Pattern hh_hl_structure

HH/HL Trend Structure

Detects trend structure shifts using swing highs and lows. Bullish: last two swing highs form a Higher High (exceeding z-score tolerance), then price pulls back to form a Higher Low above the previous swing low, confirmed when the pullback low holds for the confirmation window (default 1 bar). Bearish is the mirror (Lower Low + Lower High). Z-score tolerance adapts to each stock's volatility — tighter for low-vol, looser for high-vol stocks.

Signal family

Pattern — Formal chart-pattern detectors (double tops / bottoms, failed breakouts, HH/HL structure).

Parameters

Name Description Default Range
swing_window Swing detection window (bars each side) 10 3–30
confirmation_window Confirmation window for forming HL/LH 1 1–15
tolerance_zscore Min HH/LL move (z-scores of daily vol) 1.5 0.5–4.0
vol_window Volatility computation window 252 60–504
lookback Lookback window (bars) 252 60–504

Historical context

550,314 triggers on 23,363 tickers, 1996-07-30 → 2026-05-01. Universe: global — all covered exchanges (mcap ≥ $100,000,000, price ≥ $1). Long-only convention: BUY at open T+1, hold the horizon, compare to S&P 500 Equal Weight over the same window.

Methodology footnotes

Benchmarks shown in the detail tables: spxew (S&P 500 Equal Weight — primary, median-stock view, avoids the 2020+ megacap-concentration distortion), spx (S&P 500 cap-weighted, distorted post-2020), msci (MSCI World USD). Per-stock regime tags: trending = ADX(14) ≥ 25, high vol = 20d realized annualized vol ≥ 20%. 1d return = intraday T+1 open→close; 20d = open T+1 to close T+20.

At a glance — alpha vs S&P 500 Equal Weight, global universe

Holding-period sensitivity. Bullish columns: positive = signal worked (long the trigger beat the index). Bearish columns: negative = signal worked (the flagged stock underperformed).

Horizon Bullish α Bearish α
5-day -0.24% -0.29%
20-day -0.33% +0.13%
60-day -0.03% +0.28%
1-year +3.21% +1.10%

Sign flip across horizons. Bullish raw alpha moves from -0.24% (5d) to +3.21% (1y). Bearish raw alpha moves from -0.29% (5d) to +1.10% (1y). Check the permutation rows for those horizons before reading the longer-hold number as edge — raw alpha alone does not distinguish signal timing from universe drift.

Random-date null check (20-day): Bullish: worse than random (p=1.000).
Bearish: worse than random (p=1.000).

Where does HH_HL_STRUCTURE actually fire?

The bucket distribution often reveals what the signal really is, regardless of its textbook label. Heavy concentration in "non-trending + high vol" = it's mostly a chop-market event. Heavy in "trending + low vol" = it picks up the smooth grinds. Read the chart before the alpha numbers — context shapes everything that follows.

HH/HL Trend Structure (hh_hl_structure) — trigger count distribution by per-stock regime quadrant (trending/non-trending × high/low realized volatility) for bullish and bearish triggers, global universe

Does it work in every regime?

Trigger alpha split by the host stock's own regime on the trigger date — trending or ranging, high-vol or low-vol. The 20d alpha you'd actually capture if you took the trade. Bars matching your direction's "right" sign (positive for bullish, negative for bearish) = the signal worked in that regime; opposite sign = avoid it there. A signal with one strong-positive bar and three flat ones isn't a "20d alpha" signal — it's a "20d alpha when the stock is X" signal.

HH/HL Trend Structure (hh_hl_structure) — mean 20-day alpha versus S&P 500 Equal Weight by per-stock regime quadrant, bullish and bearish triggers 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.

Does it work in every era?

A multi-year average can hide major instability. The sample splits into three windows: 2015–2019 (pre-COVID), 2020–2022 (pandemic + 2022 bear), and 2023+ (post-ZIRP + AI megacap rally). All three matching your direction's "right" sign = the signal is durable. One era doing all the work = a regime-specific edge that may not repeat. The bigger the variance across eras, the smaller the position you should run.

