We tested 49 causal chains. One worked.
Most causal claims in finance don't survive contact with data.
Yes No Capital was built to test relationships between prediction market events. Academic literature offers dozens of plausible mechanisms — inflation depresses incumbent vote share, oil spikes precede recessions, tariffs raise consumer prices, war depresses markets, rate cuts move equities, credit spreads forecast corporate distress. Forty-nine such chains, each sourced to a peer-reviewed paper.
Each chain is tested the same way: when the cause resolves, did the effect follow more often than it would by chance? A chain earns a PREDICTIVE verdict only when it holds up across the full history, not just in hindsight. Everything else is NO EDGE, INVERTED, or INSUFFICIENT.
After more than ten thousand resolved pairs, one chain survives: nuclear policy events correlate with nuclear-related markets above baseline, stable across time-split validation. That is weak. It is also the only one that passes all the gates.
The majority of chains show no edge despite 500+ pairs each. Inflation → elections. GDP → elections. Oil → inflation. Tariffs → GDP. At scale, the "obvious" macro relationships don't discriminate prediction market outcomes. Keyword pairing is not causation.
This is why backtests are easy to publish and hard to trust. Sample-cap an hour earlier, filter by volume, restrict to a favorable window, and any chain can look predictive. The honest version of a scoreboard is rarely what publication incentivizes.
Why most of them fail
The guardrails are built to catch the usual ways a backtest flatters itself: an edge that existed once but has since decayed, a result that only appears when you cherry-pick the window, and a relationship that looks strong on average but can't actually tell a YES outcome from a NO one. Most of the forty-nine trip at least one of these. That is the point: the honest test is the one most relationships don't pass.
The scoreboard at yesnocapital.com/track-record is updated hourly as new events resolve. Chains auto-deprecate when the evidence is in — the library is self-cleaning.
One surviving chain is not a product. It is a start. The point isn't any single chain; it is that most things don't work, and making that visible is the only honest version of this kind of tool.