Week 47 · 16–22 November 2026
Nvidia’s fall catches up with venture capital
Between its record close on 6 October and 20 November, Nvidia lost 47.5% of its market value. The shock reaches venture capital through two channels. Investors now have new reference prices for AI companies, and several of the corporates that fund their largest rounds are themselves caught in the correction. At the same time, fund investors surveyed by EY are increasing their private equity allocations but asking for more evidence of the value created.
Anthropic puts a market price on AI
On Thursday, Anthropic, the developer of Claude, made its Nasdaq debut after pricing its IPO at $83 per share. On Friday, the stock closed at $88.91, down 8.5% on the day, while the US semiconductor index fell by about 11%.
The IPO raised $91.6 billion gross for Anthropic. Shareholders who sold shares in the offering collected close to $2 billion. The gap matters to fund investors: the company financed its growth, but only a fraction of the offering allowed shareholders to get their money back.
Sources: Anthropic, offering close, 20 Nov. · Axios, closing price, 20 Nov. · Reuters, markets, 20 Nov.
In Europe, concentration came before the fall
European unicorns had raised €26.8bn by the end of September, already more than in the whole of 2025 (€18.3bn). AI accounted for 84.3% of that amount and for nine of the ten largest unicorn rounds. At the other end of the market, the median pre-money valuation at pre-seed and seed fell to €4.8m.
Source: PitchBook, European venture valuations, 17 Nov.
In this issue
- 1 | Venture capital. The fall of four stocks and its effects on start-up funding.
- 2 | Limited partners. Distributions lag despite plans to reinvest.
- 3 | Climate. Why Eurazeo walked away from buying a hotel chain.
- 4 | Buyout. Rising prices and choosing a manager.
- 5 | United States. What the 4,186 small deals recorded in 2025 measure.
1 | Venture capital — How the AI correction is passed on to VC funds
Venture capital / Artificial intelligence
Nvidia lost almost 23% in the single week of 16 to 20 November. The group occupies a singular position: it sells the chips on which AI model developers depend, takes stakes in them and joins rounds that attract other investors. As of 10 November, the rounds in which it had participated accounted for 27.2% of global VC deal value in 2026, against 7.1% in 2025. PitchBook counts the full value of these rounds, co-investors included. Seven AI companies accounted for 80.3% of their value and 6.4% of the number of deals.
The four stocks at the close on Friday 20 November 2026
| Company · ticker | 2026 peak | 52-week high ($) | Close 20 Nov. ($) | Decline |
|---|---|---|---|---|
| Oracle · ORCL | 1 June | 250.25 | 101.19 | −59.56% |
| Nvidia · NVDA | 6 Oct. | 243.37 | 125.53 | −48.42% |
| Alphabet (Google) · GOOGL | 18 May | 408.61 | 265.29 | −35.08% |
| Meta Platforms · META | 24 Sept. | 779.82 | 560.84 | −28.08% |
Prices in US dollars. The decline is calculated from the 52-week intraday high. For Nvidia, it is 48.42% from that peak and 47.53% from its record close of $239.24. Alphabet is represented by its Class A share. Prices: Oracle · Nvidia · Alphabet (Google) · Meta Platforms
What the fall changes for VC funds
Valuations. Oracle has lost almost 60% from its peak, Meta 28%.
Fundraising. An AI model developer that has to pay for its next order of computing power can hardly wait for share prices to recover. If a corporate reduces its participation in a round, the lead investor will have to find another backer, accept a smaller round or renegotiate the price. Conversely, Nvidia was still one of the seven corporates that subscribed to Anthropic’s $26bn private placement, which closed on Friday. A withdrawal by Nvidia from private funding remains a hypothesis.
In Europe, rising medians and a two-speed market
PitchBook’s report on European venture valuations, published on 17 November (data as of 30 September, Europe and Israel), compares the medians for the first nine months of 2026 with those for the whole of 2025. The median European round size rises to €2.4m (+26%) and the median pre-money valuation to €7.8m (+2.7%). The increase is concentrated in late stage, while seed is the only stage to decline. The cumulative value of European unicorns reaches €654.6bn, against €525.4bn at the end of 2025, and down rounds account for 12.4% of deals, against 13.3% in 2025, with levels close to 17% in fintech.
