A Score Is Not a Prediction

The most common misunderstanding about any market score is also the most expensive one: reading a description of the present as a forecast of the future. The two look similar on screen. They are fundamentally different objects.

Descriptive versus predictive

A descriptive reading answers the question: what does the available data say right now? A predictive reading claims to answer: what happens next? The Bias Score is strictly the first kind. When a stock reads +70, the correct sentence is "public market data currently leans strongly positive on this stock". The incorrect sentence is "this stock is going up".

The distinction matters because markets price in known information continuously. Data that leans positive is often data the market has already acted on. A strong reading describes a strong data state, and a strong data state can precede continuation, consolidation, or reversal. History contains all three in quantity.

Why we do not publish hit rates

A natural follow-up question is "how often is the score right?". For a forecast, that question makes sense. For a description, it quietly changes the subject, because it grades the score against a claim it never made. Publishing an accuracy percentage would imply the score is a prediction engine, which would misrepresent what it is and invite exactly the wrong usage. We decline the frame rather than optimize for it.

What sudden score moves mean

Scores can jump substantially in a day. A jump means the underlying data changed fast: a ratings cluster shifted, news flow turned, price structure broke a pattern, fundamentals were repriced by a report. It does not mean the stock must now move in the score's new direction. Fast-moving data is frequently contested data, and sharp score changes often come with elevated disagreement between data dimensions, which is exactly what the divergence reading is for.

Using a descriptive score well

Used as designed, a bias score does three jobs:

• Triage. Across hundreds of stocks, it shows where data is aligned, where it is conflicted, and where it changed since yesterday, in seconds. The sector views make the comparison structural.
• Context. For a stock already on a watchlist, the score and its 30-day path show whether the data state is improving, deteriorating, or churning.
• Discipline. A fixed scale, computed the same way every day, is a defense against narrative drift, the tendency to remember data as having agreed with whatever one already believed.

What it cannot do is decide anything on anyone's behalf. It compresses the market's data. The judgment, the sizing, the timing, and the responsibility remain with the reader, ideally alongside professional advice where decisions have real consequences.

EquityBias is a market data research tool. Bias Scores represent aggregated public market data and do not constitute financial advice, investment recommendations, or price predictions. Past readings are not indicative of future market performance.