How we tested this pattern
- Sample
- 7,951 occurrences on 6 markets (crypto, forex, US stocks, ETFs, indices and commodities), 1973 to 2026.
- Detection
- Found by code, never picked by hand: every match of the rule on full price history, winners and losers alike.Low, peak, valley, second peak within 10% of the pattern height of the first, with the first low below the valley. Prior rise: the close one pattern duration before the first peak lies at least 60% of the height below the valley. At least 3 ATR tall, at least 3 bars between swing points. Confirmed by a close below the valley within half the pattern duration. The standard grade, which accepts most counted occurrences, widens this: levels within 15%, a prior move of at least 30% of the height over the two pattern durations before, a breakout within 0.75 of the duration. The first peak is a true high: from one first-leg duration before the pattern up to that peak, price never rose more than 10% of the pattern height above it.
- Benchmark
- Each occurrence is compared with the same assets on random bars: the move 10 bars later, and a 2:1 reward-to-risk trade.
- Checks
- Block bootstrap and a correction for the thousands of tests we run, plus a random-walk self-test that the bench passes: on random prices, no pattern should win.
- Updated
- Data updated November 5, 2026, code version c2a1683f. Read the methodology · Raw data (JSON)
Watch the video

At a glance: what textbooks say, what we measured
- Direction
- Bearish
- Occurrences tested
- 7,951 occurrences, all markets, daily
- Textbook: price should fall after it
- Too close to call · −0.04% vs the market, 10 bars
- Textbook: target reached most of the time
- 45.0% vs 44.8% on random dates
- Where it did best
- Commodities, daily: +0.34 R 2:1 trade vs random dates, not proven
- Verdict
- 10-bar move: No clear edge2:1 trade: No clear edge
On this page
- What is a double top pattern?
- Anatomy of a double top: identifying the key components
- How to identify a double top pattern on your chart
- The psychology behind the double top formation
- Real examples
- How to trade the double top pattern effectively
- Confirming the pattern with volume and technical indicators
- Common mistakes to avoid when trading double tops
- Double top vs. double bottom: key differences
- Comparing the double top to other reversal patterns
- Are double top patterns reliable?
- Best timeframes and markets for trading double tops
- Double top case studies
- Failed double top patterns: how to spot and handle them
- Pros and cons of trading the double top
- Frequently asked questions about the double top pattern
- Summary of key points
- Final thoughts on risk management
The double top pattern is a bearish reversal chart pattern: price makes a high, pulls back to a low point, rallies to about the same high a second time, then breaks below that low point. On a chart it looks like the letter M. Classical chart analysis reads it as the end of an uptrend: buyers failed twice at the same resistance level, and the break of the valley between the two peaks confirms that sellers are in control.
This guide covers the definition, how to identify a double top on a chart, how traders set the entry, stop and target, and the question that decides everything else: what actually happened after the break. Our test bench scanned full price history with a fixed rule and detected 7,951 on daily charts alone across stocks, crypto, forex, ETFs, indices and commodities, and compared the outcome with random bars on the same assets.
What is a double top pattern?
A double top is a reversal pattern made of two distinct peaks at roughly the same price level, separated by a trough, that forms after a rise. It is one of the classic chart patterns of technical analysis, described by Robert Edwards and John Magee in Technical Analysis of Stock Trends (1948) among the major reversal formations. Its mirror image, after a decline, is the double bottom.
The key point of the definition: two peaks at the same level are not yet a double top. The pattern only exists once price closes below the low between the two peaks. Until then, it is a market testing a resistance level, and it can just as well break higher.
Our bench uses a rule a computer can apply without judgment: a rise into the formation (the low before the first peak is below the valley), a first peak that is a real local high (over one first-leg duration before the pattern, price never rose more than 10% of the pattern height above it), a valley, a second peak within 10% of the pattern height of the first one (15% on the looser standard grade), and a close below the valley. The peaks and troughs are swing points measured in multiples of the average true range (ATR), at three scales, so the same rule finds small and large double tops.
The chart above is not drawn by hand. It is the median shape of the double tops our bench detected, with the position we test: entry at the close that breaks the valley, stop loss at the higher of the two peaks (the trade ends as soon as price trades above it), target at twice the risk.
