The Criterica Score is a 0 to 100 score with a defined target, horizon, reference population and mapping from model output to score, returned as not rated where the evidence is insufficient. It is not a merits opinion and not a return forecast.
A score is only useful if a reader can say exactly what it measures. The specification fixes the target, the horizon, the reference population and the rule that maps a model's output onto the scale, and it is versioned so a score always names the version that produced it.
Validity is established per asset class. A claim for an asset class waits until a model trained on the relevant realized outcomes has passed a preregistered backtest, and the claim is published with that evidence. Where support is missing, the result is not rated.
The score is a measurement of a population-level relationship, applied to one file. It is not advice on the merits of a matter, a statement about how counsel will perform or a forecast of what a position will return.
Working a file that turns on this term?
Working through a diligence process?
Institutional partners evaluating a position against this platform's outcome and duration models are welcome to reach out directly.