Gotobi Edge Portfolio
- Experts
- Version: 1.0
- Activations: 10
Gotobi Edge Portfolio trades USDJPY only, on fixed calendar days around the Japanese Gotobi days, the Tokyo fixing and the month end. It uses no indicators and no grid, martingale or hedging. Each module enters at most once per day (a large order may be split into several fills), and every position has a stop loss and a time exit.
Modules
- Gotobi-day short: a short position taken on Gotobi days and held during the Tokyo session.
- Month-end long: a long position taken in the last business days of the month and closed the same day.
- Pre-fixing long: a long position taken on Gotobi days before the Tokyo fixing and closed in the morning.
The idea is that many Japanese importers settle in dollars on Gotobi days, and the Tokyo fixing sets the reference rate. The EA does not predict price. It trades a repeating calendar pattern that was observed in historical data, and that pattern can weaken or disappear.
Risk management
- Every position has a stop loss and a time exit.
- Position size is set from the account balance and the Risk setting (default 1.0% of balance), capped by margin allocation and a minimum margin level.
- New trades are not opened when the spread is above the limit.
- A daily and a weekly equity-drop limit stop new entries.
- If a stop loss cannot be attached, the position is closed.
- If the server time zone cannot be recognised (outside UTC+2 to UTC+3), the EA does not open new trades.
Requirements
- Symbol: USDJPY. Other symbols are not validated. Any chart timeframe can be used.
- Account type: hedging.
- Server time: the usual MetaTrader server with GMT+2 in winter and GMT+3 in summer.
- The terminal must run continuously (a VPS is recommended) and algorithmic trading must be enabled.
- Suggested minimum balance: JPY 100,000 or about USD 700. Smaller balances are rounded to the minimum lot size, which makes the real risk larger than the Risk setting.
What to expect
- The EA trades on roughly six to eight days per month. It is idle the rest of the time.
- A small number of large winning days made up a large part of the historical profit in the backtest. Long flat or losing periods are normal. Do not use money you cannot afford to lose.
- The EA contains built-in calendar tables (for example Japanese bank holidays and central bank meeting dates). They are updated by new versions of the EA. The chart shows a notice when a table has expired.
Main inputs
- InpUseE1_Gotobi, InpUseE3_MonthEnd, InpUseE9_PreFixLong: switch each module on or off.
- InpRiskPercent: risk per trade, % of balance (default 1.0).
- InpMaxSpreadPips: maximum spread for a new trade.
- InpUseKill, InpKillDailyPct, InpKillWeeklyPct: stop new entries after a large equity drop.
- InpMagicBase: magic number base.
- InpAllowOtherSymbols: allow symbols other than USDJPY (not validated).
- InpWriteLogFiles: write an event log to the common data folder.
The other inputs set the filters, the stop losses and the margin limits. The defaults are the tested settings.
Backtest note
Strategy Tester simulation, every tick, USDJPY, 2016-01-01 to 2026-10-09, deposit USD 10,000, leverage 1:100, default inputs (Risk 1.0%). A simulation is not a real trading result and does not guarantee future results. Spread, slippage and execution can differ on a live account.
- Net profit: USD 321,864 on a USD 10,000 start (the lot size grows with the balance, so the result compounds).
- Profit factor: 1.40.
- Maximum drawdown: 24.4% (balance basis, closed trades). While a position is open the equity can be lower by up to the loss of one stop-loss, which is about 4% of the balance at most.
- Trades: 1,728 (date x module); a large order can be split into several fills.
- Win rate: 55.9%.
- 10 of the 11 calendar years were positive. The latest year (2026, to October 9) was -11.3%.
- A few large winning days made up a large part of the profit: the best 1% of trades gave about one third of the total log-return, and the best 5% gave about all of it. The other trades together were close to break-even.
- With the default Risk of 1.0%, the position is about 5 to 8 times the balance in notional terms.
- At a very large balance the broker volume limits reduce the position size, so the compounded result of a large account is not reproducible.
Support
Please send questions through the MQL5 private messages. Important fixes are published as updates.