Anatomy of a double top: identifying the key components
A double top has four parts: the uptrend before it, the first peak, the neckline, and the second peak. Each one has a job in the story.
- Low before the risebelow the valley: price rises into the pattern
- First peak
- Valley between the peaksthe neckline
- Second peakwithin 10% of the pattern height of the first
- Close below the valley
The first peak and the support level
The first peak forms at the top of an uptrend. Buyers push price to a new high, then selling pressure appears and price pulls back. At this stage nothing looks unusual: trending markets pause all the time. The pullback ends at a trough, and that low point becomes the support level that will matter later.
The neckline and the second peak
The neckline is the horizontal line drawn through the trough between the two peaks. From there, buyers try again and push price back to the area of the first high. The second peak is a near copy of the first: it stalls at about the same level, sometimes a little below, sometimes a little above. When price then falls back and closes below the neckline, the double top is complete and the support level becomes resistance.
How equal must the two highs be? There is no universal number. We accept a second peak within 15% of the height of the pattern (the distance from the higher peak to the neckline), and a textbook pattern stays within 10%. On our daily data, the typical pattern is not exact to the cent: a small difference between the two highs is normal.
How to identify a double top pattern on your chart
To identify a double top, check five points in order.
- A prior uptrend. There must be a rise to reverse. Two equal highs inside a sideways range are a range, not a top.
- A first peak. A clear swing high, followed by a pullback of meaningful size relative to the usual volatility of the asset.
- A trough. Mark its low: this is the neckline, the trigger level of the pattern.
- A second peak near the first. Price returns to the same resistance and fails to break it with conviction.
- A close below the neckline. The pattern is only confirmed by a close, not by an intraday dip.
That last step filters out most candidates. On daily charts, only 39.1% of the double tops our bench saw forming (two peaks and a valley) were confirmed by a close below the valley. The others broke above the peaks first or never broke at all. Counting an unconfirmed double top as a sell signal means trading a pattern that, more often than not, does not complete.
If the second high is lower than the first and the lows stay flat, the shape is closer to a descending triangle than to a double top.
The psychology behind the double top formation
In classical chart analysis, the double top tells a story of exhaustion. On the first peak, buyers who drove the uptrend meet sellers at a price level: some traders take profit, others open short trades, and price falls back. On the pullback, buyers who missed the first move see a second chance and push price up again. When price returns to the first high, the same sellers are waiting. The failure to make a higher high is the first sign that demand is fading.
The break of the neckline adds the last element: everyone who bought during the formation is now at a loss, and some of them sell, which can feed the decline. That is the theory. Whether it shows up in prices more often than chance is a question for data, and we answer it in the reliability section.
Real examples
Real occurrences from our test bench, each on its market's main timeframe. Crypto and US stock examples come from liquid assets only (the largest coins, the most traded stocks). Best and Worst are the best and worst 2:1 trade results, Textbook the cleanest shapes (chosen on geometry alone, never on the outcome), Most recent the latest detections among the assets the gallery draws from, Typical a random draw. Shaded area: the pattern. Dashed lines: entry, stop loss (SL) and 2:1 target (TP) of the tested trade. Dotted line: 10 bars after the signal. A result above +2 R or below −1 R means a bar opened past the target or the stop, and the trade was closed at that open. Stock prices are adjusted for splits and dividends, so they can differ from the prices quoted at the time. We left out 3 cards: 2 with a shape that does not read as the definition on this page, 1 with a single bar so large that the pattern or the trade levels could not be read. They still count in the statistics, and are not replaced: our bench picks a fixed number of cards per category.
How to trade the double top pattern effectively
In trading terms, the double top gives three levels: an entry, a stop and a target. The classic way to trade a double top is to sell short on the break of the neckline, with a stop loss above the peaks and a profit target measured from the pattern. This is similar to the trade our bench simulates on every occurrence.
- Entry
- Close below the valley between the peaks
- Stop loss (SL)
- Above the higher of the two peaks
- Target (TP)
- Twice the risk (2:1)
A clean case from our bench: US stocksChosen on its shape alone, never on the outcome. Stock prices are adjusted for splits and dividends, so they can differ from the prices quoted at the time.
