When does the double top work best?

In Fred's tests the best version was on EURUSD, on the 15-minute chart, in the morning (7am–noon UK), after price has been rising. It averaged +15% of the risk per trade over 392 trades, and made money in both 2020–23 and 2024–26.

Tested patterns · Sell setup

Does the double top work? Tested on 96,630 trades

Q: Does the double top work?

Not on its own, but 2 filtered versions did. Across 96,630 double top trades on 8 markets from 2020 to 2026, after real broker spreads, the setup averaged −7% of the risk per trade (−£0.72 per £10 risked) and reached a 1.5× take profit 22.5% of the time.

Not on its own, but 2 filtered versions did. Across 96,630 double top trades on 8 markets from 2020 to 2026, after real broker spreads, the setup averaged −7% of the risk per trade (−£0.72 per £10 risked) and reached a 1.5× take profit 22.5% of the time.

What it is: Two highs at about the same level, a dip between, and price now closing below the dip low.

Older years vs recent years

2020 to 2023: −6% per trade over 54,189 trades. 2024 to 2026: −9% per trade over 42,441 trades.

The versions that made money

These made money in both 2020–23 and 2024–26, which is how Fred rates a setup green. Each uses its own best stop and take profit.

EURUSD · 15-minute chartmorning (7am–noon UK), after price has been rising+15% per trade · 14% hit target · 392 trades

USDJPY · 15-minute chartafternoon (noon–4pm UK), after price has been rising+17% per trade · 8% hit target · 292 trades

Got a chart open? Upload a screenshot and Fred checks it against six years of tests.

Read my chart free

5 free reads. For learning and demo practice. Not financial advice.

By market

Market	Trades	Per trade	Hit take profit
NAS100 (Nasdaq)	3,286	−2%	27.3%
Gold (XAUUSD)	12,791	−5%	23.7%
USDJPY	13,226	−6%	23.9%
EURUSD	13,898	−7%	21.7%
US30 (Dow)	12,181	−7%	24.4%
AUDUSD	13,817	−8%	21.3%
GBPUSD	13,865	−9%	20.6%
GBPJPY	13,566	−10%	20.8%

By timeframe

Chart	Trades	Per trade	Hit take profit
5-minute	69,815	−8%	21.7%
15-minute	21,849	−5%	24.4%
1-hour	4,966	−4%	24.4%

How Fred tested it

Every time the setup appeared on MT5 price history from 2020 to 2026, Fred entered at the next candle's open, paid the real spread, placed the stop just past the pattern (plus half an average candle) and a take profit 1.5 times the risk. If neither was hit within 48 candles, the trade closed at the last price. "Per trade" is the average result as a share of the amount risked.

Back-tested results, not live trading. Past results don't guarantee future ones. Checked 28 October 2026. All tested patterns · AI chart tools compared