Can you bet on a heat wave? Yes. Kalshi lists daily temperature contracts on cities across the country, and a forecast heat wave reprices those ladders days before the hottest afternoon arrives: the high bands that were priced as long shots start getting expensive, city after city, all at once. That "all at once" is the real subject of this piece. A heat wave is not a city event, it is a regional event, and that single fact is the sharpest lesson in portfolio risk these markets have to offer. A trader can hold positions in fifteen different cities, feel diversified, and be holding one large bet without knowing it. For a five-city book priced like the one worked below, an all-five-lose evening can be as much as 400,000 times more likely than multiplying the five prices together would suggest.
The Quick Answer
You can trade heat-wave exposure through Kalshi's daily city temperature markets, taking either side of whether a city's high crosses a given band. The catch is correlation: a heat dome is one air mass sitting on an entire region, so contracts on different cities that look like separate bets can all settle the same way on the same evening. The real weather pattern that proves it, the arithmetic of what correlation does to a book of sold long shots, and what actual diversification looks like are all below.
Free: The Weekly PM Market Brief — the week's biggest market-vs-model gaps and what's spiking next, across the whole prediction-market board. One email, Sundays.
One Air Mass Moves Many Cities Together
The weather pattern behind almost every major summer heat event is a heat dome: a large, stalled ridge of high pressure that traps hot air underneath it and compresses it, like a lid on a pot. The dome does not sit over a city. It sits over a region, often several hundred miles across, and everything underneath it cooks together. Sinking air suppresses clouds, clear skies feed more heating, and the ridge holds the whole system in place for days.
For a weather trader the physics matter less than the consequence: one air mass moves many cities together. When a dome parks over the Pacific Northwest, Portland and Seattle are not having independent weather that happens to coincide. They are having the same weather, measured at two settlement stations. The same is true of a dome over Texas, or the Southeast, or the Northeast corridor. The map has many city names on it; the atmosphere is running one process.
Every Kalshi city market prices its own ladder of temperature bands, each contract settling $1 or $0 against its own named station. On the screen, Portland and Seattle are separate markets with separate order books and separate prices. Under a dome, they are the same trade wearing two tickers. Which raises the question this piece exists to answer: what happens to a book built as if they were separate?
A Real One: The June 2021 Heat Dome
The cleanest real-world illustration of correlation in weather is the heat dome that settled over the Pacific Northwest in late June 2021. A ridge of high pressure stalled over a region famous for mild summers, and the records did not just fall, they fell in formation.
Portland broke its all-time record three days in a row: 108 degrees, then 112, then 116. Seattle hit 108, five degrees past its previous all-time mark. Across the border, Lytton, British Columbia reached 121 degrees, the hottest temperature ever recorded in Canada. Dozens of stations across Oregon, Washington and British Columbia set all-time highs inside the same three-day window.
Look at that event the way a trader has to. Before the dome arrived, "Portland above 110" and "Seattle above 105" would have priced like absurdities, tails on the far edge of two different ladders. They were not two absurdities. They were one absurdity, the ridge itself, expressed at two stations, and when the ridge materialized, every one of those far-tail outcomes landed together, in the same 72 hours, across an entire region. A book short those tails in five Northwest cities would not have lost one position that week. It would have lost the book.
That is what correlation means, stripped of the statistics vocabulary. Nothing about it requires 2021 to repeat, either: any future regional heat dome would do to some region's ladders exactly what that one did to the Northwest's.
A Worked Example: Five Cities Under One Ridge
Now put arithmetic on it, because the numbers are the argument. A seller holds far-tail heat bands in five different cities, each priced at 4 cents, meaning the market puts each outcome at roughly 4%. If those five outcomes were truly independent, the chance of all five hitting on the same evening would be 0.04 multiplied by itself five times: about one in ten million. Priced that way, the wipeout scenario is not worth a thought.