Entry strategy: breakout vs. retest
There are two main entries. The breakout entry sells at the close of the bar that breaks the neckline: it is the entry we test. The retest entry waits for price to rally back to the broken neckline, which now acts as resistance, and sells there. The retest gives a better price and a tighter stop, but price does not always come back, and you can miss the move. A third, aggressive entry sells at the second peak itself, before any confirmation: it gives the best price and the highest failure rate, since most double tops never confirm.
Setting your stop loss and take profit targets
The classic target is the measured move: measure the height from the peaks to the neckline, then project that distance down from the break. If the peaks are at 110 and the neckline at 100, the target is 90. A prior support level above that target is another place to take part of the profit. Our statistics give the hit rate of the 2:1 target only: the measured move is not drawn on our chart, and the timeline gives only its median delay when reached, not how often it was reached.
Two timing figures help plan the trade. On daily charts, trades that reached the 2:1 target took a median of 41 bars; trades that hit the stop took a median of 32 bars. And of the trades that hit their stop, 9.8% had first been up 1 R. Decide in advance what you do at +1 R.
Confirming the pattern with volume and technical indicators
The close below the neckline is the only confirmation our rule uses. Two other families of filters can be added to it: volume and technical indicators.
Volume
The textbook reading wants lower volume on the second peak than on the first (weaker buying), and higher volume on the break. We measured the second part on daily charts: breaks on volume above 1.5 times the 20-bar average were followed by a lower price 10 bars later 45.8% of the time, against 44.1% for breaks on normal volume, and the 2:1 trade reached its target 11.5% of the time against 12.9%. Heavy volume on the break changed little. This breakdown is descriptive (not corrected for the number of tests we run), so read it as a warning against relying on volume, not as a new rule.
Momentum indicators and moving averages
Technical indicators are the second family of trading filters. One filter looks for a bearish divergence on the RSI or the MACD: the second peak at the same price as the first, but with weaker momentum. We have not tested divergence filters: if you use one, test it on your own market first. We did test the 200-day moving average. Double tops above it (a top inside a long-term uptrend) were lower 10 days later 44.5% of the time, with an expectancy of −0.25 R; those below it, 45.3% and −0.19 R. Neither group beat chance.
Common mistakes to avoid when trading double tops
A few repeated errors make the losses larger.
Selling before the neckline breaks
Ignoring the trend before it
A double top needs an uptrend to reverse. Equal highs in a range only describe the top of the range.
Choosing a target the pattern cannot reach
The taller the pattern, the further away the stop, and the further away a 2:1 target. On daily charts, tall double tops (above the median height) reached their 2:1 target 6.8% of the time, against 17.5% for short ones. The expectancy was about the same (−0.21 R and −0.23 R), but a tall pattern needs more time and more patience.
Trusting hindsight
On a past chart, the double tops that worked are easy to see and the failed ones are easy to forget. Count every pattern, winners and losers, before you decide the setup works on your market. This is true of all chart patterns, and of every trading strategy built on them.
Double top vs. double bottom: key differences
The double bottom pattern is the mirror image of the double top: two troughs at the same level after a decline, confirmed by a close above the peak between them. The double top is bearish and is traded short; the double bottom is bullish and is traded long.
| Feature | Double top | Double bottom |
|---|---|---|
| Shape | M, two peaks | W, two troughs |
| Prior trend | Uptrend | Downtrend |
| Confirmation | Close below the valley | Close above the peak between the troughs |
| Trade | Short, stop above the peaks | Long, stop below the troughs |
| Daily closed trades | 7,531 | 6,303 |
| 2:1 target hit (random dates) | 12.4% (12.3%) | 21.0% (22.2%) |
| Expectancy before costs | −0.22 R | +0.20 R |
The double bottom looks much better, but the same trade on random dates did almost as well (+0.19 R). The difference comes from the market, not the pattern: the assets we test rose more often than they fell over their history, which helps long trades and hurts short ones, as the random-date figures in the table show. Neither pattern's average result per trade beat random dates on daily charts by a margin large enough to separate from chance, and neither hit its target clearly more often than on random dates.
Comparing the double top to other reversal patterns
Double top vs. head and shoulders
The head and shoulders has three peaks, with a higher one in the middle; the double top has two at the same level. Both are confirmed by a close below a neckline. On daily charts, the head and shoulders trade reached its 2:1 target 15.9% of the time against 17.5% on random dates, at −0.26 R before costs, an expectancy slightly below the −0.22 R of random dates, a gap too small to pass our tests: no better than the double top. On direction, its daily row is even rated worse than chance.
