CI90 — the 90% confidence interval — is the price range between the 5th and 95th percentile of every path the quantum model simulates for a ticker at a given horizon. Read this before you read a band width as a difficulty score.
If the model is well-calibrated, about 90% of realized prices should land inside the published band over many forecasts — the other roughly 10% land outside it, by design. A wide band means the model sees genuine uncertainty for that name; a narrow band means conviction. Neither is inherently good or bad: a band that is never missed is usually too wide to be useful, and one that is missed often is overconfident.
Quantustik shows this band on every ticker page as the lower/upper bound around the expected-growth mean, at every forecast horizon.
The CI width shown on the ticker page is simply how far apart the two edges of the band are — the upper percentile (q95) minus the lower (q05), often expressed as a percentage of price. Wider means the model sees more uncertainty (earnings week, a structural break, a longer horizon); narrower means more conviction. Example (illustrative): a $95–$115 band around a $100 price is a CI width of $20, or 20% — twice as wide, and twice as uncertain, as a $95–$105 band. Width measures confidence, not accuracy: what matters is whether that width's actual coverage matches its stated 90%.
Live example: as of the last scan, AAPL's 3-month CI90 band is $281–$375 around a current price of $317 (above the current price by 1.0% at the mean). See the full AAPL forecast.
Rather than asserting calibration, we publish it: the live calibration page reports actual CI90 coverage per forecast start date against the committed TOP-20 backtest, misses included. Measured against the model as served across nine quarterly start dates from 31 Mar 2024 to 31 Mar 2026, coverage ran 78–95% of realized prices at 3 months, 71–92% at 6 months and 86–96% at 1 year — no single average is published, because when the forecast started moves the answer far more than most readers expect, and individual windows are worse still.
It means the model expects about 90% of realized prices to land inside the published band over many forecasts — not that the forecast is "right" 90% of the time, and it says nothing about direction.
Not by itself. A wide band reflects genuine uncertainty; a narrow band reflects model conviction. What matters is whether the band's actual coverage matches its stated 90%.
Against a committed, reproducible TOP-20 S&P 500 backtest, published on the live calibration page with the misses shown alongside the hits — not just an aggregate average.
No. CI90 is a coverage promise about the price range, separate from whether the model's mean forecast points up or down.
The CI width is the distance between the band's two edges (q95 minus q05), often shown as a percentage of price. A wider band means more uncertainty, a narrower one more conviction — but width is about confidence, not accuracy, so it's only useful when its coverage actually matches the stated 90%.
AAPL analysis shows this metric in context, or browse all S&P 500 tickers.
Educational research only — not investment advice.