Mean reversion rsi

RSI (Relative Strength Index)

Bullish: RSI crosses above oversold level (default 30) from below. Bearish: RSI crosses below overbought level (default 70) from above. Uses Wilder's smoothed moving average.

Signal family

Mean reversion — Oscillator-based signals that fire at overbought or oversold extremes — typically fade the prevailing move.

Parameters

Name Description Default Range
period RSI period 14 5–50
overbought Overbought level 70 60–90
oversold Oversold level 30 10–40

Historical context

185,987 valid triggers on 3,361 distinct tickers between 2015-03-03 and 2026-04-22. Universe: us_only · mcap ≥ $100,000,000 · price ≥ $1 (3,366 tickers). Entry at open T+1. 1d = intraday T+1; 20d = open T+1 to close T+20.

Benchmarks: spxew (S&P 500 Equal Weight — the primary benchmark here; a median-stock view that avoids the 2020+ megacap-concentration distortion), spx (S&P 500, cap-weighted), and msci (MSCI World USD). Per-stock regime: trending = ADX(14) ≥ 25, high vol = 20d ann. vol ≥ 20%.

At a glance (20d alpha vs S&P 500 Equal Weight, US-only)

Bullish
-0.04%
vs random-date null: worse than random (pperm=1.000)
Bearish (negative alpha = signal right)
-0.03%
vs random-date null: beats random (pperm=0.005)

Reading this: the random-date null is: for each ticker, sample N random dates and compute the same alpha — what alpha does a signal with no information produce? If the signal's observed alpha beats the null (pperm≤0.05), it's adding real information. If it's inside or worse than the null, the signal doesn't add value over random firing — any observed alpha is either noise or a universe artifact.

How often does RSI fire in each regime?

The signal's bucket distribution is itself informative. If 50%+ of all RSI triggers fire in the "non-trending + high vol" quadrant, the signal is structurally a chop-market event — regardless of what its textbook definition claims. Bullish and bearish are shown separately; counts are across the full US-only sample after the mcap and price floor.

RSI (Relative Strength Index) (rsi) — trigger count distribution by per-stock regime quadrant (trending/non-trending × high/low realized volatility) for , US-only universe

Per-stock regime quadrant — 20d alpha

Each trigger is tagged with the host stock's own technical regime on the trigger date: is the stock itself in a trend (ADX(14) ≥ 25) or ranging? And is its realized 20-day volatility high (≥ 20% annualized) or low? This is the textbook conditioning variable — "does this signal work better in trending stocks?" — answered at the level of the individual stock, not the market. Positive bars are good for the signal; negative bars mean alpha vanishes into the benchmark or worse.

RSI (Relative Strength Index) (rsi) — mean 20-day alpha versus S&P 500 Equal Weight by per-stock regime quadrant, side by side
Trending + Low vol
Stock in a clean directional move with low realized volatility. Textbook "trend-following paradise" — smooth grind with little whipsaw risk.
Trending + High vol
Violent directional moves — parabolic rallies, crisis selloffs. Trend exists but the path is noisy. Signal timing may be imprecise.
Non-trending + Low vol
Quiet chop, summer doldrums, consolidations. No directional bias but also no big swings — small edges become reliable if they exist at all.
Non-trending + High vol
Choppy and violent — the classical "whipsaw zone" for momentum signals. Crossovers and breakouts fire repeatedly without follow-through.

Sub-period check — does the signal work in every era?

A multi-year average can hide major instability. We split the sample into three non-overlapping windows: 2015–2019 (pre-COVID, normalized monetary policy), 2020–2022 (pandemic crash + recovery + rate-shock bear), and 2023+ (post-ZIRP, AI megacap rally). If a signal's alpha is positive overall but comes entirely from one era, that's a red flag — the conditions that produced it may not repeat. A robust signal shows a consistent sign across all non-empty buckets.

