Coil Break USDJPY | No Martingale, 20-Year Backtest

How far you increase it is up to you This EA has one setting: what percentage of funds you may lose in a single trade. Changing this makes the results completely different, even with the same EA over 20 years. Settings After 20 years Multiplier Deepest drawdown ─────────────────────────────────────────────────────────────── 0.5% (default) 2,650,000 yen 2.7x 17% 0.75% 4,200,000 yen 4.2x 25% 1.0% 6,470,000 yen 6.5x 32% 1.5% 14,320,000 yen 14.3x 44% 2.0% 28,510,000 yen 28.5x 55% Note: If starting with 1,000,000 yen. The same validation data from August 2006 to July 2026 was recalculated only by changing the settings.

The initial 0.5% is the safest setting. The figure of 2.7x over 20 years may seem conservative, but this is the result arranged to ensure that you do not sink more than about 20% in the worst case. If you want to increase, you can raise the setting. But as it increases, you will sink deeper. For beginners, we recommend keeping the initial value. 【If kept at the initial value, this is how it grew】 Principal 1,000,000 yen → 2,629,000 yen Number of trades 2,685 Maximum drawdown 10.61% Maximum drawdown means the depth at the deepest point. With 1,000,000 yen operating, even at the worst period you would temporarily be down about 106,000 yen.

【It was fine to miss the buy timing】 When buying this EA, the scariest thing is “what if the timing I bought at happened to be bad, what should I do?” So we divided the 20 years into 17 periods of 5 years each and tallied. If bought in 2006 and held for 5 years → +162,075 yen If bought in 2007 and held for 5 years → +18,176 yen If bought in 2008 and held for 5 years → +170,768 yen (omission) If bought in 2022 and held for 5 years → +559,433 yen All 17 were positive. The worst was starting in 2007 with +18,176 yen. It did not result in a loss. The best was starting in 2018 with +850,982 yen.

※ These are not 17 independent trials but derived from stacking 20 years into units. They are past validation results and do not guarantee future performance. 【To those who have ever skipped with averaging down】 This EA does not increase lot size to recover losses. Average methods - Averaging down: None - Martingale: None - Grid: None - Hedging: None A system that doubles the lot on a loss can make the win rate seem higher. But that merely defers losses; after 10 consecutive misses, the account could be wiped out at once. This EA confirms the losses on the spot each time. Therefore losses do not accumulate. The win rate is 43.80%. More than half of trades are losers. Instead, losses are kept small and wins are extended. Average win +7,882 yen Average loss -5,063 yen

【What to do is, just install and run it】 1. Place the files in the MT4 folder 2. Open the USD/JPY chart and drag it 3. Then leave it alone There is no need to fiddle with settings. The initial values were decided from 20 years of validation. On average 0.46 trades per trading day. Most days nothing happens. You do not need to stay glued to the screen; in fact, staying glued serves no purpose.

When attached to the chart, it looks like this.

【From what amount should you start】 It operates from 100,000 yen. Realistically, we recommend more than 300,000 yen. The annual return for the initial setting (risk 0.5%) over 20 years is about 5%. With 100,000 yen, it would only yield about 5,000 yen per year, making recovery of the EA cost take many years. 300,000 yen → 15,000 yen/year 500,000 yen → 25,000 yen/year 1,000,000 yen → 50,000 yen/year If you raise risk to 1.0%, this roughly doubles. However, the depth also doubles. Refer to the first table to decide. Note: All are based on past validation and are not guarantees for the future. 【Answering “It’s not running, but it isn’t broken, is it?”】 The most anxious moment for first-time EA users is not a loss, but when there are days with no entries. You wonder if it is broken or not. This EA includes a screen that makes that clear. ● Chart panel It shows what it is waiting for in real time. For example, “Contraction OK,” “Almost there,” “Entr y within 2.9 pips” shows how close you are to the conditions in numbers. ● Dedicated dashboard (HTML) Just open in a browser. It shows operating status, asset progression, performance 10 items, current state. It does not communicate with external servers. It will run even in a disconnected environment.● Draw buy/sell points on the chart It shows where you entered and where you closed with arrows and lines. While you are in a position, the stop-loss line is also shown, so you can see at a glance “how much you may lose now.”

【What the EA is doing】 Price movements are not always the same. There are periods of large movement and periods of little movement. After a quiet period, big moves tend to appear. It is like a spring that is compressed and then released. This EA targets that single point. ① Measure the range of recent price movement ② If that range has sufficiently contracted, go into standby ③ If the closing price of a completed bar breaks upward or downward, enter in that direction ④ Always place a stop-loss before entry That is all it does. It does not do anything complicated. 【The settlement mechanism is all written out】 While many EAs hide their exit logic, this EA will disclose it. ● Stop-loss: 3.0 times the average price movement (ATR). Confirmed at entry ● First-tier trailing: Move stop loss upward as price moves in profit direction ● Second-tier trailing: When profit equals stop-loss width, raise stop to break-even + 1 pips. Trades that reach this point generally do not lose ● Time-out: If no settlement within 48 hours, exit ● Weekend close: All positions closed by Friday to avoid weekend gaps ● Spread cap: If spreads are wide, we defer No external servers, no DLLs, none of that is used.

