to 8x.【turn0search5】【turn0search7】【turn0search3】【turn0search0】【turn0search2】【turn0search9】
• A different lens, the BVP Nasdaq Emerging Cloud Index, shows index-level revenue multiples around 8.4x in 2024–25, with broader public SaaS medians in the ~5.5x–7.0x range.【turn0search11】 This suggests that a weighted average of higher‑quality/cloud names is roughly at or slightly above the historical 8x level, while the overall universe is still below it.
• Multiple sources agree that current public SaaS valuations remain far below the 2021 bubble peak (often cited around 20–21x revenue; one summary notes current levels are still ~65–70% below that), i.e., broadly consistent with Sacks’s claim that a reversion to ~8x would be “well below” the 2021 peak.【turn0search11】

Why this is inconclusive rather than clearly right or wrong:

• The prediction is explicitly open‑ended: SaaS multiples will eventually revert to roughly 8x. There is no time horizon like “within two years,” so by late 2025 it is impossible to say with certainty that this will not occur in the future.
• On one reasonable definition (broad SaaS median), the market has not yet reverted to ~8x; medians peaked in the mid‑6x range before falling back.【turn0search4】【turn0search6】【turn0search3】
• On another plausible definition (a leading cloud/SaaS index’s average multiple), valuations are already around or slightly above 8x, which could be read as the reversion Sacks anticipated.【turn0search11】

Because both (a) the key term (“SaaS multiples”) can be reasonably measured in ways that give different answers, and (b) the forecast extends indefinitely into the future, we cannot definitively say yet whether Sacks’s prediction about an 8x reversion is right or wrong. Thus the appropriate status is inconclusive at this time.

Evidence from multiple independent SaaS and cloud benchmarks shows that sector‑wide valuation multiples have not returned to the 12–16x EV/forward‑revenue (or EV/TTM revenue) range seen in the 2020–2021 bubble period.

• Aventis Advisors’ long‑term study of public SaaS valuations reports that median EV/Revenue multiples peaked around 18–19x in 2021, then collapsed; by early 2023 the median had fallen to about 6.7x, and as of September 2025 it sits near 6.1x—far below bubble levels.(aventis-advisors.com)
• NGP Capital’s analysis of public SaaS companies finds that after the 2021–2022 peak, revenue multiples "descended to levels not seen since 2016" and stabilized in a 6–8x range; 2023 ended with an average EV/TTM revenue multiple of 7.9x and a top‑quartile multiple of 10.7x (still under 12x).(ngpcap.com)
• Software Equity Group’s SEG SaaS Index shows a median EV/Revenue multiple of 6.2x in February 2023 and 5.6x by Q3 2024, noting that Q3 2024 levels remain roughly 45% below the 2021 peak.(softwareequity.com)
• A 2025 summary of public SaaS M&A data (Aventis Advisors / SaaS M&A report) indicates that median public SaaS revenue multiples were about 5.6x in 2024 and roughly 7.4–7.5x TTM revenue in early 2025 (BVP Cloud Index and Aventis data), again well below a 12–16x sector median.(buttondown.com)
• A separate 2025 analysis of the BVP Nasdaq Emerging Cloud Index (a broad cloud/SaaS proxy) estimates an average revenue multiple of roughly 9.0x as of late September 2025—elevated versus 2022 lows but still not back to 12–16x.(profitpencil.com)
• Private and hybrid indices show similar patterns: SaaS Capital’s index of public/private SaaS names reports a median ARR multiple around 7.0x at the start of 2025; only a small handful of top‑tier companies trade at 14x+ revenue, which are outliers, not the sector average.(saas-capital.com)

One outlier estimate from Boğaziçi Ventures claims a 12.3x median EV/Revenue for SaaS companies in 2023, but this conflicts with the bulk of specialized SaaS valuation datasets above, which all cluster in the mid‑single‑ to high‑single‑digit range for 2023–2025.(bogaziciventures.com) Given the consistency of the major SaaS/Cloud benchmarks (SEG, Aventis, NGP, BVP index–based analyses), the consensus is that broad SaaS valuation multiples have not re‑established the 12–16x range since Jason’s February 2023 prediction. While a few elite names have traded above 12x, the sector median/average has remained substantially lower.

On that basis, Jason’s claim that SaaS multiples would not return to the 12–16x bubble range “for the foreseeable future” has been borne out up through late 2025.

