Best RSI screener settings, backtested
RSI screens are among the most searched setups, and the most argued about. Here is how the common RSI(2) and RSI(14) variants actually performed on the same 5-year test.
Test: 518 S&P 500 + Nasdaq-100 stocks, 2021-10-13 to 2026-10-12, each signal held 5 trading days. Hit rate = share of signals that closed higher 5 days later.
Ranked by 5-day hit rate
| # | Screen | Hit rate | Avg 5d | SPY 5d | Max DD | Signals |
|---|---|---|---|---|---|---|
| 1 | RSI(14) Oversold Mean Reversion | 57.9% | +0.88% | +0.85% | -29.9% | 22,196 |
| 2 | RSI(2) Oversold Bounce Mean Reversion | 53.9% | +0.41% | +0.32% | -22.8% | 34,000 |
| 3 | RSI(14) Overbought Momentum | 51.8% | +0.22% | +0.07% | -37.6% | 44,280 |
What to know
Short-lookback RSI(2) reacts within days, so it fits quick mean-reversion swings. RSI(14) changes slowly and fires less often. An oversold reading is a condition, not a timing signal: most of these screens also require a longer-term uptrend filter.
Standard RSI(14) under 30.
Larry Connors setup: RSI(2) below 10 while above the 200 SMA.
RSI(14) over 70 — tests whether strength persists.
Copy-paste rules for every screen above, plus Daily Fire Picks: 1-3 highest-confidence setups emailed before the open. Educational, not advice.
FAQ
What are the best RSI screener settings?
In our 2021–2026 backtest, RSI(14) Oversold had the highest 5-day hit rate of this group at 57.9% over 22,196 signals. "Best" depends on your holding period and risk tolerance; compare drawdowns too.
Where are the exact settings?
The exact backtested thresholds and copy-paste TradingView, Finviz and ThinkorSwim versions are part of SetupHits Pro ($39/mo), delivered by email. The rankings and stats on this page are free.
Are these results realistic?
They are simulated on daily bars with no fees or slippage, on today's index members (survivorship bias). Treat them as a relative comparison, not a forecast.
More guides
Simulated results; no fees, slippage or taxes; survivorship-biased universe. Not financial advice. Methodology.