RSI (Relative Strength Index) (rsi) — 20-day alpha split by historical sub-period (2015-2019, 2020-2022, 2023+) to check consistency across market regimes

↑ Bullish triggers

Bench Metric 1d 5d 20d 60d 252d
spx Stock % -0.11% +0.01% +1.09% +4.47% +15.87%
Bench % -0.12% +0.10% +1.45% +4.42% +16.24%
Alpha % -0.01% -0.10% -0.35% +0.07% -0.38%
Median alpha -0.05% -0.17% -0.65% -1.61% -9.56%
Hit rate (α>0) 48.8% 48.3% 46.9% 45.8% 39.3%
p (naive) 0.3858 <0.001 <0.001 0.4443 0.0814
p (HAC) 0.3859 <0.001 <0.001 0.6278 0.5008
N 77,415 77,331 76,560 74,444 69,790
msci Stock % -0.11% +0.01% +1.09% +4.47% +15.87%
Bench % -0.09% +0.12% +1.27% +4.08% +14.36%
Alpha % -0.04% -0.16% -0.23% +0.40% +1.42%
Median alpha -0.12% -0.28% -0.56% -1.30% -7.60%
Hit rate (α>0) 47.4% 47.4% 47.4% 46.6% 41.3%
p (naive) 0.0003 <0.001 <0.001 <0.001 <0.001
p (HAC) 0.0004 <0.001 0.0001 0.0028 0.0117
N 77,137 76,289 75,769 74,007 69,417
spxew Stock % -0.11% +0.01% +1.09% +4.47% +15.87%
Bench % -0.18% -0.06% +1.17% +4.03% +13.02%
Alpha % +0.04% +0.03% -0.04% +0.48% +3.01%
Median alpha -0.05% -0.09% -0.37% -1.15% -6.16%
Hit rate (α>0) 48.8% 49.1% 48.2% 46.8% 42.8%
p (naive) 0.0009 0.1948 0.4479 <0.001 <0.001
p (HAC) 0.0010 0.2260 0.5523 0.0003 <0.001
N 76,953 76,638 75,697 73,674 68,915
Distribution of all 20d alpha outcomes for this direction. Median and winsorized mean shown.
RSI (Relative Strength Index) (rsi) — bullish 20-day alpha histogram showing distribution of per-trigger returns
Observed 20d alpha (vertical line) against the null distribution of random-date firing. If the line is deep inside the null cloud, the signal adds no information. If it sits in a tail, the signal is doing real work in that direction.
RSI (Relative Strength Index) (rsi) — bullish 20-day observed alpha versus random-date permutation null (200 iterations)
Permutation null detail — all horizons × both benchmarks
200-iteration null: for each ticker, sample N random dates from its history (matching observed trigger count) and compute the same alpha. The null distribution's 95% CI is where a signal with no information would land. pperm = one-sided fraction of null iters with mean ≥ observed.
Horizon Bench Observed α Null mean Null 95% CI pperm
1d spx -0.01% -0.00% [-0.04%, +0.01%] 0.119
1d msci -0.04% -0.03% [-0.07%, -0.01%] 0.284
1d spxew +0.04% -0.03% [-0.07%, -0.01%] 0.025
5d spx -0.10% +0.28% [-0.07%, +1.34%] 0.995
5d msci -0.16% +0.28% [-0.07%, +1.35%] 1.000
5d spxew +0.03% +0.31% [-0.05%, +1.38%] 0.672
20d spx -0.35% +0.36% [-0.09%, +1.68%] 1.000
20d msci -0.23% +0.48% [+0.03%, +1.81%] 1.000
20d spxew -0.04% +0.55% [+0.09%, +1.88%] 1.000
60d spx +0.07% +0.52% [-0.19%, +1.89%] 0.861
60d msci +0.40% +0.98% [+0.31%, +2.35%] 0.930
60d spxew +0.48% +1.20% [+0.49%, +2.57%] 0.975
252d spx -0.38% +0.57% [-0.48%, +1.69%] 0.965
252d msci +1.42% +2.90% [+1.87%, +4.04%] 1.000
252d spxew +3.01% +4.27% [+3.23%, +5.36%] 0.995

Example triggers on US large-caps (2023+, mcap ≥ $30B)

Six recent bullish RSI triggers on US mega-caps, filtered to |alpha| ≤ 25% to exclude catalyst-driven outliers (earnings surprises, M&A, binary events). The first three are the strongest outcomes — what the signal looks like when it works. The last three are the weakest — what the signal looks like when it fails. Each chart shows the stock's price with signal-appropriate technical overlays (e.g. MACD subpanel on MACD pages, Bollinger Bands on Bollinger pages, the 52-week trailing max line on 52w-high pages), a dot marking the trigger date, and the forward window shaded (green when the signal was right, red when it wasn't). Click any chart to open full-size.

