Correlations

Correlation, in trading, is a statistic process that measures the degree to which two or more securities move in relation to each other.  The most common correlation coefficient is the Pearson Correlation Coefficient, which assumes that there is linear relationships between data points.

Correlations are that factor the trader needs to be aware of, when constructing a Portfolio Management strategy. Following the popular saying, ‘don’t keep all your eggs in the same basket’, the lack of correlation is vital for an investor. One can prevent correlation into a portfolio by diversify the strategies, the stocks, the markets, etc… Before establishing the correlations, one needs to decide on the time frame afferent to the trade that will open.

Let’s consider a portfolio formed from SPY(Standard & Poor 2500 ETF), TLT (20 year bond ETF), USO ( West Texas Intermediate Light Sweet Crude Oil ETF), FXE(price of the euro in usd ETF) and GLD(Gold ETF) and test first for 3 month correlation, then 6 months and 1 year after:

3 month correlation
6 month correlation
12 month correlation

As we can see from the above, correlations among symbols can change with the time and that’s why we need to know the time frame of the new trade. For example: the correlation between GLD, TLT and SPY are negative for the last 3 months, but they were positive before. A very important matter when we want to construct a diversified portfolio…

Peace!