How to check this later
Each row prints the first twelve characters of its rule fingerprint — a sha256 of the rule in canonical form, with members sorted and weights rescaled, so the same fund described two ways produces the same hash and a fund quietly edited produces a different one. The full hash is on each fund's own page, alongside the rule it was taken from.
Every snapshot taken afterwards carries the hash it was measured under. If a rule here were ever changed behind its own track record, the mismatch would be visible rather than merely forbidden.
What this cannot tell you
A Foundry rule is a fixed set of weights and a rebalance cadence. It has no conditional logic, so strategies that trade on a signal — dual momentum, trend-following overlays — cannot be represented here at all, and are left out rather than approximated with fixed weights their authors never published.
Where the instrument catalogue forced a substitution, the row says so in its own words. Two of them matter: there is no small-cap value fund available, which removes the tilt that half of the Golden Butterfly's thesis rests on, and which takes away the premium the Larry Portfolio is built entirely around. Read those two as approximations, not as the strategies their authors described.
Simulated results are not returns. They assume the rule was followed exactly, that rebalances happened on schedule, and that no tax or commission was ever paid. The leaderboard shows every published fund, including the ones that fail. Educational analysis only, and not investment advice.