Strongest outcomes (what RSI looks like when it works)
RSI (Relative Strength Index) (rsi) — strongest bullish trigger #1 on a US mega-cap, showing the stock's OHLC candles, moving averages, signal-specific overlays, and the 20-day forward window RSI (Relative Strength Index) (rsi) — strongest bullish trigger #2 on a US mega-cap, showing the stock's OHLC candles, moving averages, signal-specific overlays, and the 20-day forward window RSI (Relative Strength Index) (rsi) — strongest bullish trigger #3 on a US mega-cap, showing the stock's OHLC candles, moving averages, signal-specific overlays, and the 20-day forward window
Weakest outcomes (what RSI looks like when it fails)
RSI (Relative Strength Index) (rsi) — weakest bullish trigger #1 on a US mega-cap, showing the 20-day forward window where the signal failed RSI (Relative Strength Index) (rsi) — weakest bullish trigger #2 on a US mega-cap, showing the 20-day forward window where the signal failed RSI (Relative Strength Index) (rsi) — weakest bullish trigger #3 on a US mega-cap, showing the 20-day forward window where the signal failed
Stock-regime quadrants (2×2 per-stock, 20d alpha detail table)
Each quadrant groups triggers by the stock's own ADX(14) and RV(20) at the trigger date — the textbook conditioning variable (not market-level). Stock %, bench %, alpha %, and HAC p-value shown for each benchmark.
Quadrant N Stock % (spx) Bench % (spx) Alpha % (spx) p (HAC) Stock % (msci) Bench % (msci) Alpha % (msci) p (HAC) Stock % (spxew) Bench % (spxew) Alpha % (spxew) p (HAC)
Trending + Low vol Clean directional grind, low whipsaw 6,691 +0.08% +0.74% -0.64% <0.001 +0.08% +0.61% -0.51% <0.001 +0.08% +0.54% -0.37% <0.001
Trending + High vol Crisis selloff or parabolic rally 54,531 +1.24% +1.60% -0.35% <0.001 +1.24% +1.42% -0.19% 0.0083 +1.24% +1.27% +0.00% 0.9914
Non-trending + Low vol Quiet chop, summer doldrums 2,159 -0.10% +0.50% -0.56% 0.0003 -0.10% +0.39% -0.48% 0.0022 -0.10% +0.28% -0.33% 0.0289
Non-trending + High vol Classical "whipsaw zone" for momentum 14,060 +1.25% +1.37% -0.10% 0.4834 +1.25% +1.36% -0.12% 0.3980 +1.25% +1.21% +0.11% 0.4267
Sub-period breakdown table (20d alpha)
Historical clustering check. If alpha concentrates in one era, the signal's robustness is questionable.
Period N Alpha % (spx) p (HAC) Alpha % (msci) p (HAC) Alpha % (spxew) p (HAC)
2015-2019 2015-01-01 → 2020-01-01 21,821 +0.52% <0.001 +0.70% <0.001 +0.34% 0.0001
2020-2022 2020-01-01 → 2023-01-01 25,305 -0.78% <0.001 -0.39% 0.0007 -0.37% 0.0011
2023-2026 2023-01-01 → 2099-01-01 30,315 -0.60% <0.001 -0.78% <0.001 +0.00% 0.9759

↓ Bearish triggers negative alpha = signal was right (stock underperformed market)

