Were we right?
Most people who talk about stocks quietly forget their bad calls. We don't. Every forecast we make is saved here with its date — and when that date arrives, we put the real price next to what we said, right or wrong, nothing deleted. So you can judge for yourself how good our forecasts actually are.
Why some rows carry a (T+20) or (T+60) tag
From 27 July 2026 every new forecast is checked at 1 month and 3 months from the day it was published — real calendar dates. If that date is a weekend or a public holiday on that exchange, we check on the next day it actually trades. Nothing else moves it.
Before that date we counted trading sessions instead: “T+20” meant the 20th day the market was open, “T+60” the 60th. That is roughly a month and roughly three months — but only roughly, and it drifted with every public holiday, which made the check date hard to predict and hard to explain.
We did not go back and change the old forecasts. Every cohort published before 27 July 2026 keeps the horizon it was issued on, is graded on that horizon, and still counts in the score exactly as it always did — only how it's displayed changed: every row below now leads with the calendar name, 1 month or 3 months, so the table reads consistently. A row struck under the old session count carries its exact original count alongside, e.g. 1 month (T+20), so the precise published fact is never lost. A row struck directly on the calendar convention has no fixed session count to show, so it carries no parenthetical. Both are live, and the difference is small — a calendar month is about 21 trading sessions and a calendar quarter about 63, against the old 20 and 60.
The engine was re-calibrated to match: the five-year test that validates each cone now measures the same calendar window we publish, instead of a fixed 60 sessions. Full methodology →
How we check: each forecast gives a likely range. When its date comes, we see whether the real price landed inside that range, and keep a running tally. If our "9 times out of 10" ranges are honest, the price should land inside them about 9 times out of 10 — more often means we're being too cautious, less often means we're too confident.
Get told when it matters
One short email when we publish a new study, or when a forecast resolves — right or wrong. No spam, and we never sell your address.
Subscribe by email →Free, via Substack — unsubscribe any time.