Double top vs. triple top
The triple top adds a third test of the same resistance. Our bench checks for it first: if three peaks are within 10% of the height of each other (up to 25% on the looser standard grade), it is a triple top, not a double top. On daily charts the triple top trade reached its 2:1 target 9.4% of the time against 11.0% on random dates, at −0.26 R. A third test of the high did not make the reversal more reliable.
Two other tops are often confused with it. The diamond pattern widens, then narrows before the break, with no clear twin peaks. And after a sharp fall, the rebound that fails below the old support is called a dead cat bounce: a weak rally inside a downtrend, not a second test of the high.
Are double top patterns reliable?
Not on their own. On almost every market and timeframe we tested, the double top's average excess return over the rest of the market was too small to separate from chance: it was followed by a decline about as often as a random bar on most rows, and slightly less often on hourly and 4-hour charts. Daily indices are the exception, in the wrong direction: they were lower 10 bars after the break more often than random bars, yet their average move went against the pattern by a margin that passes a simple test but not our correction for multiple testing, rated weak, below chance. Here are the numbers from our test bench, which scans full price history and compares each occurrence with random bars on the same assets and with the rest of the market.
Moved the expected way 10 bars later, compared with random bars.
Measured on 7,951 daily double top occurrences across 1,715 assets. Random bars on the same assets are the honest baseline.
No clear edgeDouble top against random bars on the same assets, by market, on daily candles. Bars of almost the same length mean no edge.
- All markets
- Crypto
- US stocks
- Forex
- ETFs
- Indices
- Commodities
- All markets
- Crypto
- US stocks
- Forex
- ETFs
- Indices
- Commodities
All statistics
Full results for the double top pattern, market by market
Every occurrence our rule finds on full price history, compared with what the same assets did on random bars. A pattern only earns an edge if it beats random entries after correcting for the thousands of tests we run.
Our verdict
Not a reliable signal on its own here. Use it as context, not as a trigger.
Double top: 7,951 occurrences tested
Bearish pattern · All markets · Daily · 1,715 assets · 1973-04 to 2026-10
What this means. No measurable difference from the rest of its market over the next 10 bars.
Timeline
- Pattern length, first swing to breakout
- 64 days · middle half 38 to 118 · on 7,951 occurrences
- Breakout, after the last swing
- 3 days · middle half 1 to 8 · on 7,951 occurrences
- Time to the 2:1 target, when hit
- 41 days · middle half 21 to 64 · on 935 trades that hit it
- Time to the stop, when hit
- 32 days · middle half 17 to 56 · on 4,041 trades that hit it
- Time in the trade, all exits
- 59 days · middle half 25 to 100 · on 7,531 trades
- Measured-move target reached, when hit
- 16 days · middle half 7 to 34 · on the occurrences that reached it
- Pattern height projected from the breakout
Size and trade path
- Pattern height
- 11.8% · middle half 7.8% to 18.4% · on 7,951 occurrences
- Best open profit during the trade
- 0.60 R · middle half 0.23 R to 1.26 R · on 7,531 trades
- Stopped trades that were first up 1 R
- 9.8% · on 4,041 stopped trades
- Trades a stop moved to breakeven at +1 R would have closed flat instead of at a loss
Medians, each on the cases named on its line; "middle half" spans the 25th to 75th percentile; no value under 30 cases.
How we detect it. Low, peak, valley, second peak within 10% of the pattern height of the first, with the first low below the valley. Prior rise: the close one pattern duration before the first peak lies at least 60% of the height below the valley. At least 3 ATR tall, at least 3 bars between swing points. Confirmed by a close below the valley within half the pattern duration. The standard grade, which accepts most counted occurrences, widens this: levels within 15%, a prior move of at least 30% of the height over the two pattern durations before, a breakout within 0.75 of the duration. The first peak is a true high: from one first-leg duration before the pattern up to that peak, price never rose more than 10% of the pattern height above it.
