Kalshi Temperature Markets Explained: How Daily High Markets Work

Kalshi runs prediction markets on weather events, with daily high temperature markets for several U.S. cities. If you've heard about them but aren't sure how they work, this guide covers everything from basics to advanced patterns.

What Kalshi temperature markets are

Each market asks a simple question: What will the highest temperature at [city] be tomorrow?

The answer is divided into temperature brackets — roughly 1°F wide bins covering the realistic range of possible outcomes. For example, a Miami market in May might have brackets like:

• 79°F or lower (tail)
• 80-81°F
• 82-83°F
• 84-85°F
• 86-87°F
• 88-89°F
• 90-91°F
• 92°F or higher (tail)

You can buy YES or NO on any bracket. Each contract costs between 1¢ and 99¢ depending on market-implied probability. At settlement, winning contracts pay $1.00, losing contracts pay $0.

WeatherEdge tracks five Kalshi daily-high markets:

• KXHIGHMIA — Miami International Airport (KMIA)
• KXHIGHLAX — Los Angeles International Airport (KLAX)
• KXHIGHNY — New York Central Park (KNYC)
• KXHIGHCHI — Chicago Midway (KMDW)
• KXHIGHAUS — Houston Hobby (KHOU)

How settlement works

Markets settle based on the official NWS Climatological Daily Report (CLI), published several hours after the official observation period ends.

Each market specifies its station explicitly:

| Market | Station | Source | |---|---|---| | KXHIGHMIA | Miami International (KMIA) | NWS CLI for MIA | | KXHIGHLAX | Los Angeles International (KLAX) | NWS CLI for LAX | | KXHIGHNY | New York Central Park (KNYC) | NWS CLI for NYC | | KXHIGHCHI | Chicago Midway (KMDW) | NWS CLI for MDW | | KXHIGHAUS | Houston Hobby (KHOU) | NWS CLI for HOU |

The published peak is rounded to the nearest whole degree. So a "88-89°F" bracket resolves YES if the rounded peak is exactly 88 or 89.

This rounding matters: actual temperatures in the range [87.5°F, 89.5°F) round to 88 or 89 — that's the actual temperature window for the "88-89" bracket.

How prices form

Like most prediction markets, prices reflect the market's collective estimate of probability.

• YES at 30¢ means the market thinks the bracket has ~30% chance of resolving YES
• NO at 70¢ is the complementary view

A well-functioning market should have:

• YES + NO ≈ 100¢ (otherwise there's arbitrage)
• All bracket YES prices summing to ~100¢ (since exactly one will win)

In practice, you'll see small deviations due to spreads, fees, and inefficiency.

What moves prices during the day

Temperature markets have an interesting property: the underlying truth is being revealed continuously. This creates a predictable price arc.

Morning (6-10 AM local)

Prices reflect overnight NWS forecast and climatology. Brackets cluster around the expected peak with reasonable spread across adjacent bins. Spreads are wide because liquidity is thin.

A typical morning state might look like:

• 86-87°F: 22¢ YES
• 88-89°F: 35¢ YES (NWS forecasts 89°F)
• 90-91°F: 28¢ YES
• 92°F+: 8¢ YES

Late morning (10 AM - 12 PM)

Observations start to inform the market. If temperatures are running ahead of climatology by mid-morning, higher brackets gain probability. If overcast or marine layer holds in, lower brackets gain.

Early afternoon (12-2 PM)

This is when most price discovery happens. Volume picks up. The likely winning bracket clarifies. Adjacent brackets crash.

Mid-afternoon (2-4 PM)

Peak temperature is typically reached at these stations. The winning bracket climbs to 80-95¢ YES. Others approach zero.

Late afternoon (4 PM - close)

Convergence to settlement. Winning bracket trades 95-99¢. Spreads near zero. Almost no edge remains.

Key concepts

Spread

The difference between the best bid (where someone will buy) and best ask (where someone will sell). Tight spreads (1-3¢) indicate liquidity. Wide spreads (10¢+) indicate illiquidity and difficulty entering/exiting positions.

Edge

The difference between your estimated probability and the market price. If you think a bracket is 40% likely and it's priced at 28¢ YES, you have +12¢ of edge (in your favor).

Realistic edges in liquid Kalshi temperature markets typically range from 0 to 8¢. Anything claiming larger edges deserves skepticism — either your estimate is wrong or the market sees something you don't.

Settlement risk

The risk that the market resolves differently than expected due to factors after your trade. For temperature markets, this includes:

• Late-day temperature spikes you didn't anticipate
• Discrepancies between intraday observations and official CLI peak
• Edge cases at bracket boundaries (e.g., peak of 87.5°F could round either way)

Common strategies

The morning edge play

Buy YES on a bracket where your model says higher probability than market price suggests. Win if your model edge is real. Lose if it's not. High variance, high return per win.

Typical setup: model says 35%, market says 25%, you buy YES at 25¢. If it resolves YES, you make 75¢ per contract. If NO, you lose 25¢.

The late-day convergence play

Buy YES on the likely winning bracket at 70-85¢ in mid-afternoon when peak is essentially set. Smaller per-trade profit but higher win rate.

Typical setup: bracket trades 78¢, current peak observed is already in the bracket range, NWS hourly forecast shows minimal further heating. Buy YES, expect 90% to resolve at 100¢.

The boundary play

When current peak is sitting right at a bracket boundary (e.g., observed 89°F with brackets at 88-89 and 90-91), small temperature differences become big P&L differences. These trades are 50/50 but priced as if more certain. Skilled traders find edge here.

Things that catch newcomers

Forgetting that settlement is based on station, not city. "New York" temperature in casual conversation might mean Times Square or LaGuardia. Kalshi explicitly uses Central Park (KNYC). On heat-wave days, LaGuardia can be 5°F+ warmer than Central Park. Trade what's specified, not what you'd guess.

Confusing CLI with current observation. Markets resolve on the official NWS CLI report, not real-time observations. The CLI peak can differ from what you saw at 4 PM if a late spike happened.

Overpaying for "safe" late trades. A bracket trading 92¢ YES still has 8% implied risk. If you're buying it to "play it safe," you're paying $0.92 to make $0.08. One wrong call costs you 11 wins to recover.

Ignoring spread on the side you're buying. Bid 75¢, ask 88¢. If you buy at the ask, your real cost is 88¢. Always check the side you're entering, not the mid-price.

Where to trade

Kalshi is the primary regulated U.S. exchange for these markets. As of 2024, Kalshi was designated as a CFTC-regulated DCM (Designated Contract Market), which gave it federal authority to operate prediction markets nationwide.

Other platforms like Polymarket also offer weather markets, but they're geoblocked from U.S. users and have different settlement rules (different stations, in some cases).

Want help analyzing these markets?

WeatherEdge is built to make analyzing Kalshi temperature markets easier. It runs a daily-high prediction model alongside live Kalshi prices, highlights brackets where the model and market disagree, and includes a paper trading simulator to practice scenarios without putting real money down.

Free tier shows live KHOU data. Pro tier ($12.99/mo) adds all 5 stations + paper trading. Max ($23.99/mo) adds an AI coach that explains why each trade worked or didn't.

Not affiliated with Kalshi. Educational analysis only — not investment advice.