Hidden cost of volatility

In fact, stock market returns during 2016 and the beginning of 2017 have embodied the phrase ‘climbing a wall of worry’. 

Such periods are halcyon days for advisers – lots to talk about at client review time, whilst still being able to point to great returns. 

There are obvious, and much discussed, short-term risks to this continuing: Brexit discussions, elections in the UK and Europe and populist temptations to tamper with trade. But what about subtler longer-term risks? 

What hidden costs do advisers need to aware of to be able to help client’s meet their ultimate financial goals?

A key part of my investment philosophy is a focus on trying to achieve the highest total return for our clients per unit of risk they take on. 

This risk-adjusted return mindset is really just a cost/benefit framework and it informs our view on most of the big questions in investments by allowing us to make sensible trade-offs.

However, whilst some costs are quite easy to see, such as a fund’s ongoing charges figure (OCF), there remain many hidden costs which investors need to look out for which will influence the trade-off decisions we make. One of the most shadowy of these hidden costs is volatility itself.

Volatility drag is the statistical effect whereby volatility reduces compound returns over time below the expected return over a single year.

This effect is most clearly demonstrated by the example of an investor who has an equal chance of earning 10 per cent or minus 10 per cent in a year. 

On average the investor should be flat over that year but over many years they would tend to lose money. 

For example, if they start with £100 and gain 10 per cent in the first year and lose 10 per cent in year two (minus £11, or 10 per cent of the £110 they had at the end of year one), rather than being flat over two years they actually end up with only £99, a loss of 1 per cent overall. 

And the effect only gets worse the more volatile/extreme the returns. 

A similar bet with 20 per cent or minus 20 per cent payoffs would leave the investor down 4 per cent after both a gain and a loss.

There is a formula for this cost: volatility drag per year is roughly the volatility squared divided by two, as can be seen in the simplified examples above where the effect over two years is just the volatility squared.

If volatility is a cost, then multi-year forecast returns should be less than single year forecasts. And because financial planning is a multi-year endeavour, advisers should be using the lower multi-year forecasts.

This has implications for the ‘efficient frontiers’ we so love in investments, whereby we assume we can find the optimum portfolio of assets for each risk level and we plot their increasing expected returns vs their increasing volatilities to get a nice curve.

The problem is that these expected returns are, in most cases, single year forecasts.

A multi-year version of the efficient frontier sits below that of the single-year version and extremely risky/volatile portfolios are much less attractive than they first appear.

And whilst taking more risk still leads to higher forecast multi-year returns over most of the curve, at extreme volatility levels taking more volatility can actually decrease forecast returns – an inversion of the usual relationship.

This has pretty wide reaching implications:

* Advisers often use over-optimistic assumptions for cash flow modelling

* Investment managers don’t diversify enough and may run too much volatility in high risk portfolios

* Clients sometimes take the wrong amount of risk and/or are disappointed by the long-term returns they receive.

All-in-all, the fact that volatility negatively influences long-term returns as well as being a component of risk is pernicious and, unfortunately, inescapable.

Awareness of the issue is the only defence as it allows the effect to be carefully reflected in financial tools and financial planning and investment decisions.

Nic Spicer is portfolio manager of PortfolioMetrix