Most years, something the size of a small house slams into Earth's atmosphere and explodes in a flash brighter than the sun. Nobody sees it, because it happens over open ocean or empty sky, and the only record is a row of numbers logged quietly by NASA's Center for Near Earth Object Studies. That database is the resolution source for one of the stranger contracts on Polymarket right now: a single binary question about whether 2026 will deliver a bolide above a specific energy threshold.
The 5kt meteor strike market on Polymarket asks something deceptively simple. Will any natural meteoroid explode in Earth's atmosphere during 2026 with a total impact energy of at least 5 kilotons of TNT equivalent? Yes or no, decided by the JPL Fireball and Bolide Data repository, settled at the end of the calendar year. As of 19 June 2026, the contract favours No, trading around 74%, with Yes near 27%.
What 5 kilotons actually means
For scale, the bomb dropped on Hiroshima released roughly 15 kilotons. So the threshold here is real but not apocalyptic. It is the kind of event that produces a fireball, a sonic boom across hundreds of miles, possibly some broken windows if it happens to detonate over a populated area, and a flurry of dashcam footage uploaded to social media within hours.
The Chelyabinsk event in February 2013, which injured around 1,500 people in Russia from blown-out glass, was orders of magnitude bigger. JPL logged that one at roughly 440 kilotons. Events at the 5kt floor are far more frequent, and most go entirely unnoticed because the planet is mostly water and most atmospheres above land are empty of cameras.
What the contract is really doing is putting a price on the base rate. How often does this happen in a given year? The bolide database stretches back to 1988 and has hundreds of entries. A reader who wants to understand the analytical scaffolding can dig into how prediction market odds work, because the implied probability here is essentially a statement about historical frequency rather than any forecastable event in the conventional sense.
Why this question is structurally interesting
Most prediction markets price something a trader can plausibly know more about than the average person. An election contract rewards political nous. A football contract rewards understanding of form and lineups. A Fed contract rewards understanding of central-bank communication.
A meteor contract is different. Nobody has private information about when a previously undetected rock will arrive. The asteroids large enough to be tracked years in advance are catalogued and would already be priced into the contract; everything else is, by definition, a surprise. So the trade is almost purely a statistical one. You are betting on how a Poisson-ish process behaves over a calendar year, with the small twist that the resolution source has its own detection sensitivities and reporting lags.
That makes it an unusually clean test of what implied probability actually represents. There is no candidate to spin, no team to injure, no central banker to interpret. There is only a number that should approximate the long-run frequency of 5kt-plus bolides per year, plus whatever risk premium the market wants to charge for variance.
The resolution source matters more than it looks
One quietly important detail is the source: NASA's JPL Fireball and Bolide Data repository, with the Impact Energy field as the determining figure. That database is authoritative, but it is also slow. Events get logged after analysis, and the published energy figure can be revised. For a contract resolving at year-end, that lag is unlikely to cause drama, but it is worth knowing the rule before clicking.
The contract also excludes anything artificial. A failed satellite reentry or a spent rocket stage burning up does not count, no matter how spectacular. This is the kind of edge case that matters in space markets generally, and it is the sort of clause where reading the resolution rule carefully saves a lot of grief. The contract specifies a natural meteoroid, classified as such, with the impact energy taken from a specific field of a specific database.
The other structural feature worth flagging: this is a yearlong observation window with binary settlement. Either one qualifying event happens by 11:59pm ET on 31 December 2026, or none does. A single fireball over the South Pacific in October would settle the entire question, regardless of how quiet the rest of the year has been. That asymmetry is part of what makes the contract interesting; one event resolves Yes for the whole calendar, and nothing short of the full twelve months resolves No.
The editorial take
There is no urgency here, and that is the point. This is not a market reacting to a news event. It is a market that exists to put a number on a question almost nobody bothers to ask: what is the annual chance of a small but newsworthy bolide. The value of the contract is less in its likely resolution and more in the fact that it exists at all, forcing a price onto an otherwise vague intuition about how often the sky does something dramatic.
Markets like this one are also a quiet rebuttal to the idea that prediction venues only price politics and sport. iPredicta tracks contracts across the full breadth of what gets listed, including the obscure climate and science questions like this one, because the structural lessons travel: how a resolution source is chosen, how a binary window behaves, and how the base rate of rare events gets translated into a tradeable number.
Frequently asked questions
How often do 5kt-plus bolides actually happen?
NASA's bolide database, which stretches back to 1988, logs events of various energies every year, and impacts at the small-kiloton scale are relatively common when measured against the long record. Most happen over ocean or sparsely populated areas and get noticed only by satellites. The contract is essentially pricing how that base rate translates into a single calendar year.
What would actually resolve this market Yes?
A single natural meteoroid exploding in Earth's atmosphere with at least 5 kilotons of TNT equivalent impact energy, between 1 January and 31 December 2026, as recorded in the Impact Energy field of NASA JPL's Fireball and Bolide Data repository. Artificial objects and reentry vehicles are explicitly excluded. One qualifying event at any point in the year is enough.
