As the idiom goes, “Don’t put all your eggs in one basket”. It is common knowledge that diversification is very important when it comes to investing.
However, the questions are:
How to diversify?
How diversified are you?
How do you measure diversification?
Google “How to diversify” and there are countless articles out there “teaching” people how to diversify. In my opinion, many articles are either too vague or too qualitative. I come from an engineering background hence I prefer to quantify things. I prefer concrete numbers.
First, let us take a step back.
Why do we need to diversify?
It is because we want to reduce the risk of investing.
There are fundamentally two types of risk in investing: systematic and unsystematic. Systematic risk is the risk that affects the entire stock market. Unsystematic risk is the risk that is specific to a company or industry.
Therefore, in order to quantify how diversified a portfolio is, we need to quantify the systematic and unsystematic risk that the portfolio is exposed to. Beta is often used as a measurement for systematic risk because it measures how a portfolio would react to market movements. The closer the beta is to zero, the more insensitive it is to market movements.
As for unsystematic risk, one metric that I like to use is Value at Risk (VaR), as it tells you how much (in percentage) you can expect to lose in a very bad month. Value at risk is high when stocks in the portfolio are highly correlated. In other words, value at risk can tell us how diversified the portfolio is across unrelated companies/industries.
Start quantifying how diversified your portfolio using beta and value at risk today.
p.s. Beta and VaR are automatically calculated for your portfolio if you manage your portfolio in StocksCafe (it does not cost anything to use it). In fact, there is also an additional tool, Risk Manager, which goes an extra step and helps you calculate the Beta and VaR of your new portfolio if you were to buy a new stock. You can sign up for StocksCafe here.
A friend recently suggested to me that I should also consider using expected shortfall to measure risk. It is because while value at risk is fine for measuring risk under normal situations, it does not handle stress situations such as a financial crisis well. Hence, I have decided to add expected shortfall into many computations in SGXcafe and display them.