In my trading plan (see link to this publication under Related Ideas) I take currency correlation into account. I use a fixed fractional money management system and the risk per trade is a stable % of my trading capital. Before entering a potential trade, I crosscheck the currency correlation of that pair with my already open trades (if any) and if its highly correlated (either positively or negatively) to one of them, I do not enter the trade, even if the setup itself is valid. This is a sometimes fiercely debated topic in the forex chat and I have seen members and traders with very different opinions on it.
Some say I am crazy for doing this check, because I deliberately pass on valid trade setups this way and I could make more pips if I would just ignore correlation and always trade an opportunity when it present itself. “Think opportunity!”, they say. Others are even bolder and claim I should deliberately look for highly correlated sister pairs and always trade them both at the same time. “Double your profit in almost any trade!”, they proclaim. And then there are traders who believe I should already exclude a potential trade setup, if it’s moderately correlated (40%) to an open trade. “Anything above that % is too risky!”, they say.
Here is my reasoning:
Highly positively correlated pairs
If I were to buy or sell two highly positively correlated pairs, I would basically be doubling up on my position in one of them, since both pairs move in the same direction. I would not really get two independent chances to win or lose a trade; I would get only 1 due to the very strong correlation. In doing so, I would be doubling my traderisk, which would be a breach of my risk management rule.
Highly negatively correlated pairs
If I were to buy or sell highly negatively correlated pairs, I would basically be hedging against myself (since the two positions would cancel each other out) so the profit of one would be eaten by the loss of the other. The only exception here could be a higher timeframe trade on one of the pairs with a scalp on a lower timeframe on the other, to benefit from both timeframes.
A common misconception is that pairs with the same currency either as the base or as the quote are always highly correlated. Like EURGBP would be highly correlated to EURJPY , since buying those pairs would in both cases mean buying the euro . This is not always the case. Not only the base currency Euro has an effect here, the quote currencies Pound and Yen have an effect as well and those move independently of each other. So correlation is not fixed, it is different per timeframe (correlations on the 1h might not be in line with those on the daily) and it changes over time as well. That is why I use a real-time source to determine the correlation between two pairs on a particular timeframe. There are several sources to be found on the internet, I don’t want to endorse any trading website, but Google is your friend.
Pairs don’t trade completely independent of each other. And trading highly correlated pairs at the same time can increase your overall risk or eat your profits. All in all, I would say currency correlation does matter and its best not to ignore it. It can help you to manage your portfolio.
Interesting read. I guess I've been basing my strategy around the same principle but from a different source (COT report).
Not sure if anyone uses this? I've heard mixed opinions around that too, such as it's all a lie.. It's only there to scam retail traders etc.
Any thoughts on this?