HH/HL Trend Structure (hh_hl_structure) — 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
spxew Stock % +0.11% -0.07% +0.38% +2.31% +13.13%
Bench % +0.05% +0.17% +0.77% +2.38% +10.02%
Alpha % +0.04% -0.24% -0.33% -0.03% +3.21%
Median alpha -0.03% -0.31% -1.12% -2.31% -5.98%
Hit rate (α>0) 49.2% 46.6% 44.5% 43.8% 42.8%
p (naive) <0.001 <0.001 <0.001 0.4952 <0.001
p (HAC) <0.001 <0.001 <0.001 0.5641 <0.001
N 281,645 272,294 272,678 263,097 231,365
spx Stock % +0.11% -0.07% +0.38% +2.31% +13.13%
Bench % +0.02% +0.21% +1.01% +3.19% +13.78%
Alpha % +0.08% -0.27% -0.59% -0.90% -0.69%
Median alpha -0.01% -0.35% -1.43% -3.27% -9.96%
Hit rate (α>0) 49.8% 46.1% 43.1% 41.5% 38.7%
p (naive) <0.001 <0.001 <0.001 <0.001 <0.001
p (HAC) <0.001 <0.001 <0.001 <0.001 0.0038
N 283,315 275,282 275,661 266,944 234,090
msci Stock % +0.11% -0.07% +0.38% +2.31% +13.13%
Bench % +0.07% +0.24% +0.90% +2.78% +11.35%
Alpha % +0.03% -0.29% -0.48% -0.46% +1.65%
Median alpha -0.06% -0.39% -1.35% -2.86% -7.50%
Hit rate (α>0) 48.6% 45.7% 43.4% 42.4% 41.2%
p (naive) <0.001 <0.001 <0.001 <0.001 <0.001
p (HAC) <0.001 <0.001 <0.001 <0.001 <0.001
N 282,570 274,398 274,087 264,321 232,578
Observed 20-day lift (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 the right tail, the signal is doing real work in that direction; in the left tail it ran inverted — random dates served the bullish case better than its own triggers did.
HH/HL Trend Structure (hh_hl_structure) — bullish 20-day observed lift versus random-date permutation null (200 iterations)
Permutation null detail — all horizons × each benchmark
200-iteration null: for each ticker, sample N random dates from its history (matching observed trigger count) and compute the same alpha. Both observed and null are baseline-centered per ticker (each ticker's own baseline alpha is subtracted), which removes the universe-selection lift that all surviving names share, so the comparison is about the trigger's timing. It does not put the null at zero: that baseline is a median, while a random date's expectation is the ticker's mean, so the null settles at the gap between the two — the right-skew of the ticker's own alpha distribution, positive for equities. Read observed minus null, not the absolute position of either column — which is also why these figures do not match the α columns in the tables above (those are raw, uncentered means). pperm = one-sided fraction of null iters with mean in the "signal was right" tail (right for bullish, left for bearish), so it runs from a floor of 0.005 (trigger dates beat every random draw in the claimed direction) to a 1.000 ceiling (every random draw served the claimed direction better — a reliably inverted signal, not an absent one). The floor is 1/(iterations + 1) and moves with the iteration count; the ceiling does not. Either end says the result was reliable, not that it was large: the gap between the line and the cloud can be a fraction of a percent and still reach an endpoint, and the transaction-cost floor in the caveats below would swallow a gap that small. Read the size off the α and hit-rate columns, the reliability off pperm.
Horizon Bench Observed lift Null mean Null 95% CI pperm
1d spxew +0.17% +0.08% [+0.06%, +0.09%] 0.005
1d spx +0.19% +0.08% [+0.07%, +0.09%] 0.005
1d msci +0.18% +0.09% [+0.08%, +0.10%] 0.005
5d spxew +0.09% +0.34% [+0.32%, +0.36%] 1.000
5d spx +0.10% +0.35% [+0.33%, +0.38%] 1.000
5d msci +0.08% +0.36% [+0.34%, +0.39%] 1.000
20d spxew +0.55% +1.11% [+1.06%, +1.15%] 1.000
20d spx +0.52% +1.13% [+1.09%, +1.18%] 1.000
20d msci +0.53% +1.15% [+1.10%, +1.20%] 1.000
60d spxew +1.52% +2.38% [+2.31%, +2.45%] 1.000
60d spx +1.40% +2.44% [+2.37%, +2.52%] 1.000
60d msci +1.43% +2.47% [+2.40%, +2.53%] 1.000
252d spxew +3.31% +4.58% [+4.45%, +4.71%] 1.000
252d spx +3.55% +4.92% [+4.79%, +5.05%] 1.000
252d msci +3.54% +4.87% [+4.73%, +4.99%] 1.000

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

Six recent bullish HH_HL_STRUCTURE triggers on US mega-caps. Top three: the signal's best outcomes. Bottom three: the worst. Extreme outliers (|α| > 25%) excluded. The three best and three worst are still tail outcomes by construction — read them as the range, not the typical result.