Median pre-money valuation, nine months of 2026 and change vs full-year 2025
| Scope | 2026 median (9 months) | Change |
|---|---|---|
| Europe, all stages | €7.8m | +2.7% |
| United States, all stages | €42.6m | +41.9% |
| Europe, pre-seed and seed | €4.8m | −3.8% |
| Europe, Series A-B | €44.2m | +38.5% |
| Europe, Series C-D | €811m | +260.4% |
PitchBook, data as of 30 September 2026. Medians for the first nine months of 2026 compared with the full year 2025; some categories rest on fewer than twenty deals. In Series C-D, the median reached €897.9m in the first half. Proprietary PitchBook data, not corroborated at this stage by an independent source.
Invest Europe, which measures fund activity rather than funding rounds, made the same observation of concentrated capital in the first half: €11.5bn invested in venture capital, a record, in 1,182 companies against 1,513 a year earlier. The two series are not directly comparable: scopes, sources and cut-off dates differ.
Sources: PitchBook, Nvidia, 13 Nov. · PitchBook, European valuations, 17 Nov., pp. 3, 8 and 17 · Invest Europe, H1 2026 activity, 5 Nov. · Reuters, 20 Nov. · Anthropic, 20 Nov. · IPEV, valuation
Read on PE Research: US and European venture capital
2 | Limited partners — Investors want to reinvest. How fast are they getting their money back?
Limited partners / Allocations and fees
58% of the investors surveyed by EY plan to increase their private equity exposure by 2028 and 7% to reduce it. At the same time, 78% of private equity managers consider their pace of distributions to be below the historical norm.
The survey, published on 18 November, covers 225 institutional investors (median assets of $12.5bn) and 200 alternative asset managers (median assets of $1.5bn) interviewed in eleven countries between 14 July and 5 August. Investors devote on average 27% of their alternatives portfolio to private equity and 17% to private debt. The 2024 edition showed 22% and 14%, on a different sample. EY’s definition of private markets includes hedge funds and crypto-assets.
What limited partners are willing to pay
Illiquidity is cited by 54% of investors as a growing challenge, ahead of management fees (39%) and valuation opacity (34%). 26% name fees that are too high relative to the value created as their main source of tension with managers. Almost a third of investors, 31%, would prefer lower fixed fees and a larger share of gains going to the manager. Only 23% of managers accept this formula, and 45% remain attached to the “2 and 20” model.
Managers, for their part, are turning away from individual investors: 51% say they have no plans to market to them, against 43% in 2024, and 62% of institutional investors are concerned about greater retail participation.
What teams actually achieve
Alvarez & Marsal sheds light on this delay. In its European survey, 65% of the funds surveyed say they achieved less than half of the value planned in their value-creation plans over the past two years; in the United States, about four portfolios in ten fall short of the entry thesis. Faced with an asset that is hard to sell, 43% of European respondents cite a continuation fund and 7% a discounted sale, against 31% and 27% in the United States.
AI: adoption is moving faster than proof
The three surveys below, conducted at different dates and on different samples, cannot be compared figure by figure; they converge on a gap between adoption and measured effect on performance.
| Survey | Sample and date | What it measures |
|---|---|---|
| EY, 18 Nov. | 225 investors, 200 managers; July-August 2026 | 46% of investors already see a return gain; 19% of managers |
| S&P Global, 13 Apr. | Managers and LPs; February 2026 | 31% have integrated AI into acquisition due diligence; 27% consider it effective in creating value |
| Alvarez & Marsal, 17 Nov. | 200 Europe (Q1), 100 United States (April) | AI deployed in some functions, rarely at company-wide scale |
According to EY, 48% of investors think AI should allow fees to be lowered, 9% that it would justify higher fees.
Sources: EY, Global Private Markets Survey, 18 Nov., pp. 7 to 10, 16 and 18 · Alvarez & Marsal, Operational Alpha, 17 Nov., pp. 5 and 6 · S&P Global, 13 Apr.
Read on PE Research: Pension funds and universities, pillars of private equity financing (in French)
3 | Climate — The climate risk that made Eurazeo walk away from an acquisition
Climate / Physical risk and financing
Four of the five operating hotels of a Spanish chain studied by Eurazeo were exposed to high heat risk. Three other sites were under development. The group owned its properties on the Spanish coast: adaptation spending would therefore have weighed directly on its cash flow. In the scenario examined, it represented about 90% of one year’s EBITDA, before any development capex. Eurazeo walked away from the deal.
