Things That Should Exist: A Weather Prediction Market

What's up guys - long time no blog.

I got a little inspired recently reading about MegaETH and their plans for scaling ETH to 100k TPS.

Thinking a little about blockchain applications that aren't yet built, but that would be feasible with near-free blockspace, I was reminded of a (difficult) discussion I had with my dad one time in the car.

I was trying (unsuccessfully) to explain to him why blockchains might be a good thing for the world, using an example I almost certainly stole from The Truth Machine.

Lots of things depend on the weather. For example, in Saskatchewan, wheat yields fell from 3.5 tons per hectare in 2021 to 2.4 tons per hectare in 2022, largely due to reduced precipitation. Downstream, this meant that 13m tons less wheat was available for sale, contributing to the massive run up in wheat prices.

Farmers obviously take out insurance to protect themselves from these variations, supporting an industry worth tens of billions a year.

I don't particularly know why, but insurance salesmen (along with lawyers, bankers and most suit-wearers) annoy the fuck out of me. Because they are nothing more than middle men - their salary, their fancy (hire-purchase) car and their existence is predicated on them over-charging farmers (people who actually produce stuff) for what is basically a bet.

Farmers in Saskatchewan bet that there will not be much rain this year, so that their bet winnings cover the loss they incur if the weather is bad that year. But of course, they massively overpay relative to fair odds for this bet - enough that the insurance company makes a profit, while paying a bunch of do-nothing office staff to sit in their cubicles all day. Also, insurance is a highly regulated industry, so both parties are likely paying a hefty tax to the government. All because they want to essentially take a bet on the weather!

I've focussed on farmers in Saskatchewan because my grandfather (a farmer) was born there and because claude highlighted it as an important area for wheat. But obviously I could've used any of dozens of products and hundreds of localities. These markets are incredibly large and liquid, and absolutely ripe to be dis-intermediated by blockchain technology.

I was quite surprised to see that, (despite this article from Dec 2021 announcing a partnership between chainlink and AccuWeather), there are currently no weather oracles on any EVM-compatible chain.

Obviously in terms of prediction markets, Polymarket has attracted a lot of attention recently for being the most liquid place to bet on the Presidential Election.

I think that for the crypto space (and for the world at large), a highly liquid, fully on-chain prediction market for weather would be a massive step forward.

WeatherXM is a project I found who are building a helium-style de-pin network of weather stations, using their native token as an incentive mechanism for people to buy and maintain the infrastructure. They recently announced a funding round where they mention "development of on-chain weather derivatives". So it seems at least 1 project is looking seriously at this opportunity.

At times, I think we can all get a bit cynical about the largely meaningless money games (not to mention the hacks and phishing scams) that attract the most attention in this space. Every now and then I think it's helpful to remind ourselves that there are real-world use cases for this technology, and that there is unironically the potential for crypto to make the world a better place.

GL out there!

D