Strongest outcomes (what HH_HL_STRUCTURE looks like when it works)
HH/HL Trend Structure (hh_hl_structure) — 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 HH/HL Trend Structure (hh_hl_structure) — 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 HH/HL Trend Structure (hh_hl_structure) — 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 HH_HL_STRUCTURE looks like when it fails)
HH/HL Trend Structure (hh_hl_structure) — weakest bullish trigger #1 on a US mega-cap, showing the 20-day forward window where the signal failed HH/HL Trend Structure (hh_hl_structure) — weakest bullish trigger #2 on a US mega-cap, showing the 20-day forward window where the signal failed HH/HL Trend Structure (hh_hl_structure) — 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 % (spxew) Bench % (spxew) Alpha % (spxew) p (HAC) Stock % (spx) Bench % (spx) Alpha % (spx) p (HAC) Stock % (msci) Bench % (msci) Alpha % (msci) p (HAC)
Trending + Low vol Clean directional grind, low whipsaw 17,374 +0.17% +0.45% -0.25% <0.001 +0.17% +0.81% -0.62% <0.001 +0.17% +0.69% -0.50% <0.001
Trending + High vol Crisis selloff or parabolic rally 168,349 +0.49% +0.76% -0.20% <0.001 +0.49% +1.03% -0.48% <0.001 +0.49% +0.90% -0.34% <0.001
Non-trending + Low vol Quiet chop, summer doldrums 14,814 +0.17% +0.49% -0.27% <0.001 +0.17% +0.76% -0.56% <0.001 +0.17% +0.64% -0.43% <0.001
Non-trending + High vol Classical "whipsaw zone" for momentum 90,677 +0.29% +0.88% -0.54% <0.001 +0.29% +1.07% -0.76% <0.001 +0.29% +0.98% -0.68% <0.001
Sub-period breakdown table (20d alpha)
Historical clustering check. If alpha concentrates in one era, the signal's robustness is questionable.
Period N Alpha % (spxew) p (HAC) Alpha % (spx) p (HAC) Alpha % (msci) p (HAC)
2015-2019 2015-01-01 → 2020-01-01 79,500 -0.76% <0.001 -0.95% <0.001 -0.81% <0.001
2020-2022 2020-01-01 → 2023-01-01 90,318 -0.74% <0.001 -0.58% <0.001 -0.49% <0.001
2023-2026 2023-01-01 → 2099-01-01 121,358 +0.28% <0.001 -0.35% <0.001 -0.23% <0.001