This case is published by Invest Europe in its guide of 16 November. The chain’s annual EBITDA was estimated at €16m. The buyer had to weigh the cost of protecting each site, how long it could operate it, and the financing needs of the new hotels.
What the guide asks of managers
Invest Europe’s Climate Change Guide 2026, a 119-page document written with the consultancy ERM (Environmental Resources Management), updates the 2021 edition. Its starting point: while regulation is shrinking or fragmenting (the Omnibus directive raises the CSRD thresholds to more than 1,000 employees and €450m of turnover; the SFDR 2.0 reform is under discussion), investors’ climate expectations have become a minimum requirement, and falling behind becomes a fundraising risk. The guide recommends four practices:
| Recommendation | What it implies for a manager |
|---|---|
| A single database | Collection at each portfolio company, reused for SFDR, the ESRS standards, the ISSB and LP questionnaires |
| A dated quality upgrade | Spend-based estimates at entry; activity data required within 12 to 18 months of ownership |
| A single target framework | As chosen: the SBTi (Science Based Targets initiative, which checks that a company’s climate targets are consistent with the Paris Agreement), NZIF 2.0 (Net Zero Investment Framework, a net-zero framework for investors, version 2.0) or the PMDR (Private Markets Decarbonisation Roadmap) of the iCI (Initiative Climat International, a network of private equity managers backed by the PRI) |
| A strategic transition plan | Written as a business strategy, not as a sustainability report |
The guide publishes no market statistics. Its case studies, self-reported and often estimated, nevertheless show the gap between pioneers and the average: Astorg reports that 88% of its portfolio, by invested capital, was covered at end-2025 by emission-reduction targets validated by the SBTi, while Coller EQT finds only about 29 to 30% of managers committed to net zero among those behind the 1,439 underlying companies of its flagship secondaries fund.
Source: Invest Europe, Climate Change Guide 2026, pp. 5, 19, 77–78, 80 and 83
Why climate venture capital also depends on AI
In climate technologies, PitchBook records $22.5bn of venture capital in the third quarter. The amount rose by about 62% quarter on quarter while the number of deals fell by about 7%. Over the first nine months of 2026, the total reaches $50.1bn across 1,750 deals, already more than in the whole of 2025 ($44.1bn across 2,353 deals): the comparison is nine months against twelve, and the report preview does not give the total for the first nine months of 2025. Crusoe and Firmus, two companies linked to computing infrastructure, alone account for $5.9bn, more than a quarter of the total.
Net Zero Insights counts, on its own scope, $19.6bn of climate funding, down 7.6% over the quarter. Equity funding nevertheless rose 17%, to $15.8bn. The decline in the total comes mainly from asset-backed debt. Differences in scope and in the instruments counted explain why the two publications can describe the same quarter in different terms.
Sources: PitchBook, Climate Tech, Q3 2026 · PitchBook, Q3 2026 Climate Tech Report, data as of 30 September 2026, p. 6 (preview supplied to the editors) · Net Zero Insights, 15 Oct.
4 | Buyout — Buyout prices are rising. How to choose a manager?
Buyout / Acquisition prices and manager selection
The Argos Index published on 19 November reached 8.9 times EBITDA in the third quarter, after 8.8 times in the second. For acquisitions of control of unlisted SMEs in the euro area, funds paid a median multiple of 9.0 times and trade buyers 8.8 times. The figures cover a rolling six months and different companies depending on the type of buyer.
Source: Argos and Epsilon Research, Q3 2026, 19 Nov.
Disclosure: Gilles Mougenot is co-founder and Senior Advisor of Argos. Biography
Choosing a manager’s next fund
EQT tracked teams whose flagship buyout fund had ranked in the top quartile of its vintage. In the group studied for the next fund (131 observations), 38% fall below the median; by the third fund (106 observations), the proportion reaches half. Managers who stayed consistently above the median, fewer than 10% of the universe according to StepStone, outperform occasional champions by up to 30% in multiple over ten years. EQT also recalls, based on StepStone data cited by McKinsey, that about half of buyout value creation between 2010 and 2022 came from multiple expansion and leverage; according to its own analysis based on PitchBook, average holding periods have lengthened from three years in the early 2000s to nearly seven years today. In France, France Invest’s H1 2026 activity study (published on 13 October) reaches the same order of magnitude for the recent period: an average holding period of 7.0 years for companies fully exited in the first half of 2026, against 6.1 years in 2022 (private equity excluding infrastructure, 2,543 companies exited between 2022 and June 2026); its series does not go back to the early 2000s and therefore cannot confirm EQT’s starting point. On another measure, the median, France Invest put the holding period at 4.2 years in 2010 (figure cited by PE Research in 2025).
