Correlation and Causation

Amar Pandit , CFA , CFP

What comes to your mind when you think of Correlation and Causation?

They seem jargonish and boring right! I think so too.

Correlation is a statistic used in the investment space extensively in relation to portfolio management. It is used to measure the relationship between 2 asset classes or 2 securities or even between a portfolio and a benchmark (Sensex or S&P 500).

It is often used in diversifying portfolios and reducing the risk of a portfolio. Correlation is also used by many to demonstrate how intelligent their research and process of portfolio construction is.

I won’t bore you with the formula of Correlation, but the Correlation Coefficient’s values range from -1 to +1.

A +1 Correlation Coefficient means that the 2 variables or portfolios are perfectly correlated and that they move in the same direction. A -1 means the 2 asset classes will move in opposite directions and a 0 means that there is no linear relationship at all. Negative correlation is used in reducing the risk of a portfolio by investing in uncorrelated or low correlated assets. This is the principle of diversification.

At a simplistic level this is fine, but the problem happens when investors and even financial professionals confuse correlation with causation. Correlation might show association, but it never shows causation.

Confusing. Let me give a simple example that we will all relate to.

Umbrellas and Rain are Correlated but do Umbrellas cause Rain?

Never right.

Buying an Umbrella or carrying one will never cause rain. On the other hand, some of my stupid data observations have made me believe that the moment I step out without an umbrella (during monsoons), it starts raining ☺. Well, there is absolutely no correlation in my observation and certainly no causation.