Backtest evidence

Does an engulfing candle behave differently from a colour flip that did not engulf?

Is the Engulfing Candle Profitable? 396 Backtests on USD/JPY

Finding

On the hourly chart, in both years. Ten bars later the pattern had moved its own way 53.40% of the time in 2025 (839 signals) and 52.81% in 2024 (818), against 49.12% (2,327) and 48.98% (2,297) for pairs that flipped colour without covering the previous body. The 15-minute chart gives +0.50 and -0.37; the 4-hour gives +3.79 and -2.72, changing sign between years. Of the requirements write-ups add, covering the wicks lost to body-only in five of six cells, and restricting to signals that follow an opposite move left the weaker half on the hourly chart in both years. The only filter that improved every cell was the engulfing bar's body filling at least 70% of its own range.

Key results

Hourly 2025, engulfing against the control
53.40% from 839 signals against 49.12% from 2,327 (+4.28 points; +4.28 pips against -1.58)
Timeframes clearing the control in both years
Hourly only. 15-minute +0.50 and -0.37; 4-hour +3.79 and -2.72
Covering the wicks as well
Body-only had the higher hit rate in five of six cells
Body filling at least 70% of the range
Improved all six cells, by 0.62 to 2.98 points, at about a third of the signals
Broke the far end of the bar before the same distance the other way
Engulfing 48.40-53.27%, colour flip 53.67-57.28%, solid body 43.73-50.71%
Broke the far end on the very next bar
Engulfing 17.21-21.79%, colour flip 32.35-34.40%, solid body 10.60-15.53%
Hourly, held to the opposite pattern, profitable in both years
10 of 11 conditions. The plain pattern: 439 trades, 43.28%, +813.8 pips in 2025; 426 trades, 42.96%, +2,180.8 in 2024

Scope

  • Instrument: USD/JPY
  • Period: 2024-01-01 to 2024-12-31 and 2025-01-01 to 2025-12-31
  • Timeframes: 15-minute, 1-hour, 4-hour
  • Conditions: the plain pattern plus 10 filters

Method

Engulfing bars were counted with the same code path the backtester runs, and the control is the set of two-bar pairs that flipped colour without covering the previous body. Trades enter at the close of the completing bar and exit either on the opposite engulfing candle or after ten bars, at a fixed 0.3 pip spread, 0.1 lots and no slippage. Filters are built by switching on existing conditions beside the pattern (marubozu for the body ratio, large candle for the ATR size, MA slope for direction) and the exit always uses the plain pattern. 11 conditions x 2 exits plus one spread-free pass x 3 timeframes x 4 windows = 396 runs. Every figure in the article is checked against the run output in research/engulfing-verify.run.ts.

Limitations

  • One pair over two years.
  • Fills are at the close with a fixed spread and no slippage, so live results will differ.
  • The filters are combinations of existing conditions, not thresholds added to the engulfing test itself.
  • In the false-break test, a bar touching both the far end and the target is counted as the failure.