In the EY survey, 24% of investors cite manager selection among their difficulties and 23% the assessment of their performance, while 65% want to combine large managers and niche managers.
Secondaries funds: more capital already returned to investors, less total value
In a study published in September, Apogem Capital, a US private equity and private debt manager affiliated with New York Life Investments, compares secondaries funds with North American buyout and growth funds using performance data from Cambridge Associates, a Boston-based investment consultancy that publishes benchmarks for private equity and venture capital funds.
| North American funds | Buyout and growth | Secondaries |
|---|---|---|
| Net IRR | 13.9% | 14.9% |
| Distributed / paid-in capital (DPI) | 0.73× | 0.80× |
| Total value / paid-in capital (TVPI) | 1.67× | 1.52× |
| Average standard deviation of IRRs across funds | 14.9% | 8.1% |
Apogem calculations based on Cambridge Associates data as of 31 December 2025. Averages of medians by vintage, weighted by number of funds, for vintages 2011 to 2023; average standard deviation by vintage, weighted.
Secondaries funds have already returned 80 cents per unit of paid-in capital, against 73 for buyout and growth funds. The latter, however, show a higher total value, including holdings still in portfolio. Apogem also notes much less dispersed IRRs among secondaries funds. Two caveats apply: the “risk” measured by Apogem is limited to this dispersion across funds, without accounting for liquidity or leverage, and recent vintages, whose assets were bought at a discount, mechanically favour secondaries. The investors surveyed by EY devote only 4% of their alternative allocations to secondaries; 38% expect this share to increase. EQT, through Coller EQT, and Apogem both offer secondaries strategies.
Sources: EQT, The New Rules of Capital, 21 Oct., p. 22 and chart 8 · France Invest, Activity of French private equity players, H1 2026, 13 Oct., p. 36 (report supplied to the editors) · PE Research, Analysis 18, 9 Sept. 2025 (in French) · Apogem, A Sharpe(r) Buyout?, Sept. · EY, Global Private Markets Survey, 18 Nov.
Read on PE Research: What is a continuation fund?
5 | United States — Small US companies: what do the 4,186 deals of 2025 measure?
United States / Small businesses
In the United States, the number of private equity deals below $25m rose from 2,355 in 2016 to 4,186 in 2025. Their cumulative value grew much more slowly. These figures come from the report published on 16 November by the American Investment Council (AIC), the organisation that represents the industry in Washington.
| Deals below $25m | 2016 | 2025 |
|---|---|---|
| Number | 2,355 | 4,186 |
| Cumulative value, in current dollars | $18.8bn | $21.9bn |
PitchBook for the AIC, data as of 23 July 2026. Between the two dates, the number rises by about 78% and the value by 16%, based on the rounded amounts published.
The AIC also states that 85% of the industry’s investments go to companies with 500 employees or fewer. Company size is measured here by headcount, whereas the table classifies deals by price. As for add-on acquisitions made by companies already in portfolio, they account for 72.9% of the total number of US buyouts in 2025, all sizes combined.
Source: American Investment Council, Supporting Small Businesses, 16 Nov., pp. 2 and 9
Read on PE Research: US private equity under fire
Two games to run a fund from the inside — by PE Research
- myBUYOUTfund: you run an LBO fund. With €250m over ten years (2027-2036), your fund buys French SMEs: you negotiate prices, structure the acquisition debt, work with management, create value and choose when to exit. Your net IRR is then compared with that of 200 simulated managers. Play myBUYOUTfund
- myVCfund: you run a venture capital fund. With €100m over ten years, your fund invests in European and US start-ups: you negotiate your term sheets (valuation, liquidation preference, anti-dilution), act on the board, defend your best holdings round after round, manage currency risk and run due diligence. The rule of the trade does not change: a handful of winners must pay for all the other positions. Play myVCfund
Both games are free in their basic edition, in English and French. They are educational tools: companies, figures and returns are fictitious; neither investment advice nor an offer of a financial product.
Get the weekly Private Equity review by e-mail (free): https://mougenotg.substack.com