Bench Metric 1d 5d 20d 60d 252d
spx Stock % -0.04% +0.15% +0.41% +2.16% +12.12%
Bench % +0.00% +0.15% +0.83% +2.77% +13.22%
Alpha % -0.04% +0.01% -0.34% -0.58% -1.09%
Median alpha -0.03% -0.13% -0.81% -2.03% -8.00%
Hit rate (α>0) 48.9% 48.4% 45.5% 44.3% 39.8%
p (naive) <0.001 0.7294 <0.001 <0.001 <0.001
p (HAC) <0.001 0.7413 <0.001 <0.001 0.0124
N 108,504 108,190 107,650 105,280 94,050
msci Stock % -0.04% +0.15% +0.41% +2.16% +12.12%
Bench % -0.00% +0.12% +0.68% +2.26% +10.47%
Alpha % -0.04% +0.04% -0.17% -0.05% +1.67%
Median alpha -0.04% -0.09% -0.63% -1.50% -5.27%
Hit rate (α>0) 48.9% 48.9% 46.4% 45.7% 43.3%
p (naive) <0.001 0.0262 <0.001 0.4741 <0.001
p (HAC) <0.001 0.0336 0.0002 0.6653 0.0001
N 108,273 107,874 107,346 104,606 93,625
spxew Stock % -0.04% +0.15% +0.41% +2.16% +12.12%
Bench % +0.02% +0.14% +0.55% +1.91% +9.20%
Alpha % -0.06% +0.03% -0.03% +0.30% +3.09%
Median alpha -0.05% -0.08% -0.48% -1.12% -3.96%
Hit rate (α>0) 48.5% 49.0% 47.3% 46.6% 44.6%
p (naive) <0.001 0.1520 0.4185 <0.001 <0.001
p (HAC) <0.001 0.1716 0.5378 0.0046 <0.001
N 108,151 107,505 106,924 104,655 93,024
Distribution of all 20d alpha outcomes for this direction. Median and winsorized mean shown.
RSI (Relative Strength Index) (rsi) — bearish 20-day alpha histogram showing distribution of per-trigger returns
Observed 20d alpha (vertical line) against the null distribution of random-date firing. If the line is deep inside the null cloud, the signal adds no information. If it sits in a tail, the signal is doing real work in that direction.
RSI (Relative Strength Index) (rsi) — bearish 20-day observed alpha versus random-date permutation null (200 iterations)
Permutation null detail — all horizons × both benchmarks
200-iteration null: for each ticker, sample N random dates from its history (matching observed trigger count) and compute the same alpha. The null distribution's 95% CI is where a signal with no information would land. pperm = one-sided fraction of null iters with mean ≥ observed.
Horizon Bench Observed α Null mean Null 95% CI pperm
1d spx -0.04% -0.01% [-0.02%, +0.01%] 0.005
1d msci -0.04% -0.03% [-0.05%, -0.01%] 0.239
1d spxew -0.06% -0.03% [-0.05%, -0.01%] 0.005
5d spx +0.01% +0.17% [+0.01%, +0.93%] 0.015
5d msci +0.04% +0.18% [+0.02%, +0.94%] 0.149
5d spxew +0.03% +0.20% [+0.03%, +0.96%] 0.010
20d spx -0.34% +0.41% [+0.20%, +0.90%] 0.005
20d msci -0.17% +0.53% [+0.32%, +1.03%] 0.005
20d spxew -0.03% +0.60% [+0.40%, +1.11%] 0.005
60d spx -0.58% +1.02% [+0.72%, +1.58%] 0.005
60d msci -0.05% +1.46% [+1.15%, +2.01%] 0.005
60d spxew +0.30% +1.69% [+1.39%, +2.27%] 0.005
252d spx -1.09% +3.52% [+2.94%, +4.07%] 0.005
252d msci +1.67% +5.82% [+5.26%, +6.36%] 0.005
252d spxew +3.09% +7.19% [+6.61%, +7.75%] 0.005

Example triggers on US large-caps (2023+, mcap ≥ $30B)

Six recent bearish RSI triggers on US mega-caps, filtered to |alpha| ≤ 25% to exclude catalyst-driven outliers (earnings surprises, M&A, binary events). The first three are the strongest outcomes — what the signal looks like when it works. The last three are the weakest — what the signal looks like when it fails. Each chart shows the stock's price with signal-appropriate technical overlays (e.g. MACD subpanel on MACD pages, Bollinger Bands on Bollinger pages, the 52-week trailing max line on 52w-high pages), a dot marking the trigger date, and the forward window shaded (green when the signal was right, red when it wasn't). Click any chart to open full-size.