All measurements (20 rows)
10-bar move: did price go the expected way?
| Market | Occurrences | Pattern | Random bars | Excess | 10-bar move verdict |
|---|---|---|---|---|---|
| Crypto · 1-hour | 8,105 | 48.4% | 50.6% | −0.01% | No clear edge |
| Forex · 1-hour | 1,918 | 47.1% | 49.2% | −0.002% | No clear edge |
| All markets · 1-hour | 10,023 | 48.1% | 50.4% | −0.01% | No clear edge |
| Crypto · 4-hour | 1,584 | 49.4% | 51.9% | +0.05% | No clear edge |
| Forex · 4-hour | 418 | 47.8% | 49.3% | +0.07% | No clear edge |
| All markets · 4-hour | 2,002 | 49.1% | 51.3% | +0.05% | No clear edge |
| Crypto · Daily | 162 | 43.8% | 54.4% | +0.02% | No clear edge |
| US stocks · Daily | 7,147 | 44.7% | 45.4% | −0.04% | No clear edge |
| ETFs · Daily | 317 | 44.6% | 42.8% | +0.05% | No clear edge |
| Indices · Daily | 196 | 49.0% | 42.4% | −0.45% | Weak, below chance |
| Forex · Daily | 73 | 47.9% | 48.9% | +0.14% | No clear edge |
| Commodities · Daily | 56 | 53.6% | 47.1% | +0.02% | No clear edge |
| All markets · Daily | 7,951 | 44.9% | 45.5% | −0.04% | No clear edge |
| Crypto · Weekly | 5 | 60.0% | 53.5% | +6.95% | Not enough data |
| US stocks · Weekly | 1,168 | 38.8% | 42.6% | −0.08% | No clear edge |
| ETFs · Weekly | 38 | 18.4% | 38.3% | −1.08% | No clear edge |
| Indices · Weekly | 20 | 20.0% | 38.7% | −1.76% | Not enough data |
| Forex · Weekly | 6 | 33.3% | 48.9% | −1.07% | Not enough data |
| Commodities · Weekly | 12 | 50.0% | 46.6% | +1.27% | Not enough data |
| All markets · Weekly | 1,249 | 38.0% | 42.5% | −0.10% | No clear edge |
2:1 trade: target hit and average result
| Market | Trades | Target hit | Same trade, random dates | Per trade | 2:1 trade verdict |
|---|---|---|---|---|---|
| Crypto · 1-hour | 8,038 | 13.2% | 15.1% | +0.01 R vs −0.01 R | No clear edge |
| Forex · 1-hour | 1,911 | 14.4% | 16.9% | −0.03 R vs −0.01 R | No clear edge |
| All markets · 1-hour | 9,949 | 13.5% | 15.4% | 0.00 R vs −0.01 R | No clear edge |
| Crypto · 4-hour | 1,450 | 13.0% | 13.5% | +0.11 R vs +0.01 R | No clear edge |
| Forex · 4-hour | 413 | 14.5% | 16.6% | −0.01 R vs −0.01 R | No clear edge |
| All markets · 4-hour | 1,863 | 13.3% | 14.2% | +0.09 R vs 0.00 R | No clear edge |
| Crypto · Daily | 70 | 10.0% | 9.6% | −0.10 R vs +0.06 R | No clear edge |
| US stocks · Daily | 6,826 | 11.8% | 11.9% | −0.24 R vs −0.25 R | No clear edge |
| ETFs · Daily | 314 | 20.4% | 17.6% | −0.14 R vs −0.24 R | No clear edge |
| Indices · Daily | 193 | 20.2% | 18.4% | −0.10 R vs −0.22 R | No clear edge |
| Forex · Daily | 73 | 17.8% | 14.0% | +0.18 R vs +0.01 R | No clear edge |
| Commodities · Daily | 55 | 12.7% | 8.2% | +0.07 R vs −0.27 R | No clear edge |
| All markets · Daily | 7,531 | 12.4% | 12.3% | −0.22 R vs −0.25 R | No clear edge |
| Crypto · Weekly | 0 | n/a | n/a | n/a vs n/a | Not enough data |
| US stocks · Weekly | 808 | 10.1% | 8.4% | −0.44 R vs −0.42 R | No clear edge |
| ETFs · Weekly | 32 | 9.4% | 10.9% | −0.71 R vs −0.53 R | No clear edge |
| Indices · Weekly | 19 | 5.3% | 9.2% | −0.76 R vs −0.50 R | Not enough data |
| Forex · Weekly | 6 | 16.7% | 8.3% | 0.00 R vs −0.14 R | Not enough data |
| Commodities · Weekly | 10 | 0.0% | 2.5% | −0.45 R vs −0.41 R | Not enough data |
