SignalForge
Changelog

Updates

Every system change, recorded with its reasons and test results.

Ichimoku MTF2026-10-11

Ichimoku MTF v3.7 — per-pair risk weights + strong-yen rule

Risk per trade is now set per pair, and yen Buy trades are cut while the yen is strengthening hard. Entries and exits are identical to v3.6.

  • Return per year 28.9% → 30.3% · max DD 11.2% → 9.1% · worst year +4.3% → +7.5% (hypothetical, same basis)
  • Adding silver (XAGUSD) was tested: the return gain is real, but the lower drawdown is not reliable, so it is not used yet
  • Started running with the bot on an XM Demo account with $1,000, with stats linked to Myfxbook
Ichimoku MTF2026-10-05

Ichimoku MTF v3.6 — fewer losing streaks

4 test rounds (more than 100 variants) to shorten losing streaks without hurting the other numbers.

  • Longest losing streak: USDJPY 10 → 7 · EURJPY 8 → 6 · gold 7 → 4 · GBPJPY 8 → 5
  • Found why the Python results did not match TradingView (D1/H4 bar close time and the reset at each Replay window). Fixed: 99.7% match.
Gold Swing / Gold H12026-10-05

Gold Swing v7.2 + Gold H1 Signal V2

Gold Swing v7.2 enters only in strong big trends (ADX W1 ≥ 25, D1 ≥ 20): PF 3.89, 68.6% wins, no losing year in 20 years. Gold H1 Signal V2 was added as its partner to increase the number of trades.

  • Both together (0.6% + 0.3%): about 1.8 trades a month, PF 3.04, DD 3.5%, no losing year

Interested in a system, or want to ask first?

Talk to us directly — every message is answered personally. You trade in your own account; we never manage your money.

Chat on LINE