There is a specific kind of market-cap question that used to be answered by a glance at a Bloomberg terminal on the last day of the quarter. Now it lives on Polymarket, gets priced in real time by thousands of traders, and settles on the reading of a single tape at a single moment. The largest-company contract on Polymarket asks exactly that. Which company is worth the most in the world when the closing bell rings on 30 September 2026?
That is a narrower question than it sounds. The contract does not care about revenue, profit, headcount, strategic importance, or which chief executive is on the cover of a magazine that week. It cares about one number, on one date, sourced from a consensus of credible reporting. Everything interesting about the market flows from that constraint.
Deposit $20, get $50 to trade
Welcome bonus on the world's largest prediction market.
18+ · New users only · Availability and terms vary by region · Trading involves risk of loss · BeGambleAware.org
A snapshot, not a trend
The most important thing to understand about this contract is that it is a moment, not a period. A company could lead the market cap tables for eleven months of the year and still lose the contract to a rival that overtakes it in the final week of September. A single earnings surprise on 29 September could flip the resolution. A single guidance cut could do the same.
NVIDIA is priced as the clear favourite at 92.5% as of 17 August, with Apple at 4.5% and Alphabet at 2.5% on the same read. Amazon, Tesla, Saudi Aramco, Broadcom and Microsoft each sit at or below 0.4%. SpaceX appears on the ladder but is unquoted, with no usable two-sided book to read.
Count the names carefully, though, because the ladder has twenty-five entries and only nine of them are a market. The other sixteen are unfilled slots, quoted bid zero against ask one hundred, with no volume and placeholder labels rather than company names. They are the contract's construction showing through, not an opinion anyone holds. So the real shape is one heavy favourite, eight genuinely quoted names behind it, and sixteen slots nobody has taken a position in at all. A reader who counts the outcomes should know which of those they are counting.
The contract has traded $237,782 across its life and $38,568 in the 24 hours to 17 August. One mechanical note before anyone adds the quoted percentages up: the nine real legs sum to a little over 100, which is the spread rather than an error, since those are mid-points and the buy side of the book sums to just under 100 against a sell side above it.
The lean is unmistakable. Whether it survives six weeks of earnings, guidance, and the ordinary volatility of a mega-cap tape is a different question, and one the contract cannot answer in advance. That is the structural interest of a snapshot resolution: the closer to the date, the more the price should compress around whichever name is actually in front on the day; the further out, the more it reflects the market's read on a moving league table.
Why the resolution mechanism matters
A consensus of credible reporting is a deliberately soft resolution source, and it is worth flagging why. Market capitalisation is a calculated number, not a published one. Shares outstanding change with buybacks and issuances. Closing prices are unambiguous for listed equities but ambiguous for anything that is not currently trading on a major exchange. And the ranking depends on which currency you convert into and at what rate.
For most of the names on this ladder the calculation is uncontroversial. NVIDIA, Apple, Alphabet, Amazon, Microsoft, Tesla and Broadcom all close on major US exchanges, in dollars, and their share counts are public. Saudi Aramco is the awkward one on the listed side, priced in riyals on the Tadawul with a share float structured very differently to the US mega-caps. SpaceX is the more genuinely difficult case, because it is private, and any market-cap figure attached to it comes from secondary transactions and reported funding rounds rather than a live public tape. That is almost certainly why the market cannot settle on a two-sided price for it.
The resolution rule handles all of this by outsourcing the judgement. If credible reporting on 30 September broadly names one company as the largest in the world, that name wins. If reporting splits, or if the ordering is contested, the oracle process has to reach a decision anyway. Traders pricing SpaceX, or pricing an unlikely late surge from Aramco, are also pricing the probability that the oracle would accept the underlying calculation as valid. That is a second-order bet layered on top of the first.
The shape of the tail
NVIDIA accounts for under a quarter of the volume traded on this contract: $55,986 of the $237,782 that has changed hands across its life, on the 17 August read. That is worth stating precisely rather than loosely, because volume is not capital at risk and the two are easy to blur. It does not tell you where the money is sitting; it tells you where the trading is happening, and three quarters of it is happening away from the favourite. Most of the interesting analytical work is not on NVIDIA either. It is on the tail, where names sit at percentages small enough to look like typos and large enough to matter if you understand what they are pricing.
A sub-1% price on Amazon or Microsoft is not a claim that either company has failed. It is a claim that within six weeks, the gap between the leader and either of them is unlikely to close enough for the ordering to flip on 30 September. That gap depends on a lot of moving parts: the AI infrastructure narrative that has driven the leader's valuation, the trajectory of chip demand from hyperscalers, the mix of buyback and issuance across the mega-caps, and whether any of them delivers a genuine earnings shock between now and quarter-end. Reasonable people can read the same tape and arrive at different probabilities for how quickly that gap could close.
The fact that the tail is priced so thin, and that a wide roster of companies sits below 1%, tells you the market has picked a lead and is willing to bet against every plausible alternative simultaneously. It does not tell you the market is right.
What the contract can and cannot tell you
Used carefully, a market like this is a useful summary statistic. It compresses the aggregate view of thousands of traders about the ranking of the world's most valuable companies into a single number per name. It updates in real time as news hits. It costs nothing to read. That is more information than most retail investors will assemble from any other single source.
Used carelessly, it is misleading. A 93% price is not a promise. A sub-1% price is not a dismissal. Neither is a forecast of business quality, competitive position, or long-run prospects. All either one is telling you is what the market thinks the tape will show on one specific afternoon at the end of one specific quarter.
At iPredicta we track markets like this because the shape of the pricing tells you something useful about how the market weighs concentration risk at the top of the world's equity table, and because a single-date snapshot contract is one of the cleanest ways to see what a crowd of traders actually believes about an ordering they can watch update every second. The largest-company market on Polymarket is one of the questions we keep an eye on for exactly that reason.
Frequently asked questions
What exactly does this Polymarket contract resolve on?
It resolves to whichever company is the largest in the world by market capitalisation at the close of trading on 30 September 2026, based on a consensus of credible reporting. It is a single-moment snapshot, not an average across the quarter, so late moves in September can flip the outcome even if the ranking was stable for months before.
Why is SpaceX listed but unpriced?
SpaceX is a private company, so there is no live exchange tape and its market cap is estimated from secondary transactions and reported funding rounds rather than a public share price. That makes both the underlying number and the eventual resolution harder to pin down, and the market has not settled on a usable two-sided price for the leg.
