Ranked by %B ascending — which names are on sale — not by momentum. Measured on the AI complex over the 2023+ wave, forward 60 days: buying dips beat simply holding by +3.47 pts, while buying strength lost 2.65 pts and top-quartile 12-month momentum lost 3.01 pts. Owning the complex at all was worth +17.95%, which is roughly five times what the timing overlay adds. These are daily entries on the names as they stand today (in-sample, overlapping); the quarterly portfolio test further down did not confirm the dip rule.
Descriptive only. Nothing on this page was measured by regime, so it does not change a verdict. Broad risk-off = QQQ below its 200-day with VIX ≥ 25 or up 20% in 5 days; sector washout = SOXX below its 200-day or under 40% of the AI list above theirs.
Descriptive only; no verdict reads it. A theme leads for a few years and then hands over (the COVID winners kept dipping through late 2021 H2 and did not recover; a one-off analysis, not a committed result). "Breaking" = the theme's equal-weight basket is under its own 200-day while at least 40% of its names sit in the dip band at once; that threshold is a display choice, not a tested rule.
| Theme | Names | Basket vs 200d | Basket 20d | Names above 200d | In dip band | State |
|---|---|---|---|---|---|---|
| AI compute | 9 | +34.9% | +16.3% | 89% | 0/9 | leading |
| AI memory | 4 | +30.1% | -4.8% | 75% | 2/4 | leading |
| Chip equipment | 8 | +13.7% | +8.6% | 100% | 0/8 | leading |
| AI networking | 3 | +27.3% | +15.8% | 100% | 0/3 | leading |
| AI power | 5 | +3.5% | +4.4% | 60% | 0/5 | leading |
| AI platforms | 6 | +12.4% | +7.4% | 83% | 1/6 | leading |
| Humanoid supply chain | 16 (15 in basket) | +2.2% | +4.8% | 47% | 1/16 | mixed |
| Magnets & materials | 4 | -16.8% | -9.8% | 25% | 3/4 | breaking |
%B says where price sits inside its normal trading channel. Below 0.20 means it has pulled back to the bottom of that channel — on sale. Above 1.00 is extended: too stretched to buy, and it returned 2.65 pts below simply holding.
Two studies disagree, so this is not a filter the verdict column applies. 46 large caps, 15y, fwd 60d: dips above the 200-day hit 71% vs 68% below — better odds, but below-200d averaged slightly more (+7.16% vs +6.96%). The AI complex, 2023+: dips below the 200-day were the best bucket of all, +7.49 pts vs the null. The Meta trade in the box below was 11.8% below its 200-day. Above = steadier; below = bigger and lumpier. Pick per your tolerance.
Both halves are gates, and 2y pos is often the binding one. p10 is the 10th-percentile outcome across past 2-year holds — your unlucky-but-realistic case; a −50% p10 means one in ten 2-year holds halved. These are the exact thresholds the verdict column applies, not a rule of thumb.
| %B < 0.20 dip rule | 2026-08-21 @ 549.47 | +35.4% | 18 days before the launch, and META was 11.8% below its 200-day |
| move + volume spike | 2026-09-09 @ 653.17 | +13.9% | the day after the launch |
| follow the analyst | 2026-09-21 @ 741.25 | +0.4% | the move was over |
The 3x pair moves with SOXX, so read SOXX's %B and size with SOXL/SOXS's ATR. SOXS is inverse: its %B means the opposite.
| Ticker | Close | 1d | %B | vs 20d | vs 200d | range | ATR | $vol/d | 2y pos | 2y p10 | 60d dips | size | msgs | Verdict |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| SOXX semis ETF semis_etf |
563.28 | -3.35% | 0.57 | +1.5% | +22.3% | 76% | 2.9% | 3.9B | 95% 4w | +7.3% | 73% of 30 med +6% · CVaR -18% |
$1,210 17% | — | neutral |
| SOXL semis ETF semis_etfleveraged |
142.52 | -10.31% | 0.54 | +2.2% | +19.0% | 41% | 8.7% | 7.8B | 76% 4w | -52.2% | 63% of 30 med +9% · CVaR -57% |
— | — | leveraged · not rated |
| SOXS semis ETF semis_etfleveragedinverse |
33.79 | +10.24% | 0.40 | -7.8% | -82.3% | 0% | 10.0% | 2.6B | 0% 4w | -98.7% | 17% of 36 med -21% · CVaR -65% |
— | — | leveraged · not rated |
The ten holdings behind SOXX, their fund weight, and how much of the day each one contributed (fund weight × its move, so together they explain about 61% of the fund).
