Why execution matters as much as your forecast in weather markets, and how to decide when to cross.
Most temperature traders obsess over the forecast. Which model is right, whether the pace is running hot, whether the NWS is biased in this city. That work matters, but there's a cost sitting in front of every trade that the forecast never sees: the gap between what buyers are bidding and what sellers are asking. Kalshi shows you its fee. Nobody shows you the spread's.
Every bracket has two prices: the best bid and the best ask. The middle of the two is the closest thing the market has to a fair price. When you click buy at the ask, you're paying above that middle. A bracket at 49 bid / 51 ask costs you 1 cent above fair; at 40 bid / 50 ask, 5 cents above fair, before the exchange fee.
Kalshi's taker fee is roughly 7% of P × (1 − P) per contract: about 1.75 cents at 50 cents, less toward the extremes. The break-even on a crossed trade is roughly: what you think the bracket is worth > the ask + the fee. In the 40/50 example you need to believe the bracket is worth 52 cents or more, seven cents above the market's own middle.
Temperature markets split one number across a ladder of two-degree brackets (one degree for international cities on Polymarket). Busy same-day high markets often trade with a 1–2 cent spread near the action; low markets routinely run around 10 cents, with a fraction of the size near the best price.
If you want 20 contracts and only 11 are offered at the best price, your order either stops there or keeps buying up the ladder. A thin book with a gap raises your average price; a stacked book gives you size, but ask why so many are happy to sell there.
Resting a limit order saves the spread and usually pays little or no fee, but you tend to get filled when you're wrong. Market-makers call this adverse selection, and in thin markets it's sharp: the longer a resting order sits, the worse its fills tend to be.
Is there a clock? Is your edge bigger than the spread plus the fee? How much is actually offered? Has the price already moved? If you rest, set a timer.
A good forecast gets you to the starting line. Whether a trade makes money often comes down to how you enter it: the spread you pay, the fee on top, the depth you walk through, and the fills you do and don't get.