QWE AI Academy Confirmed double top: neckline close, height-based target, stop above the highs, and hard filters on trough depth and peak spacing. That’s the whole trade plan. Everything below is how you mark it without guessing shapes.
Two peaks near the same level after an uptrend, trough in between, then a close under that trough. Until the close, you only have twin peaks – not a tradable reverse.
Quick context: what you’re actually detecting
Buyers print a high, get rejected, pull back, try again, fail near the same resistance. Supply wins. The pattern becomes a double top only when price closes below the intervening low.
Think of it like two failed attempts to climb the same wall. The break of the valley floor between the attempts is the piece of data that says the crowd flipped.
Hands-on: reverse-engineer a tradeable double top pattern
Start from the finished plan, then work backward to the chart. That order stops early entries.
1. Pre-define your rules before you look
- Prior trend: Clear uptrend (higher highs/lows) into the pattern – multi-week when you can get it on your timeframe.
- Peak tolerance: Second high within roughly 3% of the first (StockCharts and Bulkowski identification guidelines). Exact equality is rare.
- Valley depth: Prefer at least ~10% between peaks (StockCharts ChartSchool often cites a 10-20% range). Shallower dips are weak evidence of real selling pressure.
- Separation: Weeks apart on dailies is normal; a few candles is usually ordinary resistance, not a regime shift.
- Confirmation: Close below the lowest point between the peaks (neckline). Optional noise filters from ChartSchool: ~3% penetration or a multi-day hold below.
- Volume clue: Often lighter on the second peak push, heavier on the breakdown.
Write the numbers down. Subjectivity kills edge; fixed thresholds shrink it.
2. Mark the geometry on the chart
Peak 1 (trend high so far), trough after it, peak 2 near the same price, horizontal neckline on the trough low. Still no trade.
// Simple height & target calculation (prices in your chart units)
peak = max(peak1, peak2)
neckline = trough_low
height = peak - neckline
target = neckline - height // classic measure rule
// Optional: scale by historical hit-rate from your own sample
// conservative_target = neckline - (height * hit_rate)
Target = pattern height projected down from the break. Same measure rule you’ll see on Investopedia and StockCharts ChartSchool. Large pattern, large objective – if the break holds.
3. Enter only on confirmation (two clean ways)
- Break entry: Short/sell on a close below the neckline, preferably with volume expansion. Stop a few ticks/percent above the higher peak.
- Retest entry: Break first, then wait for a bounce into the neckline from below (now resistance). Enter on rejection. Same stop logic, often cleaner risk.
Pro tip: Second peak on clearly higher volume than the first? Treat it as suspect. The usual tell is waning participation on the second try.
Stop goes above the pattern high so a push through invalidates the thesis cleanly. Size from stop distance – not from how pretty the M looks.
The catch is timing. Most damage happens when people short the second peak and hope.
Common pitfalls that turn the pattern into noise
Entering before the neckline close is the expensive one. In Bulkowski’s twin-peak work (historical samples on thepatternsite.com), unconfirmed tops keep rising roughly 60% of the time – you’re fighting the prior trend with zero proof it reversed.
Tight stops inside normal pullback range are the other silent killer. The break can look clean, then price tags the old neckline and shakes out anyone who placed risk a few ticks under entry. Plan the retest or give the stop room above the highs.
Performance numbers that actually matter
Turns out confirmation changes the base rates a lot. Thomas Bulkowski’s published samples for Adam & Adam double tops (sharp peaks, confirmed, bull-market study context; figures may differ by era and market) show about a 25% break-even failure rate, ~15% average decline, ~64% pullback rate, and ~64% meeting the full measure-rule target across 1,114 perfect trades. Eve & Eve (rounded peaks) in his tables: break-even failure near 20%, average decline around 16%, pullback near 65% (942 trades).
| Metric (confirmed, bull-market samples) | Adam & Adam | Eve & Eve (approx.) |
|---|---|---|
| Break-even failure | ~25% | ~20% |
| Average decline | ~15% | ~16% |
| Pullback rate | ~64% | ~65% |
| Meets full target | ~64% | – (not stated in cited facts) |
Not a forecast for your next ticker. Read them as base rates: confirmation beats raw twin peaks, pullbacks are common enough that break-only entries need a plan, and full height is a useful objective – not an ATM.
When NOT to use this
No clear prior uptrend? Pass. Peaks stacked a few bars apart on the daily? Pass – that’s congestion. Tick charts where every wiggle draws an M? Pass.
After a long multi-month decline, a fresh “double top” confirmation can mark a late stage of that drop rather than the open of a new major leg. Bulkowski’s Adam & Adam notes flag that context; further downside is often limited, so sizing a huge swing short off the pattern alone is the wrong read.
Raging higher-timeframe bulls with no other weakness are hostile terrain for pure pattern shorts. Wait for structure cracks or divergent momentum if you still want the side.
Tools that fit next to this: plain support/resistance, volume bars (or profile), RSI/MACD lower-high divergence against equal price highs. Separate skills – lock the geometry and confirmation rules first.
FAQ
Is a double top pattern bullish or bearish?
Bearish – once the neckline breaks. Double bottom is the mirror after a downtrend.
How close do the two peaks need to be?
Guidelines commonly allow roughly 3% variation. They should fail in the same resistance zone. Perfect twins are optional; a second peak that holds meaningfully higher is often continuation, not this pattern.
Should I enter on the neckline break or wait for a retest?
Break entry is earlier. It also eats more noise because pullbacks to the broken neckline show up in a large share of Bulkowski’s confirmed samples. Retest entry waits for that return and a rejection or lower high – tighter location, fewer runaway fills. One practical hybrid: starter size on the break, add only on a clean failed retest. Pick a rule, log it on your market, stop mixing styles mid-trade.
Open a liquid daily chart and mark every twin-peak structure from the last two years. Score each against your written trough, peak-tolerance, separation, and neckline-close rules. Note which hit the height target and which tagged the neckline again first. That drill beats another page of generic M diagrams.