Chill Out, it’s Just a Stress Test

So, what is a stress test and what form should it take? Stress testing is a simulation-based tool that puts “stress scenarios” on a portfolio to see how it reacts in that given situation. It carries a high level of complexity with hundreds of data points being pulled in all different directions. These “stress scenarios” are then run thousands of times with a Monte Carlo simulation, which slightly changes each variable by a defined probability distribution to find an average. These simulations require sophisticated software, one example is the Bloomberg’s Portfolio System.

The scenarios are limited only by the imagination of the manager but typically focus on historical events that have had sizeable impacts on financial markets. A good example would be running a current portfolio through the GFC market collapse to see how it fares.

While limitless in its options, a robust stress test should have at least the following attributes:

• Relevance – the stress test scenario should be relevant to the type of portfolio.  For instance, there is no point testing Emerging Market interest rate movements on a Developed Market equity portfolio
• Granularity – the stresses must be applied to each individual asset in the portfolio, rather than the aggregate portfolio
• Understandable – The stress scenario should be meaningful to the investor and the output explainable
• Completeness – the test must include all assets in the portfolio