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

Bench Metric 1d 5d 20d 60d 252d
spxew Stock % -0.19% +0.24% +1.57% +3.35% +12.65%
Bench % +0.06% +0.51% +1.49% +3.13% +11.08%
Alpha % -0.27% -0.29% +0.13% +0.28% +1.10%
Median alpha -0.24% -0.51% -0.71% -1.75% -6.71%
Hit rate (α>0) 44.2% 44.5% 46.6% 45.3% 41.7%
p (naive) <0.001 <0.001 <0.001 <0.001 <0.001
p (HAC) <0.001 <0.001 <0.001 <0.001 <0.001
N 249,080 238,447 239,501 233,198 210,763
spx Stock % -0.19% +0.24% +1.57% +3.35% +12.65%
Bench % +0.05% +0.49% +1.51% +3.58% +14.83%
Alpha % -0.24% -0.28% +0.11% -0.19% -2.12%
Median alpha -0.23% -0.51% -0.74% -2.21% -10.38%
Hit rate (α>0) 44.0% 44.5% 46.4% 44.2% 38.1%
p (naive) <0.001 <0.001 <0.001 <0.001 <0.001
p (HAC) <0.001 <0.001 <0.001 <0.001 <0.001
N 251,051 243,226 242,945 236,819 218,381
msci Stock % -0.19% +0.24% +1.57% +3.35% +12.65%
Bench % +0.06% +0.48% +1.38% +3.19% +12.23%
Alpha % -0.26% -0.26% +0.27% +0.23% -0.25%
Median alpha -0.25% -0.50% -0.61% -1.83% -8.28%
Hit rate (α>0) 43.5% 44.6% 47.1% 45.1% 40.2%
p (naive) <0.001 <0.001 <0.001 <0.001 0.0202
p (HAC) <0.001 <0.001 <0.001 <0.001 0.1602
N 249,712 241,299 241,363 235,655 212,210
Observed 20-day lift (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 the left tail, the signal is doing real work in that direction; in the right tail it ran inverted — random dates served the bearish case better than its own triggers did.
HH/HL Trend Structure (hh_hl_structure) — bearish 20-day observed lift versus random-date permutation null (200 iterations)
Permutation null detail — all horizons × each benchmark
200-iteration null: for each ticker, sample N random dates from its history (matching observed trigger count) and compute the same alpha. Both observed and null are baseline-centered per ticker (each ticker's own baseline alpha is subtracted), which removes the universe-selection lift that all surviving names share, so the comparison is about the trigger's timing. It does not put the null at zero: that baseline is a median, while a random date's expectation is the ticker's mean, so the null settles at the gap between the two — the right-skew of the ticker's own alpha distribution, positive for equities. Read observed minus null, not the absolute position of either column — which is also why these figures do not match the α columns in the tables above (those are raw, uncentered means). pperm = one-sided fraction of null iters with mean in the "signal was right" tail (right for bullish, left for bearish), so it runs from a floor of 0.005 (trigger dates beat every random draw in the claimed direction) to a 1.000 ceiling (every random draw served the claimed direction better — a reliably inverted signal, not an absent one). The floor is 1/(iterations + 1) and moves with the iteration count; the ceiling does not. Either end says the result was reliable, not that it was large: the gap between the line and the cloud can be a fraction of a percent and still reach an endpoint, and the transaction-cost floor in the caveats below would swallow a gap that small. Read the size off the α and hit-rate columns, the reliability off pperm.
Horizon Bench Observed lift Null mean Null 95% CI pperm
1d spxew -0.13% +0.08% [+0.07%, +0.09%] 0.005
1d spx -0.13% +0.09% [+0.08%, +0.10%] 0.005
1d msci -0.12% +0.09% [+0.08%, +0.10%] 0.005
5d spxew +0.11% +0.35% [+0.33%, +0.38%] 0.005
5d spx +0.16% +0.37% [+0.34%, +0.39%] 0.005
5d msci +0.18% +0.37% [+0.35%, +0.40%] 0.005
20d spxew +1.30% +1.15% [+1.09%, +1.21%] 1.000
20d spx +1.51% +1.18% [+1.12%, +1.23%] 1.000
20d msci +1.56% +1.19% [+1.14%, +1.25%] 1.000
60d spxew +2.79% +2.51% [+2.43%, +2.60%] 1.000
60d spx +3.07% +2.58% [+2.49%, +2.66%] 1.000
60d msci +3.08% +2.60% [+2.52%, +2.69%] 1.000
252d spxew +5.49% +5.04% [+4.87%, +5.20%] 1.000
252d spx +6.17% +5.37% [+5.18%, +5.53%] 1.000
252d msci +5.91% +5.32% [+5.16%, +5.48%] 1.000

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

Six recent bearish HH_HL_STRUCTURE triggers on US mega-caps. Top three: the signal's best outcomes. Bottom three: the worst. Extreme outliers (|α| > 25%) excluded. The three best and three worst are still tail outcomes by construction — read them as the range, not the typical result.

Strongest outcomes (what HH_HL_STRUCTURE looks like when it works)
HH/HL Trend Structure (hh_hl_structure) — 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 HH/HL Trend Structure (hh_hl_structure) — 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 HH/HL Trend Structure (hh_hl_structure) — 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 HH_HL_STRUCTURE looks like when it fails)
HH/HL Trend Structure (hh_hl_structure) — weakest bearish trigger #1 on a US mega-cap, showing the 20-day forward window where the signal failed HH/HL Trend Structure (hh_hl_structure) — weakest bearish trigger #2 on a US mega-cap, showing the 20-day forward window where the signal failed HH/HL Trend Structure (hh_hl_structure) — 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 % (spxew) Bench % (spxew) Alpha % (spxew) p (HAC) Stock % (spx) Bench % (spx) Alpha % (spx) p (HAC) Stock % (msci) Bench % (msci) Alpha % (msci) p (HAC)
Trending + Low vol Clean directional grind, low whipsaw 10,565 +0.56% +0.81% -0.19% 0.0029 +0.56% +0.96% -0.36% <0.001 +0.56% +0.89% -0.26% <0.001
Trending + High vol Crisis selloff or parabolic rally 119,839 +1.98% +1.86% +0.15% <0.001 +1.98% +1.82% +0.18% <0.001 +1.98% +1.65% +0.41% <0.001
Non-trending + Low vol Quiet chop, summer doldrums 11,874 +0.52% +0.62% -0.05% 0.3658 +0.52% +0.82% -0.27% <0.001 +0.52% +0.74% -0.18% 0.0024
Non-trending + High vol Classical "whipsaw zone" for momentum 116,822 +1.36% +1.28% +0.16% <0.001 +1.36% +1.31% +0.13% 0.0003 +1.36% +1.24% +0.24% <0.001
Sub-period breakdown table (20d alpha)
Historical clustering check. If alpha concentrates in one era, the signal's robustness is questionable.
Period N Alpha % (spxew) p (HAC) Alpha % (spx) p (HAC) Alpha % (msci) p (HAC)
2015-2019 2015-01-01 → 2020-01-01 75,856 +0.04% 0.2364 +0.02% 0.4932 +0.20% <0.001
2020-2022 2020-01-01 → 2023-01-01 81,332 +0.31% <0.001 +0.70% <0.001 +1.06% <0.001
2023-2026 2023-01-01 → 2099-01-01 101,873 +0.05% 0.2109 -0.30% <0.001 -0.31% <0.001

