Each cell in the following tables contains the correlation coefficient for two currency pairs (**currency correlations**) which are named in the corresponding fields of the upper and left-hand panel.

Correlation coefficient measures how closely two currency pairs move together. If both pairs move up and down in** perfect unison **then their correlation coefficient is** +1**. If the movement of one pair **doesn't tell anything** about the movement of the other pair then there is a **zero** correlation between these pairs. If two pairs move in **exactly opposite directions** then their correlation coefficient is **-1**. Correlations are also divided into four groups in accordance with their strength. For easy viewing all correlations in the following table are coloured to show their strength, as is noted below:

**Weak**(White): the absolute value of the correlation coefficient doesn't exceed 0,3 (i.e. it can be anything from -0,3 to +0,3).**Medium**(Grey): the absolute value of the coefficient is greater than 0,3 but less than 0,5.**Strong**(Black): the absolute value of the coefficient is bigger than 0,5 but smaller than 0,8.**High**(Red): the absolute value of the correlation coefficient is equal to or greater than 0,8.

The correlation coefficients are calculated using the daily closing prices seen over the last 40 trading days (shorter-term) and the last 120 trading days (longer term). These two periods have been chosen from among two hundred possible correlation periods **based on how well their correlation coefficients correspond with the daily price fluctuations**. As you can see from correlation simulator (Please note: This calculator requires that you have Flash installed and Javascript enabled in your browser) the actual correlation will usually diverge stronger from the target value when it is calculated for shorter time periods. This makes it important to check the short-term correlations against the longer-term correlations, which is done in the REL table below. The REL (from "reliability") table compares the short-term and the long-term correlations and shows the average of both coefficients when they stay close for both time periods. It is believed that if the short-term and the long-term correlation coefficients agree, the correlation is more reliable - **more likely to persist in the near future. **You can check how the short-term and the long-term daily correlations change over time for the most commonly traded currency pairs at the trailing correlation page (Please note: The size of this page is 1,3 Mbs and it requires that you have Flash installed and Javascript enabled in your browser).

Correlation can also be defined as the **degree of similarity** (direct similarity when correlation is positive/inverse similarity when correlation is negative) that you can expect to exist between** technical chart patterns** (e.g. trendlines, price patterns, candlesticks and Elliott waves) visible on any two currency pairs' charts. For example, you can expect to see **almost exact mirror image of the trendline **appearing on the daily EUR/USD chart when you look at the same time-scale chart of USD/CHF (because the negative correlation of these pairs is so high). Daily correlation coefficients shown here, therefore, measure the correspondence between intermediate (last 120 days) and the minor (last 40 days) chart patterns visible on the **daily charts **of the currency pairs for which they are calculated. This information will be most useful for position traders (keeping the positions open from one day to a few days) who rely primarily on the daily chart studies. If you wish to calculate correlations for other time periods you can do so in Excel, as is described at the bottom of this page.

## Currency Correlations Table

Read more on Correlation Stability Index.

Note: It is best to diversify into those currency pairs whose correlation is colored in White or Grey (with more caution) on the REL table. You can furthernarrow downthe list of candidates for the diversification by excluding those pairs which have spent the least time being weakly correlated during the last 100 trading days - as is shown on the trailing correlations page (sumof the time percentages that the 40-day and 120-day correlations stayed weak. Please note: The size of this page is 1,3 Mbs and it requires that you have Flash installed and Javascript enabled in your browser). If the correlation is colored in Red for two pairs on the REL table you can use this toselect for the trading only that pair which offers the entry with the highest reward-to-risk ratio between the two.You can also use this information toclarify the technical picture(e.g. Elliott wave counts) of the currency pair that you trade by looking at the chart of the other currency pair(s), with which it is highly correlated.

## Excel Correlations Tutorial

You can calculate **individual correlations** for any two currency pairs and for any time period by going through these steps:

- Select the currency pairs that you wish to analyze. Export the price data for each of these pairs from you forex charts (e.g. Intellicharts) to a file on your computer (the usual format for data export is CSV). Import each file into Excel by going to Data>Import External Data>Import Data and pointing to it. You might need to import the numbers as text and then replace the points with commas so that Excel can work with the prices as numbers.
- Make sure the dates in the imported time series agree for
**each**row (you can skip this step if you are working with only one price feed). - Delete the columns for Open, High and Low. Change the names of the columns with the closing prices to the names of the currency pairs to which they belong.
- Use the CORREL function to calculate the correlation. This function works on two arrays, which will be same-length ranges of closing prices for the two pairs. Simply type into one of the empty cells "=correl(" then press the "fx" button next to the formula bar and select the two ranges. The resultant formula will look like this -=CORREL(A1:A40;B1:B40) and will calculate the value of the correlation coefficient between the pairs for the chosen time period. In this example it will be 40 hours, days or weeks depending on the time scale of the charts being analyzed.

To calculate the **correlation matrix** of any number of pairs repeat the above steps 1 to 3 for each pair. Crop the whole table so that the names of the currency pairs are in the first row and the closing prices are only for the time period that you wish to analyze. Instead of using the CORREL function go to Tools>Data Analysis... and select "Correlation" from the list of analysis tools. Press the button next to the "Input Range" and then highlight the contents of all the columns. Check the mark next to "Labels in First Row". Select the output range by picking a cell to the right of the table. Press "OK".

Note: You might need to install the Data Analysis pack from your Office Installation CD if it is not loaded by default. To install it go to Tools>Add-Ins...., then select Analysis ToolPak and hit OK to start creating your currency correlations matrix.