Is Correlation Analysis Important?
Correlation coefficient (CC) is a statistical measure, or a linear regression performed on two securities (i.e., stock vs stock, stock vs sector, stock vs market, sector vs sector, etc.). In its most simplistic form, it tells how closely one security is related to the other security over a designated time period. CC typically oscillates between -1 and +1. A number that is positive (i.e., greater than 0.50) indicates a positive correlation. That is, both securities tend to move in the same direction (either up or down). A number that is negative suggests a negative correlation where the two securities tend to move in opposite direction from one another.
Although correlation analysis does not help investors predict the future returns of stocks, it can be an effective tool when used in conjunction with other technical indicators to help predict the extent to which two securities move in relationship to each other. Understanding the relationships between securities, sectors, markets are an essential component to portfolio construction, diversification, asset allocation, and risk management strategies.
Correlation analysis can identify non-correlated securities, which is the basis for developing a balanced and diversified portfolio. Understanding the correlations between securities and between asset returns are also critical to evaluating the impact and risk exposure of these new securities or assets on the overall portfolio. Once an investor understands the relationships between securities, the investor can select those that, when combined, help to reduce risk. Remember, correlations between two securities tends to fluctuate, especially during volatile market conditions. To create a true diversified portfolio, one needs to understand the historical relationship between securities and most important, understand the future relationships between these securities. The primary purpose as why an investor would want to create a portfolio with low correlations among the different assets/securities is to reduce the volatility of the overall portfolio returns, which in turn can help to reduce the portfolio’s standard deviation of return.
Enclosed below are 10-year correlation charts on major financial assets including Equities (US large cap stocks - SPX), Fixed Income (US 10-year Treasury Bond - USB), Currency (US Dollar Index – USD), Commodities (CRB index - CRB), and Gold (CME Gold contract futures – GOLD). These major assets have recorded positive correlations against US stocks (SPX). The lone exception is the negative correlation (-0.32) between CRB Index and SPX over the past year. Gold, being a hard asset, has a strong historical correlation to commodities (CRB Index) essentially trading in the same direction. However, over the past year Gold is positively correlated (0.55) to SPX. The pertinent question then becomes, will these relationships continue into the next year? What happens, if any, if these relationships begin to diverge? Does this then warn of the next economic/stock market cycle?
Also attached are 10-year correlation charts of the 11 S&P 500 sectors in relationship to SPX. As can be expected, S&P sectors tend to be positively correlated to SPX. This is reasonable as these S&P sectors are comprised of stocks, and most stocks tend be positively correlated with one another. However, it important to understand that some S&P sectors are more positively correlated to each other than others at various phases of an economic/stock market cycle. S&P Technology and Consumer Discretionary sectors have traded in sympathy with one another essentially moving step-for-step in the same direction as SPX as evidenced by their recent high correlation coefficients of 0.97 and 0.95, respectively. On the other hand, commodities sector such as S&P Materials, tend to have a much lower correlation (0.46) to SPX. We find it unusual that over the past 2 years, another commodity related sector, S&P Energy, is currently trading at a negative correlation (-0.28) to SPX. Will this negative correlation trend continue as it retests its late-2015 lows before reaching a climatic sell off bottom or does this trend signaling long-term structural change?