| All markets · Weekly | 875 | 9.9% | 8.4% | −0.45 R vs −0.42 R | No clear edge |
After the breakout
| Market | Occurrences | False starts | Same, random dates | Throwbacks |
|---|---|---|---|---|
| Crypto · 1-hour | 8,105 | 28.8% | 29.5% | 70.7% |
| Forex · 1-hour | 1,918 | 31.1% | 29.3% | 71.8% |
| All markets · 1-hour | 10,023 | 29.2% | 29.5% | 70.9% |
| Crypto · 4-hour | 1,584 | 27.4% | 29.9% | 70.8% |
| Forex · 4-hour | 418 | 30.6% | 30.0% | 74.0% |
| All markets · 4-hour | 2,002 | 28.1% | 29.9% | 71.5% |
| Crypto · Daily | 162 | 27.2% | 32.6% | 70.3% |
| US stocks · Daily | 7,147 | 37.0% | 36.9% | 71.8% |
| ETFs · Daily | 317 | 34.7% | 36.7% | 74.0% |
| Indices · Daily | 196 | 34.0% | 34.6% | 74.5% |
| Forex · Daily | 73 | 25.0% | 21.2% | 71.2% |
| Commodities · Daily | 56 | 31.5% | 34.5% | 74.5% |
| All markets · Daily | 7,951 | 36.5% | 36.6% | 71.9% |
| Crypto · Weekly | 5 | 0.0% | 68.3% | 60.0% |
| US stocks · Weekly | 1,168 | 41.6% | 42.5% | 68.5% |
| ETFs · Weekly | 38 | 57.7% | 55.1% | 59.4% |
| Indices · Weekly | 20 | 35.3% | 48.3% | 78.9% |
| Forex · Weekly | 6 | 16.7% | 25.0% | 66.7% |
| Commodities · Weekly | 12 | 50.0% | 26.5% | 75.0% |
| All markets · Weekly | 1,249 | 41.7% | 42.9% | 68.4% |
Data updated November 5, 2026 · code version c2a1683f · Read the methodology. We measure every pattern on crypto and forex (1-hour, 4-hour, daily and weekly charts) and US stocks, ETFs, indices and commodities (daily and weekly charts). A market or timeframe from that list missing from the tables above had no occurrence.
What you often read
“The double top pattern is a reliable bearish signal.”
What our data shows
On 7,951 daily occurrences, it moved in the expected direction 44.9% of the time, against 45.5% for random bars: no clear edge, since the average excess return is not statistically significant.
Table quoted in the text (verdicts and every market: see All measurements)
| Market and timeframe | Double tops tested (2:1 trades) | Lower 10 bars later | Random bars | 2:1 target hit | Same trade, random dates | Expectancy |
|---|---|---|---|---|---|---|
| All markets, daily | 7,951 (7,531) | 44.9% | 45.5% | 12.4% | 12.3% | −0.22 R |
| US stocks, daily | 7,147 (6,826) | 44.7% | 45.4% | 11.8% | 11.9% | −0.24 R |
| ETFs, daily | 317 (314) | 44.6% | 42.8% | 20.4% | 17.6% | −0.14 R |
| Indices, daily | 196 (193) | 49.0% | 42.4% | 20.2% | 18.4% | −0.10 R |
| Commodities, daily | 56 (55) | 53.6% | 47.1% | 12.7% | 8.2% | +0.07 R |
| Crypto, daily | 162 (70) | 43.8% | 54.4% | 10.0% | 9.6% | −0.10 R |
| Crypto, 4-hour | 1,584 (1,450) | 49.4% | 51.9% | 13.0% | 13.5% | +0.11 R |
| Crypto, hourly | 8,105 (8,038) | 48.4% | 50.6% | 13.2% | 15.1% | +0.01 R |
| Forex, hourly | 1,918 (1,911) | 47.1% | 49.2% | 14.4% | 16.9% | −0.03 R |
| All markets, weekly | 1,249 (875) | 38.0% | 42.5% | 9.9% | 8.4% | −0.45 R |
Read the table in pairs of columns. On daily charts, 44.9% of double tops were followed by a lower price 10 bars after the break. Random bars on the same assets were lower 45.5% of the time. The excess return over the rest of the market after 10 bars is −0.04%, with a 95% confidence interval from −0.23% to +0.14%. Zero is inside the interval: no clear edge.