U.S. interest rates have not returned to anything close to the 0% regime that prevailed pre‑2022; since early 2023 the federal funds target range has stayed roughly 3.75–5.5%, far above zero. (ycharts.com) At the same time, public SaaS / software valuations remain far below their 2021 bubble levels: median public SaaS EV/Revenue multiples peaked near ~18–20x in 2021 (with extremes like Asana around 89x revenue), then collapsed to about 6.7x by early 2023 and have since stabilized roughly 6–7x through 2025, despite some AI optimism. (aventis-advisors.com) Other surveys and indices show a similar pattern: public SaaS median multiples hovering around 6–7.4x in 2023–2025 versus 18.6x in 2021, and broader software/Cloud indices normalized well below their pandemic highs, suggesting that the lower‑multiple regime has persisted for several years. (windsordrake.com) There are notable outliers—most prominently Nvidia, whose price‑to‑sales ratio in 2023–2024 at times exceeded its 2021 peak even with higher rates—but these are a handful of mega‑caps, not a sector‑wide return to 2021‑style valuation multiples. (m.macrotrends.net) Taken together, the broad tech/growth‑equity market has not revisited 2021‑magnitude multiples in a non‑zero‑rate environment through late 2025, so Chamath’s core claim—that 2021‑style multiples wouldn’t broadly return and that early‑2023 lower valuations would be the new normal for an extended period—has so far been borne out, notwithstanding a few high‑profile exceptions.

The 18–24 month window after early February 2023 (roughly Feb 2023–Feb 2025) did in fact feature very difficult operating conditions for many startups, matching Sacks’s prediction that it would be “pretty rough for a lot of companies.”

• Severe funding pullback and down/flat rounds. Global startup funding in 2023 fell to about $285 billion, down 38% from 2022 and the lowest level since 2018, with deep cutbacks at seed, early, and late stage. Reports explicitly note that 2021-era valuations did not hold up, and that many “promising companies raised flat and down rounds,” while startups had to tighten belts and focus on unit economics as investors deployed capital more sparingly. (news.crunchbase.com)
• Large wave of shutdowns (“mass extinction event”). PitchBook data, summarized by Business Insider and others, shows about 3,200 U.S. venture‑backed startups that had raised roughly $27–27.2 billion shut down in 2023, a pattern observers described as a “mass extinction event” for startups, driven by the funding drought and investor pressure to improve economics. (tech.yahoo.com)
• Widespread layoffs across tech/startups. Layoffs.fyi data, cited in major outlets, indicates roughly 450,000 tech layoffs tracked by 2023. Coverage in early 2024 reports tens of thousands more tech layoffs just in the first weeks of 2024, showing that staff cuts and restructurings remained common as companies attempted to survive the downturn. (en.wikipedia.org)
• 2024 still tough outside of AI. In 2024, global VC funding ticked up only slightly to about $314 billion, still ~55% below the 2021 peak, with nearly a third of all capital concentrated in AI startups and exits (IPOs/M&A) remaining sparse—evidence that most non‑AI startups still faced a capital‑scarce, high‑bar environment well into 2024. H1 2024 funding remained below or flat versus the prior year, and reports consistently described the broader VC market as subdued. (barrons.com)

Because funding contracted sharply, many startups endured layoffs, flat/down rounds, and shutdowns across exactly the 18–24 month period Sacks referenced, his prediction that it would be “pretty rough for a lot of companies” is well supported by subsequent data and reporting.

Based on standard definitions, the U.S. and global economies did not enter recession in 2023.

• United States: The official arbiter of U.S. recessions is the National Bureau of Economic Research (NBER). As of now, NBER has not declared a recession for any period in 2023. Instead, U.S. real GDP grew about 2.5% for 2023, and the labor market remained relatively strong, with unemployment mostly between about 3.4–3.9% during the year—levels inconsistent with a typical recessionary downturn. (Growth and jobs data are summarized in official releases from the U.S. Bureau of Economic Analysis and Bureau of Labor Statistics.)
• Global economy: The IMF and World Bank did not classify 2023 as a global recession. The IMF’s World Economic Outlook showed positive global real GDP growth in 2023 (on the order of 3%), describing a "slow" or "weak" recovery, but not a contraction consistent with a global recession.

Because neither the U.S. nor the global economy actually entered a recession later in 2023, Sacks’s prediction that there was a “really good chance of recession later this year” did not come true, so it is best scored as wrong rather than inconclusive or ambiguous.

Evidence from multiple VC data sources shows that 2021 was an exceptional peak year for startup capital (roughly US$750B globally) and that funding volumes and deal activity in 2023–2025 have stayed well below that peak. (onmine.io) Global venture funding fell sharply in 2022–2023 and, although it recovered somewhat in 2024 to around US$314–368B, that was still roughly half of 2021’s level; one analysis notes 2024 VC activity remained about 55% below the 2021 peak. (news.crunchbase.com) PitchBook/NVCA and KPMG both describe 2023–2024 as a subdued, investor‑friendly market marked by a liquidity crunch, reduced deal sizes, and low VC fundraising—conditions explicitly contrasted with the easy‑money environment of 2021. (pitchbook.com) Structural tightness is also visible in the consolidation of capital into a smaller number of top-tier firms and a >25% drop in the number of active U.S. VC firms since 2021, which narrows funding access for many startups compared with the broad-based exuberance of 2021. (ft.com) By Q3 2025, quarterly global VC investment (~US$120.7B) is robust but still below average 2021 quarterly levels and heavily concentrated in AI megadeals rather than a return to across-the-board easy funding. (kpmg.com) Taken together, these facts support Sacks’s claim that, for the period after early 2023, capital availability for startups has remained structurally tighter and materially different from the 2021 boom, without a reversion to 2021-style easy funding conditions.