| Ticker | Fund wt | 1d | Contrib | Trend | vs 20d |
|---|---|---|---|---|---|
| NVDA | 9.43% | -2.94% | -0.278% | above 20d | |
| MU | 8.91% | -4.79% | -0.427% | above 20d | |
| AMD | 8.23% | -3.90% | -0.321% | above 20d | |
| AVGO | 7.48% | -4.35% | -0.325% | above 20d | |
| INTC | 5.09% | -5.34% | -0.272% | BELOW 20d | |
| MRVL | 4.66% | -3.52% | -0.164% | above 20d | |
| TSM earnings in 4 sessions | 4.65% | -3.01% | -0.140% | above 20d | |
| AMAT | 4.59% | -2.13% | -0.098% | above 20d | |
| LRCX | 4.27% | -2.71% | -0.116% | above 20d | |
| KLAC | 4.12% | -0.06% | -0.002% | above 20d |
Breadth is context, not a trigger. Reproduce every number here with scripts/soxx_breadth_study.py (8 years of SOXX, entry at the next open). Broad strength has no edge: top-10 weight above its 20-day > 80% was +0.04 pts vs the null at 20 days (n=727; +0.00 at 5d, −0.03 at 1d). Weakness in the holdings (breadth < 20%, or their composite %B < 0.20) pointed the same way as SOXX's own dip signal but smaller (about +0.7 to +0.9 pts vs +1.3) and rests on a few hundred overlapping observations, so treat it as noise-level: it adds nothing you cannot read off SOXX's own %B. Use this table to see why SOXX moved, broad or one name, and take the direction call from SOXX's %B. SOXL and SOXS are daily-reset products and are not rated on this page.
META, GOOGL, MSFT, AAPL, AMZN, NFLX — tracked every day whether or not they signal.
| Ticker | Close | 1d | %B | vs 20d | vs 200d | range | ATR | $vol/d | 2y pos | 2y p10 | 60d dips | size | msgs | Verdict |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| META platform ndx100ai_platformmegacap |
720.89 | -0.06% | 0.54 | +0.8% | +14.6% | 78% | 3.2% | 9.8B | 79% 4w | -28.1% | 57% of 30 med +3% · CVaR -25% |
$1,103 16% | — | neutral |
| GOOGL platform ndx100ai_platformmegacap |
348.29 | -0.63% | 0.67 | +0.9% | +2.6% | 68% | 2.5% | 8.3B | 90% 4w | -0.1% | 71% of 31 med +5% · CVaR -9% |
$1,400 20% cap | — | neutral |
| MSFT platform ndx100ai_platformmegacap |
522.61 | -1.35% | 0.80 | +2.9% | +20.9% | 92% | 2.2% | 11.1B | 96% 4w | +10.0% | 70% of 27 med +7% · CVaR -13% |
$1,400 20% cap | — | neutral |
| AAPL megacap ndx100megacap |
340.42 | +1.11% | 0.89 | +1.6% | +17.4% | 99% | 1.9% | 12.7B | 99% 4w | +16.2% | 70% of 27 med +6% · CVaR -17% |
$1,400 20% cap | — | neutral |
| AMZN megacap ndx100megacap |
254.06 | -2.25% | 0.65 | +0.9% | +5.0% | 65% | 2.2% | 8.6B | 79% 4w | -25.5% | 73% of 30 med +6% · CVaR -22% |
$1,400 20% cap | — | neutral |
| NFLX megacap ndx100megacap |
71.57 | +2.68% | 0.47 | -0.6% | -14.4% | 8% | 2.6% | 2.5B | 77% 4w | -39.2% | 74% of 31 med +7% · CVaR -37% |
$1,357 19% | — | neutral |
Any symbol, priced now. The browser cannot call Yahoo directly (no CORS), so this
goes through this site's own /api/quote and then applies the same formulas as the
nightly table, ported from the Python — so a live check and the tables cannot disagree.
