I use H3 timeframe for my setups and M5 for entries. I consider myself as a swing trader and I trade around 30 currency pairs. However I’m still kind of confused about correlation. I use this webiste to check correlation: Forex Correlation - Mataf
However sometimes I see that currency A and currency B has say 85% correlation on H4 timeframe but when I check H1 or M5 timeframe then the correlation is maybe even below 20% - basically no correlation at all. So how should I approach sitatuation like this ? How should I use correlation in similar situations ?
Short-term technicals, chart patterns etc. are less and less reliable the shorter the total period you consider. But you’re not wrong to consider correlation. Highly correlated currencies don’t always move in lock-step but they tend not to move far apart. So, except for very short-term a long position in USD is not hedged by a long position in CAD.
Plus the most correlated currencies make up pairs which don’t move dramatically per day/month etc.
The 7 most correlated pairs, ranked by their long-term historical volatility are -
AUD/NZD 0.41
AUD/CAD 0.48
EUR/CHF 0.48
EUR/GBP 0.51
EUR/CAD 0.54
USD/CAD 0.54
NZD/CAD 0.55
Other volatility data could give slightly different rankings.
Correlation has more impact on higher time-frames.
A short time-frame for price charts is like a small sample of people. With a small group of people, perhaps 10 only, maybe 5 could be blonde, and 5 could be bald. But in the wider world, these ratios from such a small sample are dead wrong - nobody would say that half the world’s population is blonde and the other half is bald.