Correlation of assets measures how an asset class moves in comparison to another asset class.
When two assets move in the same direction together, they are considered to be highly correlated. When these asset classes move in the opposite direction, they would be negatively correlated.
A positive correlation would mean that when one asset class return increases the other asset class return also increases. The closer to ‘1’ correlation, the better correlated they are. A correlation of ‘1’ would mean that the assets have a perfect positive correlation.
A negative correlation would mean when both asset class returns move in the opposite direction. That is when one asset class return increases the other decreases. A correlation of ‘0’ means no correlation and a ‘-1’ indicates a perfect negative correlation.
For example, we have considered different indices as a proxy to different asset classes. The correlation matrix for the same is given below.
We can observe from the matrix that gold is uncorrelated or minorly negatively correlated to mostly all other indices. All the equity indices are highly correlated with each other.
Given above is the representative of the rolling returns for Sensex index, gold price and Crisil Gilt index.
The correlation can be observed in this chart as well. From November 2005 to November 2007 gold returns had a high correlation with equity, as they moved in the same direction. While gold had a negative correlation with Sensex from June 2011 till date.
Adding negatively correlated asset like gold to other asset class reduces the risk considerably. The best asset allocation comes from combining negatively correlated assets.
Low correlation reduces the volatility of a portfolio without necessarily affecting the expected level of return.
Portfolio construction and diversification are heavily influenced by the correlation between different assets.