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: one-sided fraction of 200 random-date null iterations falling in the “signal was right” tail (mean ≥ observed for bullish; mean ≤ observed for bearish). 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 not beaten random timing whatever the t-test says — and because the test is one-sided, a pperm up at its 1.000 ceiling is not "no edge" but inverted edge: every random draw served the claimed direction better than the trigger dates did.

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

Not a standalone entry trigger at 20 days. Bullish 20d alpha is -0.33% — worse than random : firing on random dates would have done better. Bearish 20d alpha is +0.13% — worse than random : firing on random dates would have done better. Fires are screening context inside a composite (section 4), not entries.

These verdicts are 20-day holds vs S&P 500 Equal Weight. Longer horizons can differ in either direction — check the permutation detail tables below before extrapolating.

2 · When it works — the setups that drive it

  • Least-bad bullish cell: Trending + High vol — alpha -0.20% / 20d on 168,349 triggers — still wrong-signed; no bullish cell produced positive alpha.
  • Best bearish setup: Trending + Low vol — alpha -0.19% / 20d on 10,565 historical triggers.
  • Best era for bullish: 2023-2026 — alpha +0.28% / 20d on 121,358 triggers.
  • Least-bad era for bearish: 2015-2019 — alpha +0.04% / 20d on 75,856 triggers — still wrong-signed; no era produced negative alpha.

3 · When it fails — common false positives

  • Weakest bullish cell: Non-trending + High vol — alpha -0.54% / 20d on 90,677 triggers.
  • Weakest bearish cell: Non-trending + High vol — alpha +0.16% / 20d on 116,822 triggers.
  • Worst era for bullish: 2015-2019 — alpha -0.76% / 20d on 79,500 triggers.
  • Worst era for bearish: 2020-2022 — alpha +0.31% / 20d on 81,332 triggers.

Signal-specific failure patterns

Bullish at short horizons looks like a loser, but compounds positive at 1 year
HH/HL structure (confirmed swing-high and swing-low both rising) carries a structural cost: late confirmation. By the time the structure prints, the easy phase of the trend is over, so short-window alpha tends to be weak. Whether longer holds fare better in the current run — and whether any longer-hold number beats random-date firing — is answered by the at-a-glance table and the permutation rows, not by the raw sign alone. See the at-a-glance table for current values.
Bearish side is noise on equal-weight
Bearish HH/HL structure (confirmed lower-highs and lower-lows) does not produce a clean short signal on the equal-weight benchmark. The signal identifies stocks weaker than the cap-weighted index but not weaker than the equal-weight median — marginal as a short trigger.
Why the bullish side is broken at short horizons
The signal identifies stocks in established uptrends via swing structure. By the time structure is visible and confirmed, the trend is typically 10-20 weeks old. Buying confirmed-uptrend stocks at the 10-20 week mark catches them near the point where mean-reversion kicks in or the trend matures. The alpha emerges at horizons long enough for the trend to fully play out (1 year), not at 20 days.

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. Measured pair results — same-day co-fires put through the pair backtest — follow under “Measured pairings” below.

Trend-structure family

Higher-high / higher-low structure encodes the classical Dow-theory definition of an uptrend — a price structure of successively higher swing highs and swing lows (Dow theory as presented in Murphy, Technical Analysis of the Financial Markets, 1999; Edwards & Magee, Technical Analysis of Stock Trends, 11th ed. 2018). This overlaps with HH/HL streak, moving-average crossover, and long-term trend-break signals, which infer the same trend state from different measurements.

