Working in partnership with you

Does VaR help decision making?

There was a lot of controversy around Video Assisted Refereeing (“VAR”) at this summer’s FIFA World Cup and that has continued into this season’s Premier League. Some commentators were big fans, but others have been slow to react to the new era of technology and the practicality it offers. Used appropriately, VAR should increase confidence in the decision-making process, culminating in potentially game changing decisions!

In the pensions industry we were early adopters of our own VaR analysis. I am of course referring to Value at Risk (“VaR”). This VaR assesses the risk of potential losses to a pension scheme, over a given time period. Much like the hotly debated Video Assisted Refereeing, VaR can assist trustees in their investment strategy decision-making process. VaR is in fact a longstanding concept within the financial game, with perhaps the only debate being around which method of VaR to use.

There are two options for trustees to consider; both have potential advantages and drawbacks.

1. Variance-Covariance

A simplified approach to VaR such as the ‘variance-covariance’ method gives trustees the ability to quantify potential losses to a scheme in normal circumstances. This assumption is found at the heart of many financial theories – expected asset returns are normally distributed around their mean. Here we are assuming history is representative of the future, which of course is an unrealistic assumption to make.

Figure 1: A normal distribution curve centred around the mean (historic) return. Note standard deviation () is a measure of dispersion around the mean return, 68.2% of returns will fall between +/- 1 standard deviation of the mean based on the assumption of normality.

Using a ‘normal distribution’ (as shown above) removes the possibility of more positive, or more negative, events occurring (i.e. ‘fat tails’, or extreme events).


2. Monte Carlo

A more technical approach is the Monte Carlo simulation method, which creates a distribution of future outcomes based on assumptions around different risk factors. This approach runs up to 10,000 different economic scenarios and produces a more familiar fan chart of potential outcomes (see chart below). Perhaps the downside of this approach is that the output is limited to the range of economic scenarios input by the user.

Figure 2: Shows how stochastic analysis can generate and project thousands of different scenarios for a pension scheme’s assets and liabilities.

Has VaR improved decision making?

Whilst VaR is certainly not the only measurement of risk that trustees may wish to consider, it has become the most quoted measure, in part because it is the preferred measure of the Pensions Regulator.

Whichever method is used to calculate VaR, we should not forget that the purpose of any good risk measurement technique is to help decision makers understand the impact of adverse events. VaR aims to quantify the potential impact for a scheme as a single monetary value, over a given period of time, for a given level of confidence.

At Quantum we believe VaR helps trustees understand the risks within their investment strategy and the affordability of those risks to the sponsor. A weak covenant, for example, would mean the sponsor is less likely to be able to afford a significant fall in funding level – VaR analysis can help trustees assess the ability of the sponsor to continue to support the scheme. Working with trustees to understand VaR levels enables Quantum to formulate an investment strategy that has an acceptable level of risk for the scheme.

At Quantum we use the more sophisticated VaR method (Monte Carlo simulation). This statistical method, along with a qualitative overlay, means we can confidently report risk to our clients by examining future potential impacts on a scheme’s assets, liabilities, and overall funding level.


Stefano Carnevale, Investment Analyst at Quantum