The trade columns tell the same story. On daily charts the short trade reached its 2:1 target 12.4% of the time, against 12.3% for the same entry, stop and target on random dates, with an expectancy of −0.22 R per trade before costs (R is the amount risked on the trade). The random version lost too (−0.25 R): shorting markets that drift up is expensive. With a 2:1 target, a trader needs to win more than one trade in three to break even; the daily double top reached its target in only 12.4% of its trades. Fees and slippage would make it worse.
What changes the result, and what does not
The split by period shows no stable effect: the excess return was +0.14% in the first half of the history and −0.22% in the second. A few small samples show higher raw figures, such as weekly commodities (12 patterns, 50.0% lower against 46.6% for random bars), but with so few cases that row is rated not enough data: its numbers are not interpreted. With enough tests, some results look good by luck alone.
How we tested the double top
The detection rule is fixed in code and published, so the test can be checked and repeated. Each double top is entered at the close of the bar that breaks the valley, with the stop at the higher peak and the target at twice the risk. The trade exits at the stop, the target, or the close 100 bars later. When a bar opens beyond the stop or the target, the exit is at that open. There are no fees or slippage, and when one bar touches both the stop and the target we count the loss. Each trade is then replayed on 4 random dates of the same asset, with the same distances in percent of price: that matched baseline is the "same trade, random dates" column. These choices are one reasonable reading of the pattern, not the only one, and real results after costs would be somewhat worse.
Best timeframes and markets for trading double tops
Technical analysis applies the same way to stocks, crypto and forex, and so does our test. No timeframe and no market stood out in the pattern's favor. Its frequency does vary: one occurrence every 1,066 eligible bars on daily charts overall and one every 1,681 on hourly charts, from one every 705 on daily indices to one every 3,716 on daily crypto. On a watchlist of a few dozen stocks, signals come regularly, but not every week.
Forex had the best raw expectancy on daily charts (+0.18 R on 73 trades), against +0.01 R for random dates on the same pairs, but the gap is not significant: the row reads No clear edge. Among the hourly and 4-hour samples, forex had the highest target hit rates (14.4% hourly, 14.5% on 4-hour charts), with random dates at 16.9% and 16.6%; weekly forex shows 16.7%, but on 6 trades only, too few to be rated. Weekly charts were the worst (9.9% of targets hit, −0.45 R), because the stop above a weekly peak is far away and the target further still.
Double top case studies
Here are three real double tops our bench found on daily crypto charts, told in detail with their actual outcome. The Real examples section shows other cases.
A clean double top: TRX, August 2025
TRX/USDT made a first peak at 0.37 on August 14, 2025, fell to 0.3444 on August 18, then rallied to 0.3687 on August 24, just under the first high. On August 25 price closed at 0.3417, below the neckline: entry. The stop sat at 0.37 and the 2:1 target at 0.2851. TRX was 2.1% lower 10 bars after the break, and the target was hit 70 days after entry.
A failed double top: XLM, March 2023
XLM/USDT peaked at 0.0962 on January 24, 2023 and at 0.0981 on February 21, with a trough at 0.0817 between them on February 13. On March 8 it closed at 0.0816, below the neckline: a clean signal. Price first fell to 0.0744 on March 10, then turned, and on March 21 it traded above the stop at 0.0981, a full 1 R loss. Ten bars after the break, XLM was 4.9% higher. A confirmed double top does not protect against a market that turns.
A target too far: TRX, July 2019
TRX/USDT formed a tall double top: peaks at 0.04095 on June 2 and 0.04048 on June 25, 2019, with the lowest point between them at 0.029 on June 9. Our bench used the later, higher swing low as its neckline (0.03208 on June 18), but it only reads the second peak as a swing once price has fallen far enough from it, which happened on July 11. The June 30 close at 0.032, already under that line, came before, so the bench entered late, at the July 11 close of 0.02792. With the stop at the higher peak, the 2:1 target was 0.00186, a price TRX would have to lose more than 90% to reach. Price did fall, to 0.0112 by September 24, but our test closed the trade after 100 bars at +0.95 R. A tall pattern with a late entry gives a target that is out of reach: plan a realistic exit before entering.
