Fair Value Weather & Climate — Wednesday, July 29, 2026

A detailed daily look at some of the most interesting weather and climate prediction markets at ForecastEx and elsewhere.

• Access to live probabilities and the ability to trade is available via IBKR Prediction Markets.
• General primers on how these prediction markets work are available in the IBKR prediction markets FAQ (scroll down) and the ForecastEx FAQ.

1) In the news, a record heat dome in the Southwest

An Extreme Heat Warning issued Tuesday night covers the Coachella Valley, the San Diego County deserts, and San Gorgonio Pass through Sunday night, with temperatures forecast between 115 and 121 degrees as strengthening high pressure builds the hottest air of the year so far over the inland Southwest. That same ridge is what keep is also keeping California’s coastal markets elevated (without matching the desert’s extremes). Tomorrow’s board has the market running about 5 degrees cooler than the NWS forecast on Los Angeles’s low, and the forward scenario below has both a Los Angeles low and a San Francisco high among the board’s most interesting positions.

2) A look back, how yesterday’s daily temperature forecasts verified

Below is a scorecard for the forecasts as they stood at 5:00 PM ET on Monday, July 27th, for the target day of Tuesday, July 28th.

Each panel orders its cities by mean absolute error (MAE) for that metric, from largest to smallest.

The margins report that error, down the right for each city and along the bottom for each system.

The columns present some commonly referenced forecast systems compared to the ForecastEx market.

All four displayed commonly referenced forecast systems rely on physical numerical weather prediction models that solve the equations of the atmosphere forward in time. ECMWF (the European model) and GFS (the American model) are shown here in their raw form. Raw physical models are flexible and can handle weather situations they have never seen, but they carry systematic biases, and forecasts apply to a relatively large discrete grid box that encapsulates the weather station rather than the single point where a weather station sits.

LAMP is a model output statistics (MOS) system. It compares historical forecasts from physical models to what actually occurred and uses the errors to statistically correct systemic biases at the level of individual weather stations. The tradeoff is that a statistical fit is anchored to past situations and bends less readily to a genuinely unusual one. The National Weather Service’s National Digital Forecast Database (NDFD) is undergirded by the National Blend of Models MOS System but layers on human forecasters’ judgment, which is more flexible to dynamic weather situations but introduces subjective judgment (more on all of this here).

Minneapolis carried the largest error on highs at 4.9°F, and the American model was the day’s weakest system on that metric.

On lows, Phoenix produced an unusually large 6.9°F error, so the two grids rank quite differently.

Postmortem on the Minneapolis miss. The high reached only 88°F in Minneapolis, well under every system’s call and especially the American model’s 99.6°F, an 11.6°F overshoot.

At the lead, the National Weather Service office in the Twin Cities was describing a cold front that had just moved through, with temperatures still near 95 that afternoon but dew points already falling and wildfire smoke returning from the Pacific Northwest. By the target day, the office was writing that the region sat on the northeastern edge of a strong upper-level ridge, supporting highs closer to the low 90s with lower dew points keeping conditions less oppressive. The cooling behind that front, and the ridge that followed, ran further than the American model’s guidance had allowed for.

Over the seven-day window ending yesterday, the ForecastEx prediction market ties the Aviation Forecast for the lowest error among the twelve systems, both at 1.46°F.

This constitutes ongoing evidence that weather prediction markets may be the most accurate short-range weather forecasts available, and with the enhancement of prediction market skill becoming more apparent as lead time decreases.

This is because a market price is a confidence-weighted consensus of participants who can condition on every model shown here, plus whatever additional knowledge they bring to the table at live timescales. One way to think about it is that the prediction market is a constantly offered rewards program for anyone who can push prices and probabilities toward their true, most-calibrated values (and a penalty for those who try but fail).

On average, no public tool beats the market consistently, but the trade board below asks the narrower question of where a single tool could potentially be used to identify edge in particular circumstances.

The trade board

• Access to live probabilities and the ability to trade is available via IBKR Prediction Markets.
• General primers on how these prediction markets work are available in the IBKR prediction markets FAQ (scroll down) and the ForecastEx FAQ.

The best-paying contract on the board was No on the Atlanta low below 78 degrees, 8 cents at the Monday 5:00 PM ET price for an 11.8 times payout net of fees, and the American model was the only one of the four scorecard tools that implied it. The best-supported position was Yes on the Washington DC high above 88 degrees, which cost 12 cents and paid 8.0 times when the city settled at 90, with the American model itself calling 91 degrees.

The autopsy above shows the ladder behind the Minneapolis postmortem. The market itself was genuinely unsure at the settled strike, pricing the 88 degree line near even money, and every warmer strike from 89 up fell away sharply, leaving the American model’s 100 degree call stranded far to the right of where settlement actually landed.

3) A look ahead, tomorrow’s board

Contracts for Thursday listed at noon today, and the prices on the map are from 12:40 PM ET this afternoon. The widest gap anywhere on the board is the Philadelphia low, where the market sits 5.2°F above the NWS forecast.

• Access to live probabilities and the ability to trade is available via IBKR Prediction Markets.
• General primers on how these prediction markets work are available in the IBKR prediction markets FAQ (scroll down) and the ForecastEx FAQ.