Strongest outcomes (what RSI looks like when it works)
RSI (Relative Strength Index) (rsi) — strongest bearish trigger #1 on a US mega-cap, showing the stock's OHLC candles, moving averages, signal-specific overlays, and the 20-day forward window RSI (Relative Strength Index) (rsi) — strongest bearish trigger #2 on a US mega-cap, showing the stock's OHLC candles, moving averages, signal-specific overlays, and the 20-day forward window RSI (Relative Strength Index) (rsi) — strongest bearish trigger #3 on a US mega-cap, showing the stock's OHLC candles, moving averages, signal-specific overlays, and the 20-day forward window
Weakest outcomes (what RSI looks like when it fails)
RSI (Relative Strength Index) (rsi) — weakest bearish trigger #1 on a US mega-cap, showing the 20-day forward window where the signal failed RSI (Relative Strength Index) (rsi) — weakest bearish trigger #2 on a US mega-cap, showing the 20-day forward window where the signal failed RSI (Relative Strength Index) (rsi) — weakest bearish trigger #3 on a US mega-cap, showing the 20-day forward window where the signal failed
Stock-regime quadrants (2×2 per-stock, 20d alpha detail table)
Each quadrant groups triggers by the stock's own ADX(14) and RV(20) at the trigger date — the textbook conditioning variable (not market-level). Stock %, bench %, alpha %, and HAC p-value shown for each benchmark.
Quadrant N Stock % (spx) Bench % (spx) Alpha % (spx) p (HAC) Stock % (msci) Bench % (msci) Alpha % (msci) p (HAC) Stock % (spxew) Bench % (spxew) Alpha % (spxew) p (HAC)
Trending + Low vol Clean directional grind, low whipsaw 20,133 +0.07% +0.38% -0.32% <0.001 +0.07% +0.17% -0.09% 0.0784 +0.07% -0.01% +0.08% 0.1370
Trending + High vol Crisis selloff or parabolic rally 66,788 +0.66% +0.98% -0.24% 0.0004 +0.66% +0.83% -0.08% 0.2562 +0.66% +0.69% +0.06% 0.3679
Non-trending + Low vol Quiet chop, summer doldrums 5,350 +0.15% +0.53% -0.35% <0.001 +0.15% +0.30% -0.12% 0.1465 +0.15% +0.21% -0.03% 0.6999
Non-trending + High vol Classical "whipsaw zone" for momentum 16,273 +0.25% +0.82% -0.51% <0.001 +0.25% +0.70% -0.38% 0.0002 +0.25% +0.61% -0.26% 0.0084
Sub-period breakdown table (20d alpha)
Historical clustering check. If alpha concentrates in one era, the signal's robustness is questionable.
Period N Alpha % (spx) p (HAC) Alpha % (msci) p (HAC) Alpha % (spxew) p (HAC)
2015-2019 2015-01-01 → 2020-01-01 34,446 -0.54% <0.001 -0.30% <0.001 -0.19% 0.0023
2020-2022 2020-01-01 → 2023-01-01 31,099 -0.01% 0.9537 +0.10% 0.2613 -0.14% 0.1294
2023-2026 2023-01-01 → 2099-01-01 43,001 -0.39% <0.001 -0.22% 0.0080 +0.24% 0.0040

Methodology and caveats

How to read. Entry at open of T+1 (one trading day after the signal fires on close of T). 20d = open T+1 to close T+20. Alpha = stock return − benchmark return over the same window (Convention A, single-sided, textbook). For bullish triggers, POSITIVE alpha = signal was right. For bearish triggers, NEGATIVE alpha = signal was right (stock underperformed market). No sign-flipping; the direction of the bet determines what "good" looks like. Per-stock regime is each stock's own ADX(14) and RV(20) at the trigger date — not market-wide state.

