We can predict investment risk. So why don’t we?

 
Adam French
ITALY-NATURE-ALPS-ANIMAL
Technology, and the work of academics like Mandelbrot, mean fixed weight asset allocation is unnecessary and outdated (Source: Getty)

It doesn't take a genius to notice that we are in deeply uncertain political and economic times. The events of 2016 will have repercussions for many years, and times have never been more challenging for investment managers trying to navigate these unpredictable waters.

While many will be sticking to their rule books – typically developed in the 1950s – there is a better way of delivering smooth returns in these turbulent times.

Traditionally, investment managers build portfolios for their clients using fixed weights for each asset class. For example, a typical 60:40 (60 per cent stocks, 40 per cent bonds) portfolio would be given to a medium-risk client. The portfolio is then rebalanced periodically to keep it aligned to those original weights.

However, while this portfolio might deliver a return similar to what the client expected at the end of their investment, in the short term, it is prone to suffer from unexpectedly high losses due to significant fluctuations in the markets.

It is usually during these times that investors will get cold feet, sell their investments and sit on the sidelines for years – not the long-term plan they had in mind. It is one of the reasons why investors typically earn a lower return than the assets they invest into; they jump in and out at the wrong times.

Managers can learn a lot from the financial markets research conducted over the past two decades, but too few of them have accepted that the traditional way of building portfolios is now outdated.

Even before the 2008 global financial crisis, there were periods of unusually high volatility during which investors were punished for holding risky assets instead of being rewarded for that risk. However, the quantitative tools and the computational power required to identify these periods of high risk simply weren’t around.

Now, we have access to the right financial models and software, so it would be almost absurd not to make use of them. What investor would decline smoother returns? They simply need to get out of an asset class once it enters a period of turmoil – which sooner or later typically leads to losses – then reinvest once it’s back on track to deliver its usual, average returns. Avoiding the losses the client would be exposed to during these turbulent times helps them achieve a better return on their investment.

An adaptive asset allocation enables the risk of the portfolio to be maintained at the level expected by the client, in all market conditions. The client never takes more or less risk than they originally signed up to.

This is only possible because risk can be predicted. While it is impossible to say how a particular share price (direction and magnitude) will develop based on today’s information, the situation is different when it comes to predicting only the magnitude of moves. If the markets are very volatile today, i.e moving a lot, there is a more than a 50 per cent chance that they will be volatile again tomorrow. It is this behaviour of markets that allows us to predict the level of risk of various potential portfolio allocations.

Scientist and mathematician Benoit Mandelbrot was the first to notice this behaviour, known as volatility clustering. He noted that “large changes tend to be followed by large changes, of either sign, and small changes tend to be followed by small changes”.

In 2003, scientist Robert Engle was awarded a Nobel Memorial Prize for his work on developing this theory into models that can be used to forecast risk.

Using the work of these giants of academia allows for periods of high market volatility to be anticipated better, and for the downside risk in portfolios to be controlled.

By using market risk instead of price direction, managers that have access to the right quantitative tools can dynamically adjust portfolios according to individual risk appetite. Portfolio risk remains under control while market risk fluctuates; this is what we call dynamic risk management. Exposure to the potentially punishing temporary excess risk is far less likely.

City A.M.'s opinion pages are a place for thought-provoking views and debate. These views are not necessarily shared by City A.M.

Related articles