Chainwale Gold Channel

Chainwale Gold Channel is a Donchian unit-breakout Expert Advisor for XAUUSD, inspired by the publicly documented 1983 Turtle Trading rules (Richard Dennis and William Eckhardt). It is an original implementation, not affiliated with the original Turtles. A stop loss is placed on every order. It uses no grid, no martingale and no averaging down. It makes no DLL calls or web requests and does not use AI.

How it trades

• N is the 20-period Wilder ATR on the working timeframe (Daily by default).
• System 1: buy when price breaks the prior 20-bar high, or sell when it breaks the prior 20-bar low. The campaign exits on the opposite 10-bar channel, or on the stop.
• System 2: the same idea with a 55-bar entry and a 20-bar exit. A 55-bar break is taken as System 2. Both systems share one campaign; the EA does not run two overlapping positions on the same symbol.
• Unit size: each unit is sized so that a 1N adverse move equals 1% of equity (default). The stop is 2N from the newest unit, so one unit risks about 2% at the stop.
• Pyramiding: a further unit may be added every 0.5N, up to 4 units, if the campaign's total stop risk stays inside the 8% cap. After each add, the stop for the whole campaign is raised to 2N from the newest fill. On netting accounts this adds volume to the same position; on hedging accounts each unit is a separate ticket with the same shared stop.
• Entries are intra-bar breaks of the completed-bar channel (the classic rule). An optional closed-bar mode and an optional System 1 "skip after a winner" filter are included; both are off in the tested defaults.

Risk controls (defaults)

• If the broker's minimum lot is larger than the calculated unit, the EA takes the minimum lot only when that stop would risk at most 20% of equity; otherwise it skips the trade and writes the reason to the log. On a small gold account this usually means a single 0.01 lot and no further units, because adds must still fit the 8% total-risk cap. The stop may be tightened to no less than 1N so the minimum lot can fit the cap.
• An 8% equity loss within a server day blocks new entries until the next day.
• If equity falls 30% below its peak, new entries pause for 5 days. This pause exists to limit damage; it does not cap the open drawdown of a campaign already running.
• Entries are skipped when the spread is above 0.15 × N.
• Before any order it checks margin, free margin, volume limits and volume step. It checks the stops level and freeze level before every order, modification and close.
• It runs on both hedging and netting accounts. On netting accounts it does not open a new campaign if another position or order already exists on the symbol.

Backtest results (honest, one year only)

The main test ran in the MT5 Strategy Tester on XAUUSD.m Daily with real ticks (history quality 100%), from 2025-11-14 to 2026-11-14, starting with USD 10,000 at 1:100 leverage on JustMarkets demo data, using the default inputs. Hedging and netting produced essentially the same net result (24 unit-trades vs 9 campaign trades, because hedging counts each added unit separately).

Maximum equity drawdown over the full year was 5,403.94 (35.99%). Maximum balance drawdown was 1,127.36 (10.08%). Recovery factor 0.26; Sharpe ratio 0.47; 15 of 24 unit-trades profitable (62.50%). The equity drawdown is much larger than the balance drawdown because units are added as a trend extends; open profit can be given back before the 2N stop or the opposite-channel exit is hit.

Settings were chosen in three light tuning passes on the first eight months only (2025-11-14 to 2026-07-14). In a standalone test of that window the tester closed open positions at the cut and reported +1,588.56, profit factor 2.37. In the full-year run, one four-unit campaign was opened in May–June (inside the tuning window) and closed on 22 July, so a large part of the +1,321.48 “held-out” closed-deal figure comes from a trade that started during tuning. Later held-out campaigns also existed (entries in August and September). Treat the held-out numbers as a short sanity check, not as proof.

Daily charts were chosen over H4 after an 8-month OHLC comparison: H4 lost money with about 50% equity drawdown in that window, so it was not used for the official real-tick tests.

Small-deposit netting checks (USD 1,000, 1-minute OHLC, 2025-11-14 to 2026-11-14, same defaults):

• EURUSD H1 (signals still read on Daily): net 121.54 from 12 trades, equity drawdown 167.93 (15.90%).
• XAUUSD Daily: net 639.00 from 7 trades (all minimum-lot, no added units), equity drawdown 1,289.93 (55.88%). Two signals were still skipped because the minimum lot would have exceeded the 20% cap. This equity swing is too large for a USD 1,000 gold account; the figure is a technical check only.

None of these runs produced trade errors or a stop-out. The EA was designed and tested for gold. EURUSD is shown only as a technical check and is not a recommendation.

Recommendations

• Attach it to XAUUSD on a Daily chart (or leave WorkTF at Daily on any chart). H4 is supported but was not the tested default.
• Use an account of at least USD 10,000 so the 1% unit size is usually above the 0.01 minimum lot. Smaller gold accounts will often run a single minimum lot and can see very large equity swings.
• Choose a low-spread ECN or raw-spread account and run it 24/5 on a VPS. This is a slow system: a handful of campaigns per year, sometimes held for weeks.

Risk warning

Trading leveraged products such as gold CFDs carries a high risk of loss. These results are historical simulations over a single year. They do not guarantee future results, and live spreads, slippage and swaps can differ. Equity drawdown of about 36% occurred in the tested year even with the default risk caps; it can be larger. There is no live signal for this EA yet. Losing streaks will happen. Never risk money you cannot afford to lose, and test it on a demo account first.