Hypothetically, if one were to assume that the NWS Blend, which had the lowest error yesterday among the public tools, were to be exactly right on Thursday, the strike paying the most on the high side would be Yes on San Francisco above 71 degrees at 12 cents, an 8.0 times payout net of fees. On the low side it would be Yes on Los Angeles below 67 degrees at 12 cents, also an 8.0 times payout.

Temperature Market of the Day, Minneapolis

Tomorrow’s highlighted market is Minneapolis (KMSP), because it has the biggest forecast bust yesterday. The figure above runs from today’s noon listing through the end of tomorrow, showing observations so far, two independent hourly forecasts, and each model’s forecast high and low for the contract day, with the ForecastEx price since listing in the lower panel.

The models span 8.5 degrees on tomorrow’s high, from the French model at 96 down to the Japanese model at 87, without any single outlier driving the spread. The National Weather Service office serving Minneapolis explain the reason for the uncertainty in their forecast discussion. They highlight a weak shortwave rounding the ridge, still centered over the southern Rockies, that could allow a brief shower or thunderstorm that would inhibit the high but is still quite uncertain.

In price space, the market’s confidence falls off sharply between 88 and 90 degrees, right where the ICON model and the NWS Blend cluster their calls, even as the French and American models keep pushing for the mid-90s.

Note: uncertainty in day-ahead temperatures corresponds to significantly different energy use and expenditure across metro areas, which can inform operating and financial considerations for various entities involved in energy. Prediction markets on daily temperature allow for probability calibration across the distribution and can be used as financial instruments to hedge against the worst scenarios.

4) Tropics, a quiet basin with El Nino leaning on it

The Bahamas lead the major hurricane landfall board at 10.9%, equivalent to a Yes contract price of 11 cents, with Florida just behind.

The probability of a Category 4 or stronger United States landfall stands at 8.0% (essentially climatology, with no live storm contributing).

The season’s two named storms so far, Arthur and Bertha, both peaked as tropical storms in the Gulf, so nothing has yet reached hurricane strength.

The probability of a major hurricane making landfall within 50 miles of Miami-Dade, Florida currently stands at around 5%. Thus, a purchase of a “Yes” at 5 cents is an analog to parametric insurance, where the occurrence of a disaster based on a simple weather parameter would pay out, in this case 19X of the purchase price.

The Climate Prediction Center’s Global Tropical Hazards Outlook, a weekly product valid from August 5th to August 18th, expects El Nino conditions to keep suppressing Atlantic tropical cyclone activity, with formation unlikely over the Caribbean and the central Atlantic through the window.

Two named storms against a pace of 1.7 implied by the CSU seasonal forecast leaves the season running essentially on schedule, with the steep part of the formation calendar still ahead. On the hurricane count ladder, Kalshi prices the season above ForecastEx at every strike the two venues share, at least five hurricanes trading at 38 cents against 62, and the gap persists into the tail. On named storms, the venues flip in the middle of the ladder, with ForecastEx higher at 11 and 13 hurricanes before Kalshi takes over in the deeper tail.

Further Reading

• Hurricane Forecast Contracts
• Daily High Temperature Markets at ForecastEx
• Energy expenditure implications of next-day temperature forecasts
• Disaster Insurance Applications of Forecast Contracts
• Hurricane Forecast Contracts as a Diversifying Asset in an Investment Portfolio
• Bundling “NO” Hurricane Landfall Forecast Contracts to Reduce Downside Risk

Climate Contracts

• Global Crop Yield Forecast Contracts (view here).
• How Quickly Will The Globe Warm? Paris Agreement Forecast Contracts (view here).
• Will the Atlantic Overturning Circulation (AMOC) Collapse? Prediction Markets Can Quantify Sentiment (view here).

Appendix

ForecastEx does not publish a single forecast temperature. It lists a ladder of contracts at different strikes, and each price is the market’s probability that the day’s extreme passes that strike. Comparing the market with the other systems therefore requires converting that ladder into a central estimate. The 50 percent level is the market-implied median, so where two adjacent listed strikes bracket that crossing, the central value is interpolated between them. Where the ladder does not bracket 50 percent, because every listed strike sits far in or far out of the money, no central value is recorded and the market is left unscored rather than extrapolated beyond the quoted ladder. That reflects how the strikes happened to be listed rather than anything about forecast quality, so it should count neither for nor against the market.

About the author

Patrick T. Brown is the Head of Climate Analytics at Interactive Brokers, where his work focuses on the information discovery and risk-transfer applications of prediction markets in weather, climate, and natural disasters.

He holds a PhD in Earth and Climate Science from Duke University, a master’s degree in Meteorology and Climate Science from San Jose State University, and a bachelor’s degree in atmospheric and oceanic sciences from the University of Wisconsin, Madison. He is an adjunct faculty member (lecturer) in the Energy Policy and Climate Program at Johns Hopkins University and has conducted research at the Carnegie Institution at Stanford University, NASA JPL at Caltech, NASA Langley in Virginia, NASA Goddard in Washington, D.C., and NOAA’s GFDL at Princeton University. He has published scientific papers in Nature, PNAS, and Nature Climate Change, as well as many disciplinary journals, and his research and commentary have appeared in The New York Times, The Wall Street Journal, The Economist, CNBC, CNN, The BBC, The Washington Post, NPR, Newsweek, The Guardian, The Atlantic, Foreign Policy, and The Los Angeles Times, among other venues.