Waiting for the closed bar beats jumping in mid-bar, everywhere
Waiting for the price to close at a specific level is a more effective filter than chasing real-time price movements.
Waiting for the price to close at a specific level is a more effective filter than chasing real-time price movements. My latest round of testing confirms that the current strategy of waiting for a candle to finish before acting is superior to attempting to trigger trades the moment a price level is touched.
In my current setup, the EA uses an IsNewBar check to evaluate conditions only once a candle closes. This ensures that the engine logic and the actual trading implementation are perfectly aligned. To test if this is truly the most efficient approach, I compared this “Close/Close” method against “Touch” variants, which execute the trade the moment a price level is hit. This comparison covered three specific breakout strategies: Trend-Core FX (across four pairs and three timeframes), US3 indices, and TJL-Gold.
I maintained identical costs and sizing across all tests to ensure a fair comparison. Here is how the performance metrics compared for the Trend-Core FX strategy:
| Method | Monthly Return | PF | Drawdown |
|---|---|---|---|
| Close/Close (Current) | +0.149% | 1.27 | -6.5% |
| Touch Entry | +0.140% | 1.21 | -7.5% |
| Touch Exit | +0.122% | 1.21 | -6.8% |
| Both Touch | +0.128% | 1.18 | -7.0% |
| In other words, the “Touch Entry” method increased the total number of trades by 15 percent, from 2,142 to 2,456. However, this increase was entirely composed of “false breakouts” that the current candle-close filter successfully ignores. By entering the market the moment a price is touched, the strategy ends up catching more noise than actual trend movement. | |||
| The results for TJL-Gold were even more decisive. The current Close/Close approach achieved a Profit Factor (PF) of 2.54, representing the ratio of gross profit to gross loss, compared to 1.76 for the touch-based execution. This aligns with my previous research finding that the “edge” in these trades comes from the price drift that occurs over the days following a breakout, rather than the initial, volatile reaction to hitting a price level. | |||
| While the US3 index strategy showed a minor nominal improvement with touch-based entry, moving from a PF of 5.52 to 6.73, the sample size was extremely small at only 41 trades. The difference in monthly return was a negligible 0.008 percent, which I categorize as statistical noise rather than a meaningful improvement. | |||
| The conclusion is clear: waiting for a candle to close is not an inefficiency. It acts as a free, built-in filter that prevents the system from being whipsawed by temporary price spikes. The advantage of getting a “better” entry price by hitting a level immediately is simply not worth the cost of the increased false signals. Consequently, I will not be moving to a tick-based or stop-order system. The EA and the engine will remain exactly as they are. |
How this connects
This verification builds on earlier ones (what failed before and what I tried this time, comparisons between approaches).