Does a setting found on USD/JPY still work on another currency pair?
Do Indicator Settings Transfer Between Pairs? 3,024 Tests
Finding
The ordering does not carry. All 42 published settings ran on USD/JPY, EUR/USD, GBP/USD and EUR/JPY over 2023, 2024 and 2025 on three timeframes. Ranking the 42 by annual pips on USD/JPY and comparing with the same 42 ranked on another pair gives correlations from -0.515 to +0.47 across 27 cells, with a median of -0.011 and only 13 cells above zero. The count of settings profitable on both pairs differs from what two unrelated pairs would give by -4 to +6, median -0.2. Carrying USD/JPY's best setting of the year to another pair made money in 12 of 27 attempts. No setting was profitable on all four pairs in all three years, on any timeframe.
Key results
- Rank correlation with USD/JPY
- -0.515 to +0.47 across 27 cells, median -0.011, 13 above zero
- Profitable on both, against two unrelated pairs
- Differs by -4 to +6, median -0.2
- USD/JPY's best setting carried to another pair
- Made money in 12 of 27; ranks from 1st to 36th of 42
- Profitable on all four pairs in all three years
- 0 of 42 on every timeframe
- How far the profitable count swings
- Hourly 2024: USD/JPY 36 of 42, EUR/USD 5, GBP/USD 6, EUR/JPY 8
- Median year by pair, hourly
- USD/JPY is the only pair positive in all three years
- Mean bar range, hourly
- USD/JPY 24.85 pips, EUR/USD 15.27, GBP/USD 17.12, EUR/JPY 24.72
Scope
- Instruments: USD/JPY, EUR/USD, GBP/USD, EUR/JPY
- Period: 2023-01-01 to 2025-12-31, split by year
- Timeframes: 15-minute, 1-hour, 4-hour
- Conditions: the 42 published settings
Method
Each article carries a "run this setting" button holding the rules it published; 44 of those exist and the three Bollinger articles publish identical rules, which leaves 42. Only the pair and the window were changed: rules, lot size, fill and spread stay as published. Pips are counted in each pair's own pip, and the year is also divided by that pair's mean 14-bar true range so four pairs sit in one column. 42 settings x 4 pairs x 3 timeframes x 3 years x 2 spreads (0.3 pips and 0) = 3,024 runs. Rank correlation is Spearman with tied ranks averaged. Every figure is checked in research/pairs-verify.run.ts.
Limitations
- Four pairs over three years.
- Fills at the close, one 0.3 pip spread for every pair, no slippage; real spreads differ by pair.
- One published setting per article, not that article's full parameter sweep.
- Only one pair in the set leaves the dollar out, so the yen-versus-dollar question rests on EUR/JPY alone.