But if the five cities sit under one ridge, the five contracts are not five separate questions. They are one question, does the dome deliver, asked five times. Under near-perfect correlation, the honest worst case for a book under one dome, the chance all five resolve against the seller is no longer one in ten million; it is roughly the chance of the heat event itself, on the order of 4%, one in twenty-five. A real dome delivers correlation somewhere short of perfect, so the truth sits between the two rows below, but it lives far closer to the bottom one than any independence math admits.
| The Same Five 4¢ Tails | Chance all five lose together |
|---|---|
| Five Independent Cities | about 1 in 10,000,000 |
| Five Cities Under One Air Mass | about 1 in 25 |
One row of that table is the trade the seller thinks they made. The other is far closer to the trade they actually made, and the gap between the rows is a factor of roughly 400,000. Nothing about the individual prices was wrong. Each contract at 4 cents could be perfectly fair on its own. Correlation does not live inside any single price; it lives between the prices, which is why apparently independent contracts can resolve together while every one of them was priced correctly in isolation. No single market on the board will ever warn you about it.
Diversification That Isn't
This lands hardest on sellers because of the shape of the trade. As we walked through in the asymmetric payoff piece, selling an unlikely outcome collects a few cents of premium against most of a dollar of risk, so one loss erases the premiums from roughly thirty or forty wins. Sizing is the entire game for that style of trading, and correlation quietly breaks the sizing. A seller who caps each position at "a bad day in the log" has done the arithmetic per position. Under a dome, five of those bad days arrive in one settlement window, and a sizing plan built on independence turns out to have been five times too large.
So the honest unit of risk in a weather book is not the position. It is the air mass. Every contract that one system can settle is, for sizing purposes, one contract, and adding a sixteenth city under the same ridge is not diversification, it is doubling down with a different ticker. What actually spreads risk in these markets:
- Different Climate Regions, far enough apart that one ridge cannot cover them. A Northwest dome says nothing about Florida's afternoon.
- Both Directions. A book that is short heat everywhere loses everywhere when the heat comes. Positions on both sides of different ladders cannot all be beaten by one pattern.
- Both Books. Daily highs and overnight lows are set at opposite ends of the day by partly different physics; a dome punishes both, but unevenly, and a humid city and a dry city under the same heat behave differently again.
The item I keep coming back to is the second one, because it is the one sellers resist. Being short the same direction in fifteen markets feels like fifteen small opinions. The 2021 map is the reminder that the atmosphere is happy to grade all fifteen with one stroke.
Count the systems, not the cities. The honest unit of risk in a weather book is the air mass. Every position one ridge can settle is one position for sizing purposes, and a sixteenth city under the same dome is not diversification, it is more of the same bet.
What The Board Does While The Dome Approaches
Correlation also changes how a heat wave trades before it arrives. Weather models see a major ridge building days out, every trader on the platform reads the same public forecast, and the repricing happens across the whole region at once: far tails that were 3 cents are suddenly 15, then 40, as the model runs converge. There is no kickoff and no closing bell; the market just drifts toward the answer as evidence arrives, which is one reason these markets track reality as well as they do and one reason nobody should grade themselves on a single dramatic week. By the time the dome is on the doorstep, the "long shot" prices are gone, and with them the version of the trade this piece warns about. The dangerous version is sold early, in calm weather, at calm-weather prices, which is exactly when a whole region's tails are cheapest to accumulate and the book looks most harmless.
Worth saying plainly: none of this is a pick, and this page is not telling you which side of any band to be on. It is telling you what the board does to people who treat a regional event as fifteen local ones.
The Same Correlation Lives In A Parlay
If you bet sports, you already own a version of this problem. A parlay is correlation you assembled on purpose: several legs that must all win, where the honest question is how related the legs really are. Sportsbooks price same-game legs as related because pretending otherwise would bankrupt them, which should tell you something about how seriously the professionals take resolution-together risk.
Run the five-city arithmetic on a Sunday card and the direction of the error is identical. A bettor takes the over in six different outdoor games, feels spread across six matchups, and calls that diversification. If a cold front is sitting on four of those six stadiums, those four legs are partly answering one weather question. Treated as six independent coin flips, an 0-for-6 Sunday looks like a one-in-sixty-four disaster; to the extent the front binds four of them together, the real number is worse than that, and the bad Sunday shows up more often than the multiplication promised. A front is a much looser link to a game total than a dome is to a thermometer, so the honest claim here is the direction of the mistake, not a second precise figure. Same lesson, slower motion: a bettor spreading futures across a single division is running one bet with four tickets on it.
