Tokyo London Fix
- Experts
- Version: 1.1
- Activations: 5
Follow the Tokyo fixing and the London month-end fixing from one USDJPY chart in MT5. The panel shows the next check and the last decision. Before buying, replay 21 minutes of the Tokyo session in the free demo and see the difference between a signal in the Journal and a trade in deal history.
Tokyo checks three yen pairs on scheduled Japanese settlement dates. On the last weekday of the month, London checks how far a US equity index has moved since the start of the month. A qualifying move triggers a check of US dollar pairs. Either strategy can run on its own (guides A and H).
Replay the Tokyo fixing in the free demo
- Set up the replay (B–D). In desktop MT5, get the Market demo and press Ctrl+R. Choose Tokyo London Fix, USDJPY, M1 (one minute), 29–31 October 2026, real ticks and visual mode. Match the historical server time settings to your broker.
- Watch the decision (E). Set “Observe only (no new orders)” to true and start. In the reference test, all three yen pairs logged “observation: sell” at 00:54 UTC on 30 October. Zero new orders is the expected result in this mode.
- Compare the fills (F). Set observation to false and run the same dates again. The reference simulation sold USDJPY at 157.492 and bought it back at 157.295 at 01:15 UTC. Compare symbol, direction and time first. Prices vary by broker. The Journal explains skipped orders (G).
How the balance changed
When comparing settings, we evaluate how steadily the balance grew as trades closed and weigh net profit against the largest fall in account value. In the version 1.01 backtest, JPY 1,000,000 became JPY 1,109,532. Nine of ten months ended higher. June lost JPY 6,112 and October gained JPY 44,934. The monthly and balance charts show that path.
Exness MT5 history, 1 January–31 October 2026, JPY account, 1:500 leverage, 100% real ticks reported by MT5, no added delay. The 219 trades returned JPY 109,532 after booked costs. Gross gains were 2.18 times gross losses. Including open positions, the largest drop from a previous peak was JPY 19,093 (2% as displayed by MT5). These are historical simulations.
Inputs and default values
Start here and general settings
- Observe only (no new orders) / InpObserveOnly (false) — Set to true to record decisions without opening new positions.
- InpMagic (26111201) — Base number used to identify orders from this EA.
- Historical server winter offset, minutes east of UTC / InpTesterUtcMinutes (0) — Winter server offset from UTC in minutes for historical tests. Use 120 for UTC+2.
- Historical server daylight-saving rule / InpTesterDst (FixedOffset) — Broker daylight-saving rule for replays. Select fixed time, European summer time or New York summer time.
- InpMaxDDPct (15%) — Drop in account equity from its peak (%). At the threshold the EA stops new entries and tries to close its positions. Zero disables this protection.
- InpMaxLot (20) — Maximum lots per order.
- InpSlippagePoints (20) — Requested execution price deviation, in points.
Tokyo fixing
- InpG_Enable (true) — Enable the Tokyo strategy.
- InpG_Symbols (USDJPY, EURJPY, GBPJPY) — Yen pairs to check, separated by commas. Use your broker’s symbol names.
- InpG_RiskPercent (0.45%) — Base risk as a percentage of equity for sizing from the stop distance. Tokyo applies a weight to each pair.
- InpG_MaxSpreadPips (2.0) — Maximum entry spread in pips, the gap between buy and sell prices.
London month-end fixing
- InpM_Enable (true) — Enable London. Set to false for a Tokyo-only replay.
- InpM_Symbols (EURUSD, GBPUSD, AUDUSD, NZDUSD, USDJPY, USDCAD, USDCHF) — Comma-separated list of US dollar pairs for London.
- InpM_EquityIndex (US500) — Your broker’s cash S&P 500 equivalent. If the panel shows “No index”, check its exact Market Watch name.
- InpM_RiskPercent (0.25%) — Risk per London order as a percentage of equity, used to calculate position size.
- InpM_MaxSpreadPips (2.0) — Maximum London entry spread in pips.
Scheduled exits require MT5 to stay connected, locally or on a VPS. Simultaneous positions, price gaps and execution costs can push losses above the planned amount. Support is available in the MQL5 product comments and messages.
