The heaviest single input to a stock’s conviction score (30% of it). It asks one question of the forecast itself: does the model lean upward, and does it lean far enough to be worth trading? A gentle drift up is not the same as a real move, so the component only pays full marks when both are true at once.
The model simulates a large number of possible price paths for the stock, and those paths are not all equally likely — each carries a weight. Two numbers come out of that: what share of the model’s path weight ends above today’s price (the upward lean), and how big the expected move is (the expected growth over the horizon).
The component scores 1 — full marks — only when at least 65% of that weight sits above today’s price AND the expected move is at least +10%. If the model merely leans upward (half or more of the weight above) but the expected move is small, it scores ½. Otherwise it scores 0. In plain terms: the model has to be both pointing up and pointing up hard.
This is the most important sentence on this page, so it is worth being blunt: the share of simulated paths ending above today’s price is not the chance you make money, and we will never present it as such. It treats a one-cent rise exactly the same as a doubling. It is silent on the size of the loss on the paths that went the other way. And it is the output of a model whose own confidence bands are only approximately calibrated.
A stock can have most of its paths ending slightly higher while the minority that fall, fall catastrophically — a coin-flip that pays you a dollar or costs you fifty. That is precisely why direction is only 30% of the conviction score and never the whole of it, and why a zero on the tail-risk component overrides it entirely.
Quantum direction is one of the two gate components in the conviction ledger. If it scores 0 — the model sees no upward edge worth naming — the whole conviction score is capped at 3.0 out of 10, no matter how bullish every technical indicator looks. A perfect chart cannot manufacture a thesis the forecast does not support. The other gate component is quantum tail risk.
Read it as the answer to "does the forecast itself actually back this trade?" — before you look at any chart pattern. If it is 0, nothing further on the page can rescue the setup and you should expect a WAIT or AVOID verdict. If it is 1, the model is pointing up with conviction, and your next question should be the one the tail-risk component answers: how bad is the downside if it is wrong?
No. It means the model’s simulated paths lean upward and the expected move over the horizon is at least +10%. That is a statement about the forecast, not a promise about the outcome — the model can be wrong, and a full score is not a guarantee of any kind.
Because this component reads the forecast, not the chart. The trend, MACD, RSI and 52-week-position components read the chart, and they are deliberately outvoted: if quantum direction is 0, the conviction score is capped at 3.0 however good the chart looks.
Deliberate coarseness. The underlying forecast is not precise enough to justify a smooth score, and a three-step cut-off is honest about that. A finer number would imply a precision the model does not have.
Browse all S&P 500 tickers to see this metric applied to individual companies.
Educational research only — not investment advice.