Clean per-leg pricing is worth having, and OddsShopper's top bets screen shops every line across every major sportsbook available in your state and strips the vig out to show the fair number. But notice what that does and does not solve, because it is this article's whole argument: fair pricing fixes the leg and does nothing at all about the correlation between legs. Each of those five 4-cent contracts was fairly priced too. The Parlay Builder gives you legs that have already been through the fair-price check, which is the leg-level half of the job; deciding how many of them share one driver is still yours, and that is the half that decides whether a bad day is survivable. If you would rather watch that thinking applied to actual games first, our free expert picks cost nothing and show the process in public.
OddsShopper Pro includes a free trial week, so you can look at how the tools price a card before your first payment, and the code HEATWAVE20 takes 20% off that first payment if you continue past the week.
Heat Wave Markets FAQ
Can you bet on a heat wave? Yes. A Kalshi heat wave trade is a position on daily temperature contracts, which are CFTC-regulated event derivatives rather than sportsbook wagers. Each city's market settles $1 or $0 against a named weather station, and you can take either side of any band. Our guide to how weather markets work covers the mechanics from the start, and the fee schedule covers what a trade costs.
Why do weather contracts on different cities move together? Because the cities share weather systems. Contracts settle at separate stations and trade in separate order books, but one ridge or one front drives the temperature at many stations simultaneously, so the outcomes are correlated even though the markets are mechanically independent. A dome that holds for days can settle the same way across a region several times in a row, which compounds the problem for anyone positioned the same direction everywhere.
Does a heat wave make weather markets easier to profit from? No. Everyone reads the same public forecasts, so regional heat gets priced into every affected ladder quickly, and an exchange does not bake sportsbook-style vig into the price, though it does charge a stated fee you can see. What a heat wave mostly does is reveal whose book was secretly one bet. We trade these markets ourselves and we do not publish performance figures; the rest of the series lives on our weather markets hub. Whether this kind of trading is investing or gambling is a fair question with a longer answer.
Count The Systems, Not The Cities
A heat wave does not change what any single weather contract is. It changes what a collection of them is, collapsing many tickers into one outcome, which is why the five-city book at the top of this piece was never a one-in-ten-million risk and sits far closer to a one-in-twenty-five one. The atmosphere does not care how your positions are labeled; it settles by air mass. We trade these markets ourselves and we do not publish performance figures, and the rest of the series sits on our Kalshi weather markets hub. The sizing rule we hold ourselves to is the one this piece argues for: count the systems, not the cities.
Disclosure and fine print. Stokastic trades Kalshi weather markets and holds positions in them; where a settled position is described in this series, we were the seller. We have no affiliate or commercial relationship with Kalshi, and we do carry sign-up offers for some other prediction-market and betting platforms. Kalshi weather contracts are CFTC-regulated event derivatives traded on a designated contract market, not sportsbook wagers, and a position can lose its full value. Selling an unlikely outcome collects a small premium and risks most of a dollar, so one loss can erase the premiums from thirty or forty wins, and correlated positions can lose together on the same evening; size accordingly. 18+, available where Kalshi operates; the risk of loss is real and, on the side we trade, individually large. This series is an open research log of a strategy we have not proven, and we do not publish performance figures for it. Nothing here is trading advice, and nothing on this page is a pick or a recommendation.
Trade prediction markets on Polymarket too. Most readers here already have Kalshi — Polymarket is the other major venue, and code OS4 on the Polymarket US App: Deposit $10, get a $20 trading bonus. Get the Polymarket US bonus →
Affiliate disclosure: OddsShopper earns a commission on signups through this link. Terms as stated by Polymarket; 18+, availability varies by state.
18+ Only. Restrictions and eligibility requirements apply. Not available in all jurisdictions. Trading is risky. 100% loss can occur. See polymarket.com/tos for more information. The Polymarket US App serves as an independent software provider and affiliate of Polymarket US and Polymarket Clearing, the CFTC-regulated exchange and clearing organization.
