# Currencies Correlations 4

Salve a tutti,

In my previous post I described how to calculate and use the currency score that shows strength/weakness of each currency.

But there’s also an even more important analysis that we should perform before deciding to enter a trade: the correlation between currencies.

Using the new indicator that I developed – which will be available to everyone soon – you can also read the correlation score between each couple of currencies.

The correlation is a value that tells you how related are two data series (in our case the “scores” of two currencies) over a specific period of time.

The correlation score can go from -100 to +100 (or from -1 to +1). The higher the score, the more related the two data series are, meaning that they tend to “move” in the same direction.

Vice versa if the score is negative it means that the two data series are inversely correlated. So the lower the score, the more the two data series move in opposite directions.

If the correlation is around 0, it means that the two data series are not correlated and so they will tend to move independently from each other.

The analysis of correlations between currencies let us understand more in depth the general “rules” that move the Forex market. At least for the period taken into account.

Let’s take a look at the correlation score on the daily timeframe (it uses the last 20 days of data, that is around one month of trading).

We can divide the 8 currencies in 3 “families” based on the direct correlation between them.

## Group 1 – AUD, NZD, EUR and (usually) GBP Group 2 – USD, JPY and (lately) GBP Group 3 – CAD and CHF

Note that I had to add a “usually” before the GBP as in general it tends to have a direct correlation with EUR, AUD and NZD but lately it is directly correlated to the currencies in the second group (USD and JPY). This proves that correlations among currencies can change over time due to many different reasons. To understand if the change is only a temporary divergence from the regular behavior, we should look at the correlation value over a longer period or on a higher timeframe.

But why is the currency correlation so important for our trading? In my previous post I wrote that we can have a good entry signal when we see a cross between the lines of two currencies. But if we want a higher percent of success of our trades we need “momentum”.

You can think about Forex market as a “war of currencies” where directly correlated currencies are allied and fight the currencies that are inversely correlated to them. So Group 1 is against Group 2, and Group 3 is somehow “neutral”. Nothing political in that but I always think it is funny that CHF (Swiss Frank) is in the neutral group 😉

We have “momentum” between currencies that “hate” each other (inversely correlated). So we have a greater chance that our trade will be a profitable one because the two currencies will tend to move away from each other. Remember that currency pairs are nothing else than “ratios” between currencies. So when we look at EURUSD price we are looking at the ratio between the value of EUR and USD.

So the rules are simple: look for crosses of currencies scores among inversely correlated ones (negative correlation score, the lower the better).

That took me to a simple but interesting finding: the more two currencies are negatively correlated, the greater is the average daily range of their cross.

For example: EURUSD, GBPUSD, EURJPY, GBPJPY have a high daily range (so they are said to be more “volatile”), while EURGBP and USDJPY have a low ADR as they are crosses between “allied” currencies.

To talk a little about inter-market correlations, you’ll see Group 1 currencies to be also directly correlated to the stock market. So when world stock indexes grow, they’ll tend to grow as well. And vice versa Group 2 will tend to become weaker. The opposite applies to bearish stock market… Group 1 will fall and group 2 will grow. That is for example, why when there are good news for the US economy, the stock market grows and the USD loses ground… It took me time to get used to that when I started trading Forex 😀

CAD and CHF instead (Group 3), are known as “commodity currencies” as they are more related to gold (CHF) and oil (CAD) than to other currencies. They usually are not correlated to group 1 and group 2 but occasionally there can be direct/inverse correlations like it happened this summer when gold was inversely correlated to the stock markets.

Those are general rules. There are periods when correlations change and those are good occasions to trade “contrarian” waiting for the “out of the rules” currency to come back home (like GBP now for example).

To better check correlation evolution between currencies I also coded an indicator showing the correlation as a line.

That way you can check the score of two currencies waiting for a cross, as well the correlation between them.

You can use that as an “integration” to you trading system to filter pairs that are in the “perfect” conditions for trading. Personally I trade the system as it is (using an automatic strategy applied to all the 28 possible crosses). But you can do the same by applying the indicators to the 28 charts and wait for an alert to popup. More about the indicators soon 😉

### I’m also happy to announce that Thursday December 1st the indicators shown in my last posts will be available here in PympMyEA.com website.

So you’ll be able to trade using currency correlation and strength/weakness.

## 4 thoughts on “Currencies Correlations”

• Dan

Hi Andrea, good stuff!

I am very much looking forward to trying out these indicators.

In the EU chart above would the best and safest trades be in the May and June periods when the correlation extreme at its greatest (ie: less than -50)?

Do you have trading rules such as only trade a pair when the correlation is less than -50 for example?

I’m also interested to hear the benefits of trading 28 different crosses when so many of the pairs are often tightly correlated to each other.

Thanks, Dan

• Andrea Post author

Hi Dan,

if the correlation is below a certain value (-50 is a good choice) you can expect to have a “safer” entry signals as being inversely correlated currencies will tend to go in opposite directions. So I use the correlation as a filter for entries.

We use all the 28 pairs to calculate the currency scores (and so their correlation). We won’t trade all the 28 pairs but only those inversely correlated between them. But you have to consider that based on the timeframes the scores and the correlations may vary.

You don’t need to use them all but the more you have the more possibilities to have a trading opportunity you have.

Also another way to trade is to trade the “bounce from extreme values” of the scores.

In a few days I’ll post some screenshot showing you the different uses of the indicators and their signals.

• Leyenda

Hi,

I am new to this website and trying to understand the principles behind correlation. Most of the websites show correlation between two currency pairs, not between two currencies, and I do not understand how to get from one to the other.
Also, the 2PFX strategy depends on creating correlation groups. I have read everything you wrote about that, but I still do not understand how you get from the list of correlation scores to the 3 groups. Of course, this is an essential step in trading the 2PFX strategy, so that leaves me stuck for now. Can you please enlighten me?

Thanks!

• Andrea Post author

Ciao,

well the correlation between two currencies cames before the correlation between two currency pairs (that is a simple “ratio” between two currencies).

What it tells you is how “related” two currencies are and lets you understand in depth how the Forex market behaves and moves. Knowing for example that “in general” (there are always exceptions”) AUD/EUR/NZD are positively correlated between them and so are USD/JPY makes you understand that it’s like having two main “parties”. GBP used to be part of the first group but now it seems to be quite stably part of the second.

When we look for good trading opportunities we look in general for “trending” pairs. To have trend there must be some “struggle” between the two currencies. That’s why we look for pairs made out of group1-group2 mix (like EURUSD, EURJPY, AUDUSD, etc).

So checking correlations between single currencies will help you pick up the right “pairs”. Those with a better potential to develop a strong trend. Other indicators (like the PotenzaFX Powerlines) will help you deciding when those trend develop and in which direction. In theory that’s all you need for trading

The 2PFX strategy relies on a little more complex strategy but not so much. Is made out of mixing together two strategies: PotenzaFX (that I just described) and PaniereFX. You can find all the info about them in the blog. We described everything in details.

But I hope that things are a little more clear now 😉