Three p-values, three robustness tests. (a) p_naive: scipy one-sample t-test on winsorized alphas. Optimistic because overlapping 20d windows on the same ticker inflate effective N. (b) p_hac: Newey-West HAC with lag = horizon — corrects for the overlap and is the academic-finance standard. (c) p_perm: fraction of 200 random-date null iterations with mean ≥ observed. Tests whether the signal beats random date selection at all. A signal that clears all three (pnaive, phac, pperm all < 0.05) has real information; a signal that fails pperm has zero edge even if the t-test says "significant."

Caveats. (i) Universe reflects today's active tickers; delisted losers pruned → survivorship bias. (ii) Mcap ≥ $100M filter uses today's snapshot, not point-in-time — mild lookahead on which stocks enter the sample, not on returns. (iii) Means and p-values use winsorized alphas (1/99 percentile) to prevent data errors from dominating. Medians and hit rates use raw data. (iv) Zero transaction costs assumed. Realistic bid-ask + commissions remove 20–40bps from 20d alpha on US large-caps, more on small-cap. Sub-20bps alpha is noise in practice. (v) Past performance does not predict future results.

How to use this

1 · When to reach for this signal

Caution recommended. Bullish 20d alpha is -0.35% and worse than random — triggering on random dates would have produced better long-side returns. Either direction fails the "beats random" test. Don't use RSI (Relative Strength Index) as a standalone entry trigger. It may still be useful as part of a composite (section 4).

2 · When it works — the setups that drive it

  • Best bullish setup: Non-trending + High vol — alpha -0.10% / 20d on 14,060 historical triggers.
  • Best bearish setup: Trending + High vol — alpha -0.24% / 20d on 66,788 historical triggers.
  • Best era for bullish: 2015-2019 — alpha +0.52% / 20d.
  • Best era for bearish: 2020-2022 — alpha -0.01% / 20d.

3 · When it fails — common false positives

  • Weakest bullish cell: Trending + Low vol — alpha -0.64% / 20d on 6,691 triggers.
  • Weakest bearish cell: Non-trending + High vol — alpha -0.51% / 20d on 16,273 triggers.
  • Worst era for bullish: 2020-2022 — alpha -0.78% / 20d.
  • Worst era for bearish: 2015-2019 — alpha -0.54% / 20d.

Signal-specific failure patterns

Bullish RSI works in sideways markets, fails in megacap bull markets
Sub-period breakdown reveals RSI bullish had +0.52% alpha per 20d in 2015-2019, a normal dispersed market. Then −0.78% in 2020-2022 (COVID whipsaws) and −0.60% in 2023-2026 (AI megacap rally). The signal fires on oversold laggards; when index returns are concentrated in a handful of winners, buying oversold laggards is exactly the wrong bet. Over the full 2015-2026 window the average is −0.35%, p(HAC)<0.001.
evidence: 20d bullish vs SPX: 2015-2019 α=+0.52 (positive), 2020-2022 α=−0.78, 2023-2026 α=−0.60
Bullish fails HARDEST in the 'textbook-favorable' trending regime
Trending + low vol, the cleanest setup by classical TA, produces the most negative bullish alpha (−0.64%, N=6,691 at 20d). The intuition — 'dip in an uptrend, buy the pullback' — does not hold at the equal-weight large-cap US level. The best bullish regime is non-trend high-vol (−0.10%), i.e. chaotic markets where mean-reversion sometimes works. Against every benchmark, bullish RSI fails the permutation null (p_perm=1.000 at 20d vs all four benchmarks).
evidence: 20d bullish by regime vs SPX: trending_low_vol −0.64 (worst), nontrend_high_vol −0.10 (least bad)
Bearish has real edge, magnified at longer horizons
Bearish RSI fires when a stock closes above the 70 threshold (overbought). Under Convention A (single-sided long benchmark), a 'good' bearish signal produces negative alpha — the stock underperforms the market. RSI bearish delivers α=−0.34% at 20d (p(HAC)<1e-13, p_perm=0.005), widening to −0.59% at 60d. That's a real, persistent short-side edge. Most of it comes from stocks that were stretched high and then mean-revert while the broader market continues up.
evidence: vs SPX: bearish 20d α=−0.34 p_perm=0.005, bearish 60d α=−0.59 p_perm=0.005
Bearish 2020-2022 was noise
The bearish edge disappears in 2020-2022 (α=−0.005 at 20d). COVID-era fundamentals broke mean-reversion — stocks that hit overbought kept going up because fiscal/monetary stimulus was re-flating everything. Bearish RSI should be de-emphasized or paused during QE-driven liquidity surges.
evidence: bearish 20d vs SPX: 2015-2019 α=−0.54, 2020-2022 α=−0.005, 2023-2026 α=−0.39

