The Return Forecast

Forecasts as context, not promises

A forward return forecast is a model's estimate of where a stock's price might stand over some future horizon, given current data. The key word is estimate. Any quantitative forecast is built on historical patterns, and history is an imperfect guide: companies restructure, macro conditions shift, and markets reprice assets on information that no model has seen in advance.

ŷRobot's return forecast is honest about this. Wide error bars are not a weakness to be hidden — they are a truthful description of how much uncertainty surrounds any single stock's near-term direction. A forecast that appears confident but is built on fragile assumptions is more dangerous than one that openly carries a wide range of plausible outcomes. The practical implication is that the return forecast is most useful as one piece of a multi-signal picture, not as a standalone directive.

Reading the forecast as probabilistic context rather than a promise changes how you use it. A forecast that leans in a particular direction is worth paying attention to — especially when it aligns with other independent signals — but it is not a reason to act on that single number alone. The score's architecture is designed specifically to prevent any one number, including this one, from dominating the final read.

Its place in the score

The forward return forecast carries a 20% weight in the AI score — the second-largest of the eight components. Only the fair-value gap, at 30%, contributes more. That weighting reflects a deliberate judgment: a directional forward view adds meaningful information beyond what fundamentals and historical valuations already capture, but it should not dominate the composite on its own.

When the return forecast aligns with the fair-value gap, chart signal, and other components, that agreement is reflected in the score's confidence measure — cross-signal agreement is one of the two ingredients in it, alongside data completeness. When the forecast disagrees with the other signals — pointing one direction while the rest point another — the overall score reflects that tension rather than resolving it artificially. That is by design: cross-signal disagreement is itself information about how confident any read on this stock should be.

Because the return forecast is a forward-looking model output, its contribution to the AI score updates as underlying data refreshes. Stocks with sparser data histories may show lower confidence alongside the score, which captures how much the system was able to work with when producing both the forecast and the composite.

Where to see it

The return forecast is surfaced on each stock's page as a forecast chart. The chart plots the model's directional estimate over time alongside the stock's actual price history, giving you a visual read on how the forecast has tracked against realized price movement — and how much the forecast has varied.

The forecast chart is separate from the Intrinsic Value card, which shows the fair-value gap. They are answering related but different questions: the fair-value estimate asks whether the stock appears priced above or below what the business seems worth based on fundamentals; the return forecast asks what direction the model sees from here. The two charts address related but distinct questions about the same stock.