We built models to beat markets. They don't.
Pigglet is the honest record of that — two markets, tested the same unforgiving way, both saying the same thing. It leads with the null result instead of hiding it, because a tool that shows a ranking while concealing that the ranking predicts nothing is a slot machine with a chart. Pick a world.
The ranking funnel
A four-factor model over ~2,000 US names, rebalanced monthly and backtested on survivorship-free, point-in-time data. It has no measurable ability to predict returns — so it narrows the market to a few names worth reading about, and never says buy.
Enter Stocks →The prediction markets
The same question where the evidence is stronger: markets that resolve in days, so months buy thousands of scored predictions. The arena is the honest result so far; an odds and prop-research tool is being built on top of it.
Enter Sports →