Cached 10 minutes.
%B below 0.2, liquid, and passes the 1–3 year hold gate (≥80% of rolling 2-year windows positive with a 10th percentile above -10%).
| Ticker | Close | 1d | %B | vs 20d | vs 200d | range | ATR | $vol/d | 2y pos | 2y p10 | 60d dips | size | msgs | Verdict |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ORCL platform ai_platform |
135.69 | -5.48% | 0.16 | -4.8% | -15.9% | 11% | 4.8% | 3.9B | 98% 4w | +14.1% | 60% of 30 med +3% · CVaR -23% |
$731 10% | 30 | BUY ZONE |
| VRTX ndx100 |
503.25 | -0.47% | 0.16 | -2.1% | +6.0% | 64% | 2.3% | 542M | 92% 4w | +1.7% | 60% of 30 med +3% · CVaR -12% |
$1,400 20% cap | 0 | BUY ZONE |
| ROST ndx100 |
225.20 | -0.15% | 0.19 | -2.2% | +3.3% | 73% | 2.2% | 504M | 80% 4w | -7.0% | 64% of 28 med +6% · CVaR -20% |
$1,400 20% cap | 0 | BUY ZONE |
A move of 5%+ on 2×+ the 20-day average volume. This is a reaction, not a setup — something happened, and the rule does not know what or which direction it resolves.
| Ticker | Close | 1d | %B | vs 20d | vs 200d | range | ATR | $vol/d | 2y pos | 2y p10 | 60d dips | size | msgs | Verdict |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| NOVT humanoid humanoid_disclosed |
136.33 | -5.09% | 0.25 | -3.7% | -3.2% | 53% | 4.3% | 52M | 75% 4w | -21.7% | 67% of 33 med +10% · CVaR -22% |
$808 12% | 0 | neutral SPIKE -5.1% · 2.0x |
Dips that fail the hold gate, plus names with no measured edge either way.
| Ticker | Close | 1d | %B | vs 20d | vs 200d | range | ATR | $vol/d | 2y pos | 2y p10 | 60d dips | size | msgs | Verdict |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| WDC memory ndx100ai_memory |
393.31 | -2.99% | -0.00 | -10.1% | -2.8% | 44% | 7.0% | 3.2B | 53% 4w | -43.6% | 47% of 30 med -2% · CVaR -28% |
$497 7% | 13 | DIP · FAILS HOLD |
| ALB materials humanoid_materials |
102.10 | -0.63% | 0.10 | -6.7% | -33.6% | 12% | 4.0% | 265M | 49% 4w | -62.4% | 56% of 32 med +2% · CVaR -27% |
$872 12% | 1 | DIP · FAILS HOLD |
| STX memory ndx100ai_memory |
774.83 | -4.05% | 0.13 | -10.0% | +17.3% | 64% | 7.1% | 3.8B | 80% 4w | -13.2% | 68% of 31 med +12% · CVaR -24% |
$496 7% | 6 | DIP · FAILS HOLD |
| SBUX ndx100 |
93.21 | -0.40% | 0.15 | -2.3% | -5.0% | 52% | 2.6% | 657M | 75% 4w | -8.6% | 47% of 30 med -1% · CVaR -18% |
$1,324 19% | 22 | DIP · FAILS HOLD |
| QCOM ndx100 |
176.01 | -0.63% | 0.15 | -5.5% | +4.8% | 42% | 4.3% | 1.8B | 75% 4w | -13.6% | 52% of 29 med +0% · CVaR -22% |
$817 12% | 9 | DIP · FAILS HOLD |
| ALNY ndx100 |
224.53 | +0.05% | 0.17 | -6.6% | -25.0% | 7% | 4.5% | 368M | 79% 4w | -8.6% | 53% of 32 med +4% · CVaR -23% |
$786 11% | 0 | DIP · FAILS HOLD |
| BOTZ humanoid ETF humanoid_etf |
34.83 | -2.35% | 0.21 | -1.5% | -5.5% | 30% | 1.6% | 23M | 69% 4w | -28.2% | 55% of 29 med +4% · CVaR -21% |
$1,400 20% cap | — | neutral |
| DXCM ndx100 |
84.41 | +0.13% | 0.21 | -2.4% | +15.0% | 79% | 2.6% | 315M | 64% 4w | -38.3% | 58% of 31 med +10% · CVaR -27% |
