Thoughts? Suppose you design a technical system that works well on most of the major pairs, but performs badly on crosses. Or perhaps you design a system that’s worked on one pair for many years, but fails on almost everything else. Why do you think this would be? I’m looking for some thoughts and ideas. 
pairs are like women. They are not all the same
I think it really depends on your strategy. In the end, at least for me, price action is price action and 50 pips are 50 pips. Maybe the biggest disruption is do to fundamental factors. I apply one strategy across 40+ currency pairs and overall am pleased with results.
I’m finding the same thing on backtesting my technical system. I guess it makes sense as every pair is independent of each other (ignoring the obvious mathmatical link with crosses), and each behave differently, even if they are positively correlated. I put it down to different markets that may be open, news announcements, volatility differences, and any large orders to hit the markets in certain pairs.
I would like to have faith in my backtesting to have a robust technical system that works well on most pairs and most timeframes but I’m just not finding this to be the case. Logically some strategies work better on some pairs. E.g. Big Ben breakout on GBPUSD (I dont actually trade this strategy but it makes logical sense).
Hope that helps!
Cheers,
steve
^ Hey Steve. I’m stuck in the same boat.
Here’s one of my thoughts. Suppose we’re looking at the EURCAD, which is going be influenced by the EURUSD and USDCAD. If we examine the three pairs only, 75% of traders might focus on the EURUSD, 20% on the USDCAD and 5% on the EURCAD alone.
The EURCAD measures the difference between the EURUSD and USDCAD. If 95% of traders are trading the EURUSD and USDCAD, perhaps those 95% of traders will have a greater influence on the EURCAD than the 5% of traders who focus on the EURCAD alone? Hence why trading cross-pair charts might be more difficult? (the traders influencing the cross-pair will be trading the majors, so their influence won’t be obvious on a cross-pair chart).
Like speculating it’s all a matter of perspective. Some might say pairs are like men. They are not all the same lol
I find the most volatile pairs to be more consistent.
Yeah that makes sense too! It is reasonable to consider crosses as “following” the trends of the majors and minors that make up the pair.
The other thing I am finding with my backtesting is how sensitive some input parameters are. So if I am looking at optimising 1 currency pair, and set TP to be 1R, the results are quite different to if I set TP at 3R, for example. Same thing applies for Bollinger Band Settings, PSAR, MACD, Moving Averages, etc.
Sure the TP varying results makes sense but small changes in a moving average leads to large changes in final returns of the backtesting. Makes it difficult to know what settings to go live with - I need to settle on a happy medium.
I think that fundamentals is the reason, because fundamentals is what really moves the market and certain pairs are moved by different fundamental reasons.
I think pairs behave differently. You need to get experience on them before giving analysis. It is very hard for them to learn or analyze more pairs. So they will stick with one pair