Failed double top patterns: how to spot and handle them
A failed double top is one that confirms, then reverses: price breaks the neckline, then rallies back above it and through the peaks. The XLM case, under A failed double top: XLM, March 2023, is one. Classical analysis watches three signs, which we have not tested as filters:
- price closes back above the neckline within a few bars of the break;
- the market rallies above the higher peak, which invalidates the pattern and triggers the stop;
- the larger trend is still strongly up, on a higher timeframe.
How to handle it: respect the stop. Another option is to reverse and buy when price closes above the peaks, treating the failed top as a breakout; we have not tested that trade for the double top. What our data does show is that failure is common: on daily charts, only 12.4% of trades reached their 2:1 target.
Pros and cons of trading the double top
Advantages
- Easy to identify: two peaks, one valley, one trigger level.
- Defined risk: the stop above the peaks is visible before entry.
- Available on every market and timeframe.
Limitations
- On its own, it did not beat chance on any market or timeframe we tested.
- Fewer than half of the formations confirm with a close below the neckline.
- Tall patterns put the stop and the target far from the entry.
- Short trades work against the long-term upward drift of the assets we test.
Frequently asked questions about the double top pattern
Is the double top pattern bullish or bearish?
The double top is a bearish reversal pattern: it marks the potential end of an uptrend. In our data, it was followed by a decline about as often as a random bar on the same assets.
How do you confirm a double top breakout?
With a close below the neckline, the low between the two peaks. An intraday dip below it is not enough. A stricter version also waits for a retest of the broken neckline from below.
Can I use the double top pattern for crypto and forex?
Yes, the definition is the same on any market. On hourly forex (1,911 trades on 25 currency pairs since 2012), the 2:1 trade reached its target 14.4% of the time against 16.9% on random dates. On daily crypto (162 occurrences on 124 coins since 2019, 70 closed trades), 10.0% against 9.6%. Neither hit rate is clearly above random dates, and neither average result per trade differs from random dates by a margin that passes our statistical test.
Is the double top pattern profitable?
Not on its own in our tests. Before fees, the 2:1 short trade lost money on most markets and timeframes: −0.22 R per trade on daily charts and −0.45 R on weekly charts, all markets together. The few positive rows, such as daily forex at +0.18 R against +0.01 R on random dates, are gaps that are not significant, and no row beat chance.
What is the difference between a double top and an M pattern?
None: "M pattern" is a nickname for the double top, from its shape. The double bottom is called a W pattern.
Summary of key points
- A double top is two peaks at about the same level after an uptrend, confirmed by a close below the valley between them.
- On daily charts, only 39.1% of the formations confirmed.
- On 7,951 confirmed daily double tops, 44.9% were lower 10 bars later, against 45.5% for random bars.
- The 2:1 short trade reached its target 12.4% of the time against 12.3% on random dates, at −0.22 R before costs.
- Volume on the break and the 200-day moving average did not change the picture.
Final thoughts on risk management
If you trade the double top, treat it as a place to define risk, not as a reason to sell. Wait for the close below the neckline, put the stop above the peaks, size the position on that distance and keep the risk per trade small. Our data shows that the pattern alone does not tilt the odds; any edge has to come from what you combine it with, and you should test that combination on your own market before risking money. Nothing on this page is investment advice: trading involves the risk of losing money, and past results of a technical pattern do not predict future ones.
About the author and the data. Written by Amaury, founder of PatternsWizard, who is passionate about finance, quantitative trading and data analysis. He built the PatternsWizard test bench, which detects each setup with fixed, published rules on full market history: crypto, forex, US stocks, ETFs, indices and commodities. Trades and win rates are compared with random bars on the same assets, excess returns with the rest of the market, with error margins and a correction for multiple testing, and the whole pipeline is checked on random-walk prices, where nothing can be predicted. Statistics on this page were generated on November 5, 2026. The full methodology explains every step.