$1,323 19% | — | neutral |
| REGN ndx100 |
739.57 | -0.34% | 0.22 | -3.7% | +0.7% | 62% | 2.8% | 503M | 63% 4w | -31.3% | 48% of 29 med -0% · CVaR -25% |
$1,259 18% | — | neutral |
| KHC ndx100 |
22.48 | +2.27% | 0.25 | -4.3% | -3.2% | 30% | 2.6% | 359M | 37% 4w | -58.1% | 40% of 35 med -1% · CVaR -17% |
$1,357 19% | — | neutral |
| NOVT humanoid humanoid_disclosed |
136.33 | -5.09% | 0.25 | -3.7% | -3.2% | 53% | 4.3% | 52M | 75% 4w | -21.7% | 67% of 33 med +10% · CVaR -22% |
$808 12% | 0 | neutral SPIKE -5.1% · 2.0x |
| AXON ndx100 |
417.31 | +2.79% | 0.27 | -4.9% | -15.5% | 17% | 4.9% | 429M | 98% 4w | +20.8% | 59% of 32 med +20% · CVaR -28% |
$711 10% | — | neutral |
| CMCSA ndx100 |
21.21 | +1.29% | 0.29 | -4.3% | -17.8% | 3% | 2.8% | 656M | 56% 4w | -27.1% | 47% of 32 med -0% · CVaR -19% |
$1,267 18% | — | neutral |
| PCAR ndx100 |
109.66 | +2.62% | 0.30 | -3.4% | -8.8% | 40% | 2.5% | 355M | 98% 4w | +8.1% | 61% of 31 med +6% · CVaR -17% |
$1,400 20% cap | — | neutral |
| BIIB ndx100 |
218.43 | -1.67% | 0.30 | -1.6% | +11.3% | 88% | 2.7% | 184M | 33% 4w | -45.3% | 50% of 32 med +0% · CVaR -26% |
$1,309 19% | — | neutral |
| INTC ndx100soxx_top10 |
107.08 | -5.34% | 0.31 | -5.9% | +29.5% | 68% | 5.6% | 10.3B | 49% 4w | -37.8% | 58% of 33 med +3% · CVaR -24% |
$623 9% | — | neutral |
| HON ndx100 |
206.60 | -0.73% | 0.33 | -1.3% | -9.8% | 19% | 2.1% | 598M | 88% 4w | -2.2% | 57% of 30 med +2% · CVaR -12% |
$1,400 20% cap | — | neutral |
| BKNG ndx100 |
159.97 | +2.63% | 0.33 | -2.5% | -11.5% | 10% | 3.6% | 1.1B | 84% 4w | -5.0% | 75% of 32 med +4% · CVaR -19% |
$980 14% | — | neutral |
| CPRT ndx100 |
27.48 | +3.23% | 0.33 | -3.5% | -17.4% | 4% | 3.1% | 282M | 86% 4w | -18.4% | 62% of 32 med +8% · CVaR -17% |
$1,122 16% | — | neutral |
| TTD ndx100 |
12.34 | +2.07% | 0.34 | -5.4% | -44.2% | 1% | 4.1% | 272M | 69% 4w | -48.0% | 57% of 30 med +11% · CVaR -40% |
$848 12% | — | neutral |
| JBL humanoid humanoid_disclosed |
299.16 | -0.10% | 0.34 | -1.8% | -1.3% | 55% | 4.4% | 319M | 95% 4w | +8.4% | 61% of 31 med +6% · CVaR -14% |
$792 11% | — | neutral |
| BKR ndx100 |
56.30 | +1.61% | 0.36 | -0.9% | -5.0% | 49% | 2.9% | 409M | 66% 4w | -44.0% | 43% of 30 med -0% · CVaR -29% |
$1,209 17% | — | neutral |
| VRT power ai_power |
243.73 | -1.12% | 0.36 | -1.3% | -8.5% | 42% | 4.7% | 1.1B | 80% 4w | -26.0% | 80% of 25 med +17% · CVaR -34% |
$739 11% | — | neutral |
| WDAY ndx100 |
187.81 | +1.93% | 0.36 | -1.0% | +19.9% | 57% | 3.8% | 568M | 67% 4w | -28.7% | 70% of 33 med +8% · CVaR -34% |
$909 13% | — | neutral |
| PH humanoid humanoid_association |
946.96 | -0.61% | 0.37 | -1.0% | +0.3% | 65% | 2.2% | 559M | 97% 4w | +7.4% | 68% of 31 med +7% · CVaR -23% |
$1,400 20% cap | — | neutral |
%B above 1.00 — the stall zone, −2.65 pts vs the null at 60 days. Not a short signal; a not-today signal.
