Apocryphal, perhaps, but illustrating the point

We admit to wiping away a little tear at hearing that this story is regarded as apocryphal:

The dubious story of a student who won $100,000 betting on the weather

Urban legend has it AI-based system scanned aviation data and weather-betting sites, then identified mismatches to get good odds

The sadness comes from the fact that if the story were true it would be an excellent illustration of the benefit of such prediction markets:

Online platforms such as Polymarket allow gamblers to bet against one another on practically anything, including weather wagers such as the “highest temperature in London tomorrow”. According to an online urban legend, a Chinese student successfully gamed this system to make a pile of money.

The student supposedly used Metar, meteorological aerodrome reports. These are used in the aviation industry and are updated hourly, unlike most other publicly available weather forecasts which only update every few hours. The student is said to have used an AI-based system running on two PCs in his dorm room to download Metar data, scan weather-betting sites and identify mismatches between the two where he could get good odds. He made a series of successful bets and soon amassed more than $100,000 (£73,500).

But the story is dubious because betting on weather is such a niche area, and it is unlikely the student would find enough takers. The lack of identifying details is also the hallmark of urban legend. The story seems more likely to be an interesting idea of what someone could do in theory, rather than something someone has actually done.

The efficient markets hypothesis simply says that markets are efficient at processing the information about what prices should be in a market. It isn’t - as some affect to profess - that markets are the efficient way to do everything. The reason the EMH stands is because people can see ways to make money out of information - the finer the better - and thus information held gets coded into prices. If hourly weather reports are where it’s possible to make money as against three hourly then prices reflect information about immediately future weather better than without the inclusion of the hourly information. This is good, markets are now more efficient at processing that information.

So, even if this is not being done, or has not been, it should be. We’re also not sure that the explanation for why it hsan’t stands. Even in thin markets an edge will produce profit. Over time a decent edge could quite well produce 6 figure profits.

This is not, of course, investment nor even speculation advice.

Tim Worstall