Double Top Pattern
A double top is a bearish reversal pattern: two peaks at roughly the same level, separated by a pullback, that completes when price closes below the low between them (the neckline).
Across 111 scored detections, the Double Top moved the way it predicts relative to SPY 56% of the time over the next 20 trading days, with a mean excess return of -0.21%. A coin flip scores about 50%.
23 symbols in our coverage universe have a recorded Double Top detection history. AVGO has formed it most often — 4 times.
How to identify a double top
- A prior uptrend for the pattern to reverse.
- First peak, followed by a pullback.
- Second peak near the level of the first — often within a few percent — frequently on lighter volume.
- Neckline: the low of the pullback between the two peaks.
- Enough time between the peaks. Two highs a few sessions apart are usually just noise, not a double top.
Is a double top bullish or bearish?
Bearish. Price tested the same high twice and could not get through; a close below the neckline confirms that selling at that level won out.
Price target
Take the height from the peaks down to the neckline and project it downward from the breakout point.
When the pattern fails
A close back above the neckline, or a push to a new high above both peaks, invalidates the pattern. A breakout above a failed double top is sometimes read as a bullish signal in its own right.
How has this setup performed historically?
| Window | Occurrences | Win rate vs SPY | Mean excess return |
|---|---|---|---|
| 5 days | 134 | 53% | 0.15% |
| 20 days | 111 | 56% | -0.21% |
| 60 days | 60 | 52% | -6.12% |
20-day figures: 111 settled signals detected from 2026-02-18, scored through 2026-10-29; 24 more still inside the window. Every one is listed in the track record.
Win rate = share of occurrences where the stock beat SPY over the window in the direction the signal predicted — for a bearish setup, a win means it underperformed. Measured on our own detections, which puts the coin-flip line at 50%. Sample sizes are small; treat these as context, not as an edge.
Recent Double Top detections
From our daily published list over the past 12 months, newest first. Past detections, not recommendations — open a symbol for its chart and current status.
| Symbol | Direction | Detected | Detail |
|---|---|---|---|
| WDAY | Bearish | 2026-10-29 | Status & history → |
| MTCH | Bearish | 2026-10-28 | Status & history → |
| NDSN | Bearish | 2026-10-22 | Status & history → |
| DVA | Bearish | 2026-10-20 | Status & history → |
| FTEC | Bearish | 2026-10-16 | Status & history → |
| NVDA | Bearish | 2026-10-16 | Status & history → |
| XLK | Bearish | 2026-10-16 | Status & history → |
| KEYS | Bearish | 2026-10-15 | Status & history → |
| ANET | Bearish | 2026-10-14 | Status & history → |
| DELL | Bearish | 2026-10-14 | Status & history → |
US stocks that have formed the Double Top most often
Counts are past detections by our scanner, not signals firing today. Follow a symbol through to its own page for the current reading.
Frequently asked questions
- What does a double top mean in stocks?
- That a stock rallied to the same price twice and was turned back both times. Supply at that level outlasted demand, and the break of the neckline is read as the trend turning down.
- What is the difference between a double top and a double bottom?
- They are mirror images. A double top has two peaks and is bearish; a double bottom has two lows and is bullish. Each is confirmed by a break of the neckline between them.
- What is the success rate of a double top?
- Across 111 scored detections, the Double Top moved the way it predicts relative to SPY 56% of the time over the next 20 trading days, with a mean excess return of -0.21%. A coin flip scores about 50%.
Further reading
Other patterns
Pattern Vista is a technical-analysis research tool. Everything on this page is automated pattern detection and historical statistics published for educational purposes. It is not investment advice, not a recommendation to buy or sell any security, and past performance does not indicate future results. Full disclaimer