4 · Pairing inside a screen

The statements below describe how this signal relates to others by construction — which indicator family it belongs to, and where same-family redundancy might reduce the independence of evidence inside a Daily Report. These are taxonomic classifications drawn from standard technical-analysis texts; they are not pairing backtests. A multi-signal convergence backtest is planned but not yet run.

Oscillator-family redundancy

RSI belongs to the momentum-oscillator family alongside Stochastics, Williams %R, and CCI — each is constructed from closing price over a short lookback, normalised to a bounded range (Murphy, Technical Analysis of the Financial Markets, 1999; Pring, Technical Analysis Explained, 5th ed. 2014; Kirkpatrick & Dahlquist, Technical Analysis, 3rd ed. 2015). Stacking two or more of these in the same direction within a single Daily Report produces correlated rather than independent evidence.

What would likely rescue this signal

This block calls out the data or conditions that could turn a technically weak signal into a usable one in a composite screen. Based on signal mechanics and the observed failure patterns above; individual combinations are not yet backtested.

  • Regime-gate the bullish side — Bullish RSI worked in 2015-2019 (α=+0.52) and broke in the post-COVID era. Conditioning bullish RSI on 'market breadth > 60%' or 'SPX not within 5% of ATH' might restore historical alpha. Testable hypothesis, not yet run.
  • Fundamental filter for bullish mean-reversion — An oversold stock with improving fundamentals (positive EPS revision, rising revenue) is a different population than one with deteriorating fundamentals. The signal would likely rescue bullish if filtered to names whose earnings estimates have been raised in the last 30 days. Requires commercial fundamentals feed (not yet wired).
  • Tighter bearish cooldown — Bearish RSI alpha widens from −0.04% at 5d to −0.59% at 60d. The edge is a slow drift, not a sharp crack. Exit rules that hold through the full 60d window (rather than stopping out on 10% bounces) better capture the signal's value. Bearish signals on stocks that are already in structural decline compound more reliably.

See also Why technical-only signals don't survive on their own for the broader argument.

5 · Before you act — a 5-point checklist

  1. Normal trading day? Rule out earnings (within ±3 days), ex-dividend, or known corporate-action dates — the signal is almost certainly reading noise, not momentum, in those windows.
  2. Where is price vs its own 50 / 200 DMA? A mean-reversion signal firing against the long-term trend (e.g. oversold in a clean uptrend) is much more reliable than one firing with it.
  3. What's the sector breadth doing? An isolated signal in a broadly down-trending sector is a lower-confidence setup than one firing with the rest of its peer group.
  4. Is ADV20 enough for your size? If the trigger is on a $500M name and you want to move $1M notional, you're the tape. Consider adv20d ≥ 5% of your intended position.
  5. What invalidates you? Define a price level (for longs: a close below the trigger-day low; for shorts: close above the trigger-day high) and honor it. The backtest alpha is an average; any one trade can be at either tail.

Execution notes

Entry = open T+1 (day after signal fires on close). 20d horizon offers the clearest bearish alpha (−0.34%); 60d compounds more (−0.59%) but takes on more event-driven variance. Bullish horizons are a waste of time at every tested length — 60d even flips slightly positive (+0.065%) but not significantly. If you must trade bullish RSI, pair it with a structural filter (fundamental improvement, sector breadth) — the raw oscillator on its own is a value trap.