Measured pairings — Bonferroni survivors

Beyond the literature pairings above, these are the same-day co-fire combinations involving HH/HL Trend Structure that cleared the pair backtest's Bonferroni cut on the full 2016–2026 sample (549 pairs × 5 horizons = 2,745 hypotheses), on universes filtered to ADV ≥ $5M, price ≥ $5 and market cap ≥ $100M. That cut is two-sided: it asks only whether the co-fire's α is reliably different from zero, in either direction, so a pair can survive by reliably underperforming — 6 of the 15 rows below do exactly that (negative full-sample α). The same run holds out 2023+: the Test columns are that held-out window, printed for every row with enough held-out co-fires to measure, so a survivor that did not repeat out of sample is visible rather than hidden. All α figures here are for holding the stock long after the co-fire — no shorting assumed, and no sign flip for bearish legs. So positive α means the co-fire was followed by outperformance and negative α by underperformance, whichever way either leg points — a bearish leg does not flip the reading. Survivors are rare by design — absence of a pair here means it did not clear the cut, not that it was untested. Ranked by held-out (2023+) α. Historical tendencies, not recommendations.

US (NYSE / NASDAQ / AMEX)

Pair (same-day co-fire, long) Full α Test α (2023+) Test N p_perm test
hh_hl_structure bullish + weekly_change bullish -2.78% -1.15% 192 0.437
hh_hl_structure bullish + stochastics bearish -2.69% -2.74% 82 0.006

2 of this universe's 18 surviving pairs involve this signal · α vs ^SPXEW.

Europe — 1 surviving pair
Pair (same-day co-fire, long) Full α Test α (2023+) Test N p_perm test
hh_hl_structure bearish + rsi bullish +4.18% — — —

1 of this universe's 20 surviving pairs involves this signal · α vs ^STOXX.

China A-shares — 12 surviving pairs
Pair (same-day co-fire, long) Full α Test α (2023+) Test N p_perm test
hh_hl_structure bearish + weekly_change bearish +3.55% +4.88% 130 0.002
hh_hl_structure bullish + weekly_change bearish +1.97% +4.24% 2,410 0.002
hh_hl_structure bearish + vwap_cross bullish +2.49% +4.07% 969 0.002
hh_hl_structure bearish + macd bullish +1.61% +3.03% 799 0.002
hh_hl_structure bearish + weekly_change bullish +1.95% +2.68% 1,683 0.002
hh_hl_structure bearish + stochastics bearish +1.60% +2.12% 1,705 0.002
hh_hl_structure bearish + vwap_cross bearish +0.73% +1.56% 1,939 0.002
hh_hl_structure bearish + williams_r bearish +0.77% +1.54% 2,711 0.002
hh_hl_structure bullish + williams_r bearish -3.42% -3.45% 134 0.002
bollinger bearish + hh_hl_structure bullish -3.87% -3.56% 87 0.002
hh_hl_structure bullish + weekly_change bullish -4.80% -4.40% 305 0.002
hh_hl_structure bullish + stochastics bearish -4.64% -5.03% 63 0.002

12 of this universe's 138 surviving pairs involve this signal · α vs 83188.HK.

China A-share survivor α runs large but skews toward small-caps, where trading costs and thin liquidity claim a large share of any measured edge — screening context, not a capturable spread.

“—” in the test columns means the held-out 2023+ sample fell below the 20-observation minimum this run requires before it computes any statistic, so no out-of-sample figure exists for that pair — not that it never co-fired again. Those pairs rank last.

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.

  • Use as a regime classifier, not a trigger — HH/HL structure is better thought of as what kind of trend a stock is in rather than whether to buy or short now. Apply as a filter to other signals.
  • Early-trend bullish refinement — The signal underperforms at short horizons but works long-term. Subset to triggers within the first 3 months of trend formation (trendiness-days counter) to capture earlier entries.

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? Pattern signals carry their own structural context; check that the implied support/resistance levels have historical relevance, not just the most-recent 3-month range.
  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

Best use: regime classifier rather than entry trigger. Apply as a filter to other signals (e.g., take bullish RSI only when HH/HL structure is also bullish). Treat the signal as confirmation of trend rather than a timing tool at any horizon. Before leaning on a longer-horizon number, check that horizon's permutation row — a positive 1-year column that fails the random-date test is not edge. The test is one-sided, so read which end of the p range it failed at: near the ceiling means the trigger dates did worse than random dates in the claimed direction, which is inverted rather than merely absent.