Correlation of Returns
Click any correlation number for a time-series chart option.
About CorrelationCorrelation is a measure of the strength and direction of the relationship between two price series. In other words, it measures to what extent the prices of two securities move together. The measure ranges from -1.0 to +1.0. A value of +1.0 indicates the funds move up and down in a near perfect relationship. A value of -1.0 indicates they move inverse to one another. A value near 0 indicates no recognizable relationship between the price movements.
So how is correlation used in investing? Primarily to insure portfolio diversification. If your goal is to hold 5 ETF's for diversification, it does little good if 3 of them typically move together. By monitoring the correlations among your holdings and your potential new buys you can better maintain a diversified portfolio and thus reduce your risk. And remember, diversification is one of the few strategies that receives near unanimous support among investment professionals.
Correlations are typically presented in a matrix like shown above. The default symbols shown are from the ETF Market View segment on the home page, and represent a rather diverse sampling of world equities. You read the matrix by finding the intersection of one symbol on top and the other symbol on the side. You will generally find the relationship among the U.S. equity index funds relatively strong, where commodity and bond correlation with the equity indexes is usually low. Gold and bonds often have a very low or negative correlation with equities, because investors flock to these markets during times of crisis.
Correlations on this site are measured using daily returns for the past 126 days, or about 6 months. This would be considered a short term measurement by many and was chosen because we want to focus on what relationships exist today, not two or three years ago when market conditions might have been different. It is important to understand that correlations are not static and they will change dramatically with market conditions.
Correlation data is calculated for what we consider the actively traded ETF's. That means they average about a million dollars or more of daily trading and, in sum, account for almost 99.9% of all ETF trading activity when measured in dollar terms. This data is made available in several ways on the site, as summarized below:
1) On the price chart pages you will find the 10 most correlated and 10 least correlated funds tabulated near the bottom of the page. This lets you quickly see what funds are similar and what inversely related funds might be available.
2) There is also a complete report for a specific ETF, showing all calculated correlations ranked from highest to lowest. If the top and bottom 10 given on the price chart do not answer your needs this report will allow you to search deeper in the list to find the fund relationships you are looking for. This list is also available from the price chart
3) You can view a matrix like the one above for a group of funds. This is an option directly from most Performance Pages on the site, or you can enter a group of symbols in the form above.
Monitoring fund correlations and incorporating them into one's fund
selection process is a good way to manage portfolio diversification.
The screener available on this site is a good tool for identifying
potentially profitable trades but many times the funds listed are
highly correlated. By adding a correlation review step one can
hopefully reduce risk.