A company is worth $50 trillion

Where things stand

No company is anywhere near $50T. On CompaniesMarketCap (https://companiesmarketcap.com/, fetched 12 Oct 2026) the largest companies are Nvidia at $5.54T, Apple at $4.91T, Alphabet at about $4.3T and Microsoft at $3.97T. SpaceX listed on 12 June 2026 at a $1.77T valuation and is now worth $2.14T. Anthropic is preparing an IPO that could value it at more than $2T (Yahoo Finance, early Oct 2026). OpenAI has pushed its IPO to 2027 and is valued privately at roughly $1.2–1.4T. To reach $50T, the current leader would have to grow about 9x by the end of 2031, roughly 54% a year. The newer AI listings would have to grow 20–35x.

Base rate (computed)

I took year-end market caps for 12 megacaps from CompaniesMarketCap's history pages and tracked the largest company in each year from 1996 to 2026:

• The largest company's value grew by an average of 0.116 a year in log terms, with a standard deviation of 0.276.
• The best 5-year gain was 4.76x (2016 to 2021).
• No company at the top has ever grown 9x in 5 years.
• The record market cap has doubled about every 3.5 years ($1T in 2018, $4T in 2025). On that trend, $50T arrives around 2037.

So the historical base rate is well under 1%. A YES essentially requires a regime change.

What markets imply

Implied volatility on Yahoo's January 2029 call options is about 42–44% for Nvidia, about 40% for Alphabet and about 57% for SpaceX. I assumed log-normal price paths with a chance of touching $50T at any point, covering 12 candidates including Anthropic and OpenAI after their IPOs, and ran a correlated Monte Carlo (correlation 0.55). The chance that at least one company closes at or above $50T by the end of 2031 is:

• 2.4% using option-market (risk-neutral) pricing;
• 4.3% adding a 5% equity risk premium;
• 6.0% adding an 8% premium.

In every version the probability is close to zero before 2029. Mainstream spending forecasts point lower still. Goldman expects big-tech AI capital spending of $1.2T in 2027 and $1.4T in 2028 (Bloomberg, 25 Sep 2026), which would put Nvidia at roughly $10–20T by 2031. Getting to $50T needs about $1–2T of annual profit, or a valuation multiple like the 2000 bubble on top of a very strong boom.

AI takeoff and other forecasters

There are real signs that AI is speeding up:

• Nvidia's quarterly revenue was $96.2B, up 106% on a year earlier, and it is guiding to about 70% growth next fiscal year (https://nvidianews.nvidia.com/news/nvidia-announces-financial-results-for-second-quarter-fiscal-2027).
• Anthropic's annualized revenue reached $65B in July and is expected to pass $100B in 2026.
• In a September 2026 essay, Anthropic's chief executive Dario Amodei says AI systems have started to improve the next generation of AI themselves ("recursive self-improvement") (https://darioamodei.com/post/we-must-pace-the-frontier).

Against that, Nvidia's stock is up only about 13% over the year even though its revenue doubled. The 10-year Treasury yield is 5.24%. OpenAI's revenue figure was revised down this month.

Other forecasts:

• Metaculus: the community's median date for Nvidia reaching $10T is February 2032, and it gives 22% to a US company reaching $10T before 2028.
• Manifold: 14% for a $30T company before 2030.
• Kalshi: about 13% that Nvidia CEO Jensen Huang becomes the second trillionaire before 2033.
• James Anderson, a well-known Nvidia bull, puts the odds of Nvidia reaching about $49T by 2035 at 10–15%.

My scenario estimate: a 12–15% chance that markets recognise an AI takeoff by 2029, times a 30–40% chance that one company then reaches $50T, plus a small chance of a pure market mania, gives about 5–6%.

Forecast

I put the chance at 0.2% by the end of 2027, 0.5% by the end of 2028, 1.5% by the end of 2029, 3.3% by the end of 2030 and 5.7% by the end of 2031. The quarterly chance rises because, short of an AI takeoff, a 9x gain needs several years.

The strongest case for a higher number is that AI revenue growth has no precedent, new high-volatility AI companies are listing, and prediction markets are more bullish than my estimate. The case for a lower number is that nothing like this has happened before, size limits bite, and in a real takeoff interest rates would rise, competition and regulation would grow (including the industry "pacing" Amodei proposes), and the value would be split across many firms.

A YES would be reported immediately by Bloomberg and Reuters, so delays in publication do not matter here.

Question. By the last day of each quarter, has any company closed a trading day with a market capitalisation of US$50 trillion or more?

Definitions.

• Company: a single publicly traded corporate issuer anywhere in the world, including a listed holding company or state-controlled company. It need not be an AI company or a covered developer. Funds, ETFs, cryptocurrencies, and sums of separately listed companies do not count.
• Market capitalisation: total outstanding common-equity value, not enterprise value, free-float capitalisation, or a fully diluted valuation. Multiply outstanding shares, including restricted and government-held shares but excluding treasury shares and unexercised options, by the official regular-session closing price. Sum distinct common-share classes at their respective closing prices; value an unlisted class at its listed economic equivalent, adjusted for its conversion ratio. Do not double-count cross-listings or depositary receipts. Use the primary listing for each class and the latest completed regular-session close as of the close being tested. Share counts must be applicable to that trading date, with documented issuances, cancellations, and splits reflected.
• $50 trillion: US$50,000,000,000,000 in nominal dollars, without inflation adjustment. For non-dollar shares, use the same trading date's closing exchange rate reported by Bloomberg or LSEG; use Bloomberg if both are available and disagree, and the latest preceding business day's rate if no rate is published for that date.
• Close a trading day: the official regular-session close on a public stock exchange, including its closing auction. Intraday, pre-market, and after-hours prices do not count.

Resolves YES if a qualifying dated source establishes a closing market capitalisation at or above the threshold for at least one company, either explicitly or through closing prices, applicable share counts, and exchange rates sufficient to calculate it. A report explicitly stating that the company closed at or above $50 trillion counts; a rounded display of “$50T” alone does not if its underlying value could be below the threshold. The Observation period convention applies: the qualifying close must occur after the opening date.

Resolves NO if no qualifying publication is available by the horizon. Private financing valuations, IPO offer valuations without a qualifying trading close, forecasts, intraday crossings, and valuations of entire corporate groups obtained by adding separately listed issuers are excluded. A merger counts only once the resulting issuer actually has the qualifying closing capitalisation.

Resolving source. Official exchange closing-price and capitalisation records and issuer regulatory filings are primary evidence. Dated Bloomberg, Reuters/LSEG, or Financial Times reports or historical market-data records also qualify when explicitly tied to a trading close and consistent with the definitions above. Official exchange prices and issuer-filed share counts prevail over conflicting secondary figures. The CompaniesMarketCap global ranking may identify candidates but an undated or intraday ranking alone is insufficient. These financial sources govern rather than Public confirmation (A2). Apply the Resolution window convention and Resolution date (A3): resolution is dated to publication of the qualifying evidence, not backdated to the trading session. Later-published corrections or historical data can first qualify on their publication date; previously qualifying YES resolutions remain YES under the site convention.