Risk is not a number

There is no single correct way to measure risk. Even defining risk is difficult. The definitions are different, can be inconsistent with each other, and are often specific to individuals. This applies in life, as well as in finance.

Skydiving, for example, has a very, very small chance of going wrong. But when it goes wrong, the results are terminal. Regular commuting by bike has a much higher chance of going wrong than skydiving – but is far less likely to be fatal.

A keen skydiver might believe that commuting by bike in London is crazy, while a regular cyclist might vow never to jump out of a plane. Neither of them are wrong about how risky the activities are, they just have different perspectives. There’s no single answer about which is riskier.

The same is true in investing. There’s no single answer. There’s no one number that can completely describe the risk of a portfolio. But the broader financial industry doesn’t always reflect that. And we often see volatility and risk used interchangeably, when they aren’t the same thing.

Volatility and risk

Most of the time volatility is a pretty good indicator of risk. Very simply, it describes the range of returns of an asset around the average.

Imagine that two stocks have the same return of 5% in a year. Stock A has a volatility of 2%, and Stock B has a volatility of 10%. This tells you that the price of Stock B has moved around a lot more than Stock A.

With a little maths (if returns are ‘normally’ distributed, as statisticians say), it tells you even more – that 95% of the time, returns from Stock A were between 1% and 9%, while Stock B returns were in the -15% to 25% range. That is useful information, but doesn’t give you the whole picture.

There are lots of potential limitations to using volatility to measure risk. We’ve picked out three common ones below.

Tail risks

Volatility is pretty good at capturing the big averages – as with the ‘95% of the time’ above. But that other 5% is pretty important. The ‘tails’ of the distribution can’t be ignored.

Imagine a lottery where you win one million pounds 95% of the time. Would you play? Or would you ask what happens if you don’t win? The outcome in that 5% of the time might matter – if you’re dragged off to prison for life the lottery would look a lot less attractive. The painful, unpredictable, and often disproportionate outcomes in that 5% is what investors call ‘fat tails’.

The COVID-19 crisis is the perfect example of this. Coming into 2020, most economists’ predictions for growth this year had been centred on the long-term average. After all, it worked the year before. And the year before that. And the year bef… you get the picture. There was the odd economist who thought growth was slow, but not even the uber-bears were predicting a recession. Fast forward three months and we are now experiencing the worst recession in almost 100 years.

We use stress tests to simulate the tails in finance – fat or otherwise. We think about nasty scenarios from history and apply them to portfolios. These include the Dotcom bust, the Global Financial Crisis, the Brexit referendum. We also come up with our own scenarios – where the Risk Management Team and the Investment Management Team make the world look as bad as possible, and test how our portfolios would cope. Volatility numbers won’t tell us that.

Measurement

Volatility is a statistical calculation. It simply reflects the data fed into the equation. A couple of things can easily shift the final number around – frequency and time horizon.

Frequency is how often the data is measured. Usually, the more often you look at something, the higher its measured volatility. A good analogy is how we respond to the news. Read the headlines every day and you’ll think the world is falling apart most of the time. Check in every January and you’d get a very different picture.

Time horizon is just as important – the longer a period you measure over, the more data you have, so the more likely you are to have a representative sample. If you wanted to show someone how good you are at your job, would you want them to see a random hungover Friday or a whole year of your efforts?

The FTSE 100 has a volatility of 24% when measured daily over the last year, but a volatility of 12% when measured monthly over 10 years. As with the skydiver and the cyclist, neither is right or wrong. We use multiple time horizons and frequencies when analysing our portfolios – looking at the numbers side-by-side and with a bucket of salt, because what happened in the past might not be relevant today.

Liquidity

You can keep measured volatility down by recording prices only now and then. Think about houses. Most of us are convinced that houses are a stable investment because we only look at the value of our homes once or twice a year, if that. And we sell our homes even less often.

If an estate agent provided estimates of how much our houses were worth every five minutes, then house prices wouldn’t seem so stable. In fact, they would jump around wildly.

There are a number of asset classes where prices are very stable normally. But in periods of stress, no one wants to trade, so prices don’t move… until they crash.

Volatility won’t capture that. But it’s a key risk for some asset classes. We’ve seen it recently in physical property funds – whether it’s Brexit or COVID-19, investors are suddenly locked in. So prices don’t move for a long time – but that’s not the kind of low volatility you want.

At 7IM, we watch portfolio liquidity closely – the Risk Management Team monitors the prices and liquidity of over 600 securities every day. But more importantly, we assess liquidity before anything even enters the portfolios. This applies to our Strategic Asset Allocation, as well as any tactical positions we might take.

Not just one number

The problem isn’t with volatility. The problem is that risk is too complicated to be captured with just one indicator. There’s too much nuance and too many shades of grey. It’s like rating the best football player in the world based only on goals scored. Messi wins. But while goals are a good start, you need to dig deeper. Using broader measures, a defender like Virgil van Dijk might be just as good.

The Risk Management Team and the Investment Management Team have lots of different people looking at our portfolios in many different ways – we discuss and analyse portfolios from as many different angles as possible.

At 7IM, we certainly didn’t see the coronavirus coming, nor did we predict that the rest of world would soon follow the Chinese model of lockdown. But constructing resilient portfolios is about preparation, not prediction. Acknowledging that risk isn’t a single number is a vital step.

So, while our clients spend their time taking the kinds of risk they enjoy – whether it’s cycling to work or, erm, skydiving at the weekends – they can rely on us to manage the risks of their portfolios.