The drawer
A piece of paper sits inside a locked drawer with a single word written on it in pen. Nobody on earth knows the word. A prediction market on “what word is in the drawer” trades at 50/50 for every possible answer. The market is permanently and completely wrong.
Someone could pay a hundred dollars to break into the house, pick the lock on the drawer, read the paper, and trade on the information. If the market holds enough volume to make the hundred-dollar break-in economics profitable, someone does this and the price collapses to 99/1 on the correct word. If the market is too small, nobody breaks in. The cost of discovering the truth exceeds the profit available from trading on it. The market stays at 50/50 while the correct answer sits twelve inches away inside a piece of particle board furniture.
This is the information cost ceiling. A prediction market surfaces only information that costs less to acquire than the trading profit it generates. If the truth costs more than the gain, the market sits at an uninformative equilibrium constructed entirely from gossip and the pocket of public analysis that has not yet been economically invalidated.
Polymarket does not know the future
Every Polymarket contract displaying a probability percentage is a robust aggregation of publicly available information. It communicates nothing about the true probability of the underlying event. A contract showing 22 percent odds of a specific political outcome is the clearing price of the most recent informed trade. That price was set by information cheap enough for someone to acquire and act upon.
The true probability could be below 1 percent. It could be above 99 percent. The market cannot reach either number because reaching either number requires spending more to discover the truth than the market’s volume financially justifies. A market with ten thousand dollars of total volume cannot economically support spending one hundred thousand dollars on intelligence collection. It can support spending up to ten thousand. If the truth costs ten thousand and one dollars, the market remains permanently wrong and nobody will correct it because correcting it is unprofitable.
Volume is a ceiling on accuracy. Not a theoretical ceiling. A mathematical one.
The animal
There is an animal standing in a field. No one can determine whether it is a horse or a donkey without approaching closely enough to inspect its reproductive anatomy. A prediction market on the species trades at 50/50.
At ten dollars of market volume, nobody walks to the field. The market stays permanently at 50/50. At ten million dollars, someone charters a helicopter, flies to the field, performs the relevant inspection, and trades on the result. The market collapses to 99/1. The animal did not change species between the two scenarios. The only variable that moved was the financial incentive to acquire private information about the animal’s biology.
Apply this economics to any domain. Launching a satellite to monitor oil tanker levels becomes rational when the oil futures market justifies the launch cost. Hacking a private server for confidential data operates on identical arithmetic. Physically coercing an information holder operates on identical arithmetic. The market does not distinguish between legal and illegal information acquisition. It distinguishes between profitable and unprofitable acquisition. If the expected profit from a trade exceeds the expected cost of committing a crime multiplied by the probability of evading consequences, the crime becomes economically rational for at least one participant.
What the market actually is
Prediction markets are not truth-finding machines. They are profit-finding machines. They aggregate information up to the point where additional information costs more to acquire than it returns in trading gains. Below that point, the market reflects reality. Above that point, the market reflects the cheapest available narrative about reality. These are not the same thing.
The horse knows what it is. The market does not know what the horse is. The market will not know until someone pays to walk to the field and check. Nobody is paying because the market volume does not justify the walk. The horse is patient. The horse does not know it is the subject of a financial contract. The horse is standing in a field being a horse. The market is trading on the gap between what the horse knows and what the market can afford to learn. The gap is permanent.