Full Kelly assumes your win-probability and reward:risk estimates are exactly correct, which real-world estimation error makes unlikely — half or quarter Kelly is standard institutional practice. This is a sizing framework, not investment advice.