The Forecast Board: four rules, locked in first, graded in public
On 3 November 2026 we launched the Forecast Board. It tracks four rules that were written down, fingerprinted and timestamped before any of their results existed, and it grades them in public.
Why a board
A back-test looks at the past, and the person running it can try many versions until one looks good. A forward test avoids that. The rule is fixed first, then time passes, then the result is scored. The board holds only forward results. Back-tested numbers stay on the evidence page, kept apart because a back-test is not a forward result.
The four rules
- 12-1 momentum. Each month it picks the four of 75 large US stocks that rose most over the past year, skipping the latest month, and holds them equally weighted. It went live on 2 October 2026.
- Earnings-growth momentum. Each month it picks the fifth of 171 large US stocks whose profit grew most over the past year relative to their assets, using only filings already public. Scoring starts on 1 December 2026.
- Low accruals. Each month it picks the fifth of 171 large US stocks whose reported profit is best backed by cash flow, using only filings already public. Scoring also starts on 1 December 2026.
- 1-year range forecast. Once a month it writes a 50% and an 80% price range for the next year for 75 stocks and the S&P 500, then checks how often the price landed inside. The first batch has 76 calls, made on 3 November 2026.
These are fixed, automated rules, not recommendations. The board describes how they perform against a benchmark and is not a call to buy, sell or hold anything.
How a rule is locked in
Each rule has a written document, dated before its first result. The page shows the SHA-256 fingerprint of that document, so anyone can hash the file and compare. Each rule also has a timestamp proof file, which can still be waiting on a Bitcoin block when the page loads.
Each rule has its pass line written in advance too. For the three stock-picking rules, there is one read, on or after the read date. The rule has to beat the equal-weight benchmark after costs and clear its p-value threshold (below 0.05 for 12-1 momentum, below 0.025 for the two new stock rules). It also has to beat a control: a dumb control for 12-1 momentum, and an eligible-universe twin for the other two. Fewer than 240 scored sessions means no verdict.
The range forecast is judged per monthly batch, once every call in the batch is a year old. The 80% range has to hold 70% to 90% of the time, the 50% range 40% to 60%, and the up-probability has to beat a 50/50 guess (a Brier score below 0.25). Anything else is published as a miss.
What you can see today
Nothing is graded yet. The board says so instead of showing early figures. As of the 30 October price data, 12-1 momentum had 21 scored sessions, and no figures are shown for it until 63. Its pass-or-fail read comes on 1 October 2027. The two new stock rules read on 1 December 2027, and the first range-forecast batch resolves on 3 November 2027.
Scores update monthly, against the S&P 500 and an equal-weight basket. A rule that fails stays on the page.
Follow it
The rules, fingerprints and progress bars are on the Forecast Board. Our general method is in the testing standard.
About Tickfloor
What is Tickfloor?
Tickfloor is a stock and crypto indicator for beginner and self-directed traders: it scores any stock or coin out of 10 from ten common chart checks, and publishes every backtest behind the method, the failed ones too. So far that is 1,342 strategies tested, and 0 passed the luck check.
Is Tickfloor free?
Yes, to start. You get 3 free scores a day with a free account. No card. The example read, the evidence and results pages, the glossary and the first two course lessons need no account at all. Paid passes unlock the rest and are listed on the pricing page.
How is Tickfloor different from other indicator sites?
Tickfloor shows its own test record next to the score, failures included: 1,342 strategies in 77 families against a benchmark that pays no costs. 265 finished ahead on the raw number and 0 of 1,774 tests were left once luck was allowed for. The score describes what a chart shows today. It is not a tip or a prediction.
General information, not personal advice. Tickfloor does not hold a financial services licence.