Jon B Good posted this comment and asked this question:

QUESTION: Hi Greg. Great analysis (he was speaking to this post). Thanks for your sound professional approach. In mentioning the EurUSD fall in your analysis I noticed you didn’t refer to the SNB negative rate news and the rise in USDCHF and correlated fall in EURUSD. To what extent would you suggest that traders take such inverse correlated moves into account or do you just ignore it in relation to today’s intraday moves?

ANSWER: Great question…

Much is done studying correlation, relationships of things like the USDCHF and the EURUSD and there effect on the EURCHF, and in even intramarket relationships (i.e., Stocks and the dollar, commodities and the dollar, one bond vs another bond, etc.).

You are right that USDCHF went up, EURUSD went down… TODAY. The USDCHF impact was greater which caused the EURCHF to move higher. That is math as you can be sure there are arbitrage traders who will take the spread out “if the pieces get out of whack”. Today, one went higher, the other went lower and overall, the dollar went higher against both in that window after the headline came out.

Having said that, the EURCHF can go higher by having the USDCHF move higher, the EURUSD staying unchanged or even going higher at a slower pace.

Someone asked me the other day “What happened when the EURCHF moved 1000 pips from 1.1000 to 1.2000 three years ago?”

So I looked it up, and found that the USDCHF moved up by around 750-800 pips, and the EURUSD moved lower by 100 pips.

A trader that day, who hedged the EURCHF exposure by selling EURUSD (in an effort to lower risk) because the USDCHF was in a fast, illiquid market and he could not get out, really had no hedge. The move was ALL (well mostly) in the USDCHF. This is where correlation and all the studies in the world, don’t work.

Think back to the mortgage debacle. Mortgage backed securities tend to move with US treasuries until you have a mortgage crisis. So traders who hedged mortgage securities porfolios with US treasuries saw the price of the mortgage bonds fall dramatically and the price of the US bonds rise dramatically. Whoops! Losses on both sides. OUCH!!!!! That is the definition of LARGE LOSSES and I can tell you, the feeling is not nice.

I have been burned in my career. Thankfully, not to the degree that others have. I learned my lesson. How?

When you get burned, you learn to try not to get burned. I burned myself as a kid. I went downstairs in the morning, picked up matches, lit a match and caught my pajama pants on fire. True story. I learned not play with matches.

To try and not get burned you need to answer the question “Why did I get burned?” In the financial markets (not the match lighting example) the answer is “I relied on correlation” and “Why did correlation not work?”. Because it sometimes does not work and when it does not work, it can be REALLY painful. “Why is it really painful?” Because you cannot define risk using correlation.

In other words, you have an idea it might work but you really don’t know you are really getting burned when it does not work. You not only get burned, you can blow up.

So, the long answer to your short question is that I will prefer to look at each pair separately to define and limit risk and make my assessment on bullish or bearish bias. I may (and often) do have an understanding of what is driving EURUSD, USDCHF and EURCHF (i.e., the correleation) , but I know, if I take a position because of that correlation, and want to get out of that position, my risk defining levels will be on the pair I am long or short – not on something else that may not be correlating “like it should”.

I hope this lesson into the way I think, is clear. Think about it. Chew on it.

As I mention each and every day, it is all about defining and limiting risk.