【To avoid stopping midway】Here are some negative numbers. The biggest reason for losing with an EA is not that the EA is bad. It is that you cannot tolerate a bad period and stop. Therefore, I will first note how long the hardest period lasted. Maximum consecutive losses: 17 times Longest period without new asset peak: 962 days (about 2 years 8 months) Years with losses in 21 years: 6 There was indeed a period with 17 consecutive losses. For about 2 years and 8 months, there were times with no growth. Still, the 17 five-year periods are all positive. Difficult periods will come. If you know they will come, you can avoid stopping. These numbers are provided for that purpose. 【Compared with other EAs】 Compare with EAs that have actually been running around USDJPY for about 10 years. Product Years of operation PF Max DD tagosaku USDJPY 13 years 1.23 20.22% nekoboo FX 11 years 1.19 17.70% WAVE For USDJPY 9.7 years 2.26 3.47% Flame Angel 9 years 1.12 28.27% Nanairo USDJPY 4.9 years 1.39 19.06% ───────────────────────── Coil Break 20 years 1.21 10.61% PF (Profit Factor) is total profit divided by total losses. A PF of 1.0 means break-even. Veteran EAs typically sit around 1.1–1.5. The PF of this EA at 1.21 is in the middle, and its drawdown is less than half. ※The other 5 are performance on their sales pages, while this EA uses backtest figures. Since the sources differ, they are not directly comparable. 【What is included】 ● EA file ● User manual (PDF) Covers how to install, what settings mean, and how to troubleshoot ● Backtest validation report (PDF, 8 pages) Yearly results, even items that didn’t meet targets are included ● Dedicated dashboard (HTML) 【Operating environment】 MetaTrader 4 (Build 600 or later) Currency pair: USDJPY only Timeframe: Any, judged automatically using the 30-minute chart Recommended funds: 300,000 yen or more (works with 100,000 yen) VPS: recommended (designed for 24/7 operation) Also supports brokers whose symbol ends with a suffix (USDJPYm, USDJPY., etc.) 【Who it is suited for】 ● Have ever skipped on averaging down ● Want an EA you can understand ● Can tolerate occasional losses per year ● Can leave it for months to years 【Not suitable for】 ● You must have a monthly profit ● You want results within a few weeks ● You want to multiply funds quickly in a short period 【Frequently Asked Questions】 Q. Can beginners use it? A. Yes. No need to change settings. The manual explains how to install in MT4 with pictures. Q. From how much can you start? A. It operates from 100,000 yen, but we recommend 300,000 yen or more. Since the initial annual return is about 5% in validation, 100,000 yen yields only about 5,000 yen/year, taking too long to recoup the EA cost. Q. Do I need to keep my computer on? A. Yes. It is designed for 24-hour operation, so VPS use is recommended. Q. I won’t enter for days. Is it broken? A. That’s expected. On average 0.46 trades per trading day, and there is an 8.4% chance of no entries over 5 trading days. The included dashboard shows how close you are to entry conditions. Q. Can I change the settings? A. You can, but the initial values are those from 20 years of validation. In particular, please do not reduce the width of the first-tier trailing relative to the stop-loss; the reason is in the manual. Q. Can I run it with other EAs? A. Yes. The magic number uses values up to +159, so keep at least 200 away from other EAs. Q. Will the performance be the same as backtests? A. No. Backtests assume fixed spreads, but live trading sees spreads widening and deferrals. The number of trades will be fewer than 2,685. This is noted in the validation report. 【Forward Test】 Period: 2026-09-02 to 2026-09-04 Account: Rakuten Securities Demo / USDJPY 30-minute chart / initial settings (risk 0.5%) Trades: 2 (1 win, 1 loss) / P/L +70,979 yen 2026-09-02 Buy 0.73 lot −15.7 pips −11,461 yen 2026-09-04 Sell 0.36 lot +229.0 pips +82,440 yen Because there were only 2 trades, the results have no practical meaning. Win rate or PF calculated from this sample would be meaningless. The four confirmed points were: - Initial stop-loss placed as designed (ATR 3.0x) - Lot calculation works as set to 0.5% - Two-stage trailing and break-even lock operate - No errors from order to close. Auto-close on Friday The sell on September 4 extended to about 3x the planned stop-loss width (3.03R). It behaved as designed: losses are kept small and wins are extended. But this is a single result; it will not happen every time. For long-term performance, refer to the 20-year backtest above. 【Disclaimer】 The performance described for this product is based on historical data and validation. It does not guarantee future performance. Market moves may incur losses. Please trade at your own risk and judgment.──────────────────────────────── Production by Kururuni ────────────────────────────────