Using the Correlation Analysis Tool

The Correlation Analysis Tool is based on one of the simplest yet most powerful concepts in understanding diversification: the Correlation Factor.  The Correlation Factor is a unitless measure that indicates the degrees of historical price dependency between two assets.  The possible values for a correlation factor range between -1 to +1.  A value of +1 indicates a perfect positive correlation, meaning the two assets always move together, in the same direction.  A value of -1 indicates a perfect negative correlation, meaning the two assets always move in precisely opposite directions.  A value of 0 indicates no correlation whatsoever.  Each of these three values is an extreme, and it is highly probable that for any pair of assets, the actual correlation usually falls somewhere in between. A correlation factor of 0.9 represents a very strong correlation, while a correlation factor of 0.2 represents a weak positive correlation.  

The goal in achieving sound diversification is to build a portfolio of solid performing assets that have weaker correlation factors with each other.  This reduces the volatility and risk of the overall portfolio.     To understand how it does so, see the article on the Importance of Diversification.  While looking for weaker correlations, you also should strive to select individaul assets that are expected to have a good long-term track record, whether they are mutual funds, ETFs or individual stocks.  It doesn't make sense to pick an asset that is weakly correlated with others but delivers poor returns.

To use the Correlation Tool to determine the historical Correlation Factors for the assets in a portfolio, perform the following steps:

• Select the Correlation Analysis link.
• Add your list of securities to the portfolio, either by selecting the security in the asset selection tree, or by using the search feature.  Select the “Add” button or “Remove” button to add or remove each security from the list.
• Once you have added the securities, select the “Next” button.  
• Select the time range to analyze historical data to calculate Correlation Factors.  It is important to select a time range long enough to provide a good representation of correlations under different market conditions (up and down).  Monthly closing prices are used in the calculation.
• Select the “Calculate” button.
• At this point, the tool will calculate the Correlation Factors for all pairs of assets in your portfolio.  The results will be presented to you in table format.  Below is an example.  The first table shows the correlation factors between each pair of assets.  To see the Correlation Factor between any pair of assets, simply pick the row for one and column for the other and find the sell that intersects.  You can ignore the diagonal identity cells with 1.00, since those indicate the correlation of an asset with itself (always 1.00).  Note how this portfolio has 5 ETFs, and that over half of the correlation factors are above 0.90.  This is generally a warning sign, since too many correlation factors close to 1.0 are an indication of a lack of diversification.  The ETF that offers the best diversification to this portfolio is the SPDR 500 fund, since it has historically had the lowest correlation factors with the other assets (ranging from 0.58 to 0.69). 

 

It is very important to also study the second table.  It shows the expected return and expected standard deviation of each asset in the portfolio.  This is strictly based on past performance over the historical time period you selected.  It is important to look at this table in conjunction with the first table, since what you want to look for is better performing assets that have low correlation factors with respect to each other.  At this point, you may want to make adjustments to the portfolio by going back to the first page of the Correlation Analysis and removing or adding assets to see how they contribute to better diversifying your portfolio.  You can experiment with a virtually unlimited variety of hypothetical portfolios.

Finally, be aware that both tables are based on past price history behavior of each asset, and that past history is not always an indicator of future behavior.  Market conditions can change, as can managers of mutual funds and companies.  This is why it is important to select a time period for analysis that you feel will be a good representation or model for the future.  In addition, you should always do your homework on each individual mutual fund, ETF or stock by visiting sites such as Morningstar or Yahoo Finance, which provide tools for researching individual securities.

Once you are done building and analyzing your portfolio, you can choose to save it if you wish.  To do this, select the “Save portfolio” button.  If you are not already logged-in, you will be prompted to log-in (or register) so that the portfolio can be associated with your user-name.  You can now reference this portfolio in My Portfolios, and track its performance or do further analysis on it in the future.