A transparent research score
A priority score you can inspect—not a promise you must trust.
AheadPick scores where research time may be most useful. The model does not forecast a return. It combines observable inputs, applies quality gates and keeps uncertainty visible.The scoring model
The base score combines four positive factors and then subtracts an uncertainty penalty. Inputs are normalized before weighting so one large raw number cannot dominate the queue by itself.
- Event surprise — 30%: the size, speed and persistence of the probability move.
- Exposure strength — 25%: how directly the event can affect the mapped equity or ETF.
- Reaction gap — 25%: how much of the event move appears not to be reflected in the asset response.
- Market quality — 20%: liquidity, spread, activity and clarity of the source market.
- Uncertainty penalty: alternative explanations, weak mappings and incomplete price history reduce the score.
Five gates before a signal can rank
A high arithmetic score is not enough. A candidate must first pass freshness, liquidity, rule clarity, asset-mapping and price-history checks. Failed candidates stay outside the ranked queue.
- Fresh enough to represent a current repricing.
- Liquid enough that a single trade is less likely to distort the move.
- Clear settlement terms and an identifiable source market.
- A defensible path from the event to the exposed equity.
- Enough price history to compare event movement with asset response.
How to read the result
A higher score means the event–asset pair deserves earlier review. It does not mean the asset must move, that the mapping is causal, or that a trade will be profitable.
The full research view keeps the factor breakdown, source context and timestamp together so the original reasoning can be challenged later.