The statistics most tools skip.
Every output on this page is backed by a documented method — click through to the full methodology for the math.
Cointegration screening
An escalating test hierarchy — Johansen, Gregory–Hansen, Phillips–Ouliaris, and a Hurst-exponent proxy — so one method's blind spot doesn't decide the result.
Z-score & half-life
ADF stationarity testing plus an Ornstein–Uhlenbeck half-life estimate — how fast statistical analysis shows deviations revert, not just whether they might.
Regime & breakdown detection
PELT change-point detection partitions the history into regimes, so a long, stationary current regime can be told apart from a heavily fragmented one.
Earnings-aware gating
A pair is excluded when either leg has an earnings report imminent, since event risk swamps the mean-reversion statistics. Instruments without earnings pass automatically.
Read the methodology →Multi-asset (equities / FX)
The same price-ratio pipeline runs across equities and FX — one consistent statistical treatment regardless of asset class.
Transparent backtests
Backtests net out commission, bid–ask spread, and market impact — no frictionless-fantasy performance numbers.
What RINS actually screens.
The universe, the inputs, and the cadence — stated plainly, so you know exactly what the statistics are computed on.
+ 21 sector & index ETFs
& EM · Professional tier and above
candidate combinations
platform minimum 0.55
30 / 60 / 90-day MA windows
ready before the open
Coverage is what the engine computes on — not a claim about returns. See Disclaimers & Risk Disclosures.
See it on your own watchlist.
Research tool, not advice.