1. Background
2. How Contracts Work
3. Settlement
4. Station Map
5. Reading Prices
6. Fees
7. Gotchas
8. Edge Concepts

01 — BACKGROUND

Weather derivatives trace back to 1997, when Enron and Koch Industries executed the first over-the-counter (OTC) temperature swap. In 1999, the CME launched standardized HDD/CDD futures, primarily serving energy companies hedging seasonal demand.

The underlying data infrastructure dates to 1991, when the FAA, NWS, and DOD jointly deployed the Automated Surface Observing System (ASOS) network — 900+ airport weather stations across the US providing 1-minute temperature, wind, pressure, and visibility readings.

Kalshi, a CFTC-regulated Designated Contract Market (DCM), brought retail access to weather contracts. Instead of $20,000 CME futures notionals, you trade $1 binary contracts on whether a city's official high temperature will be above or below a specific threshold.

02 — HOW CONTRACTS WORK

Each Kalshi weather contract is a binary option that pays $1.00 if the condition is met, $0.00 if not. The contract price represents the market's implied probability of that outcome.

Contracts are structured as mutually exclusive range ladders. For a city's high temperature, you might see: "70–74°F", "75–79°F", "80–84°F", etc. The YES prices across all ranges should sum to approximately $1.00 (the sum-to-one constraint).

Markets typically open around 10:00 AM ET the day before settlement, with final settlement the following morning after the NWS CLI report publishes.

The order book is a Central Limit Order Book (CLOB) — you trade directly against other participants, not the house.

03 — SETTLEMENT

The temperature data flows through a three-stage chain:

1. Real-time METAR — 1-minute sensor readings, available immediately
2. Preliminary Summary — automated daily summary, usually available by 7 AM local
3. Final CLI Report — official NWS Daily Climate Report, published by mid-morning. This is what settles the contract.

Settlement delay conditions: Equipment malfunction, data review, or unusual weather events can delay the CLI. There is no recourse if the preliminary and final differ.

The F→C→F rounding problem: ASOS sensors record in Celsius. The conversion to Fahrenheit can cause 1°F rounding differences vs. what weather apps display. A reading of 23.9°C rounds to 75°F in the CLI, but some apps show 75.0°F from their own conversion.

DST trap: During Daylight Saving Time, the climatological observation day can extend into the next calendar day. The "high for April 6" might include readings from early April 7.

04 — STATION MAP

05 — READING PRICES

YES/NO prices represent the market's implied probability. A YES price of $0.72 means the market assigns a 72% chance the temperature will land in that range.

YES + NO prices for a single contract sum to approximately $1.00 (the difference is the spread).

• Tight spread (1–2¢) — liquid market, easy to enter/exit
• Wide spread (5¢+) — illiquid, may be stuck even if market moves your way
• Price near 1¢/99¢ — edge is gone, market has decided
• Price near 50¢ — maximum uncertainty, highest fees
• High volume — conviction signal, more reliable pricing

06 — FEES (effective Feb 5, 2026)

Kalshi fees follow a parabolic curve — highest at 50¢, lowest at extremes. This is counterintuitive: the most uncertain contracts cost the most to trade.

Zero settlement fee (as of Feb 2026). Zero membership fee. 2% debit card deposit fee only.

07 — GOTCHAS

1. Rounding trap — The F→C→F pipeline means 75°F on your app can settle at 74°F on the CLI
2. Wrong station — Denver airport ≠ Denver metro: 3–5°F difference
3. Preliminary vs final CLI — ~5–10% of days differ by 1°F, no recourse once final posts
4. Wide spread = stuck — Can't exit profitably even if market moves your way
5. GFS runs at 6am/6pm ET — Prices move 30 min after each run
6. Low volume ceiling — Can't size large without moving market against yourself
7. DST window — Observation day extends into next calendar day during DST
8. Fees highest at 50¢ — Not lowest (common misconception)

08 — EDGE CONCEPTS

NWS hourly vs daily: The hourly forecast curve gives better temperature resolution than the daily high/low. A city forecast for "high near 80" might show an hourly curve peaking at 78°F — the hourly data is more informative.

Model consensus: When GFS, NAM, and ECMWF agree, the forecast is more reliable. When they diverge, uncertainty is higher and contract prices may not reflect the full range of outcomes.

Cross-platform pricing gaps: Kalshi and Polymarket sometimes price similar weather events differently. The WxMkt Consensus panel highlights these gaps.

Sum-to-one constraint: Since all range contracts must sum to ~$1.00, going long one range and short an adjacent range is a relative value trade on where the temperature lands within a narrow band.