| Ticker | Close | 1d | %B | vs 20d | vs 200d | range | ATR | $vol/d | 2y pos | 2y p10 | 60d dips | size | msgs | Verdict |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| COST ndx100 |
947.92 | +0.60% | 1.04 | +3.8% | -1.4% | 41% | 1.6% | 1.9B | 100% 4w | +26.5% | 60% of 30 med +7% · CVaR -9% |
$1,400 20% cap | — | extended |
Buy the dip names at each quarter-end, hold to the next, versus the same number of names drawn at random from the same list (2000 draws), after 0.20% costs. Thresholds were written before each run. Neither variant cleared them, so nothing here changes the verdicts above. The universe is today's AI leaders, which flatters dip entries; the mean is carried by one quarter (2026-03-31) in v1.
| Test | signal qtrs | mean net excess | qtrs positive | ex best qtr | p vs random | result |
|---|---|---|---|---|---|---|
| v1 · dip (%B < 0.20) | 23 | +4.2 pts | 52% | +1.0 pts | 0.071 | not promoted (failed: P1, P4) |
| v2 · dip + above 200-day | 18 | +3.3 pts | 39% | -4.8 pts | 0.259 | not promoted (failed: P1, P3, P4, P5, P6) |
The 1–3 year durability gate cannot be tested point-in-time on 10 years of data (needs 2016 prior sessions); it is judged forward only. Method and thresholds: docs/experiments/2026-09-24-preregistration-qr-s1-v{1,2}.md.
Daily entries at the next open on the 34 AI stocks over 10 years, edge vs every session for the same names. The rule was fixed before this ran: the grid can only demote the default, never pick a different cell. Result: not demoted — it sits on a plateau. Deeper dips test a little better and every cell is positive, but the default edge is +1.31 pts over 10 years, well below the +3.47 pts measured on 2023+ alone.
| edge vs null, pts | 20d | 40d | 60d | 90d |
|---|---|---|---|---|
| %B < 0.10 | +1.24 | +1.52 | +1.72 | +1.34 |
| %B < 0.15 | +1.13 | +1.35 | +1.52 | +1.09 |
| %B < 0.20 | +0.98 | +1.17 | +1.31 | +0.63 |
| %B < 0.25 | +0.85 | +1.05 | +1.28 | +0.57 |
| %B < 0.30 | +0.77 | +1.00 | +1.31 | +0.59 |
By year (default cell): 2016 +0.5% · 2017 +2.2% · 2018 +2.1% · 2019 +4.1% · 2020 +8.1% · 2021 +3.2% · 2022 +3.8% · 2023 +4.4% · 2024 +3.5% · 2025 -3.3% · 2026 +4.2%
By market: QQQ above 200 +0.72% · QQQ below 200 +2.16%
Same rule on indices since 1999 (no hand-picked names): QQQ +0.58% · SPY +0.85% · XLK +0.39% — positive on at least 2 of 3. Episodes are counted per name, and the names dip together, so the independent count is far lower than the raw one. Source: scripts/quarterly_rotation/robustness.py.