Does the score predict anything?
Raw return after an event is not evidence. If the whole market rose, every event looks like it worked. Every outcome below is measured as excess return against Bitcoin over the identical window, and reported whichever way it comes out.
The short version
After Tickerz opens an event, crypto assets underperform BTC by about 6.0% over the next 7d.
4 of 20 events beat BTC. A coin flip would give about 10. The odds of seeing a split this lopsided by chance are about 85 to 1. In plain terms: a spike in attention has been followed by the asset lagging its market, not leading it. That is the opposite of what a list of unusual tickers looks like it is telling you. One month of data is not every market, and this is a measurement of what happened, not advice about what to do.
Excess return against each market, by horizon
| Horizon | Class | Measured against | Events | Mean excess | Median excess | Beat benchmark | Significance |
|---|---|---|---|---|---|---|---|
| 24h | crypto | BTC | 50 | +0.74% | +0.73% | 29 of 50 | p = 0.32, not distinguishable from chance |
| 24h | equity | SPY | 6 | +1.36% | +0.94% | 4 of 6 | p = 0.69, not distinguishable from chance |
| 7d | crypto | BTC | 20 | -4.83% | -5.97% | 4 of 20 | p = 0.012 |
| 7d | equity | SPY | 1 | +9.90% | +9.90% | 1 of 1 | n=1, too few to test |
| 7d | crypto, ordinary day | BTC | 1131 | -- | +2.02% | 62% | the comparison |
| 7d | equity, ordinary day | SPY | 374 | -- | +0.24% | 51% | the comparison |
How to read this
Excess return is the event asset's return minus Bitcoin's return over the same window. Significance is an exact two sided binomial test against a coin flip on how many events beat the benchmark. It tests direction only, not size. Equities are measured against the same crypto benchmark for now, which is the wrong benchmark, so treat those rows as unreported rather than as a result. Nothing here is a forecast or advice, and none of it is annualised.
Updated 2026-09-22 23:15 UTC. Every event behind this table is in the record, and the method is on the methodology page. The same test on four years of market data, 3,213 spikes found with the live formula, is in Attention fades.