The idea that you can recommend a risk-based portfolio on the basis of a single metric is about as far-fetched as judging someone’s health solely on their body mass index (BMI). In investment risk profiling, as in health, to think that you can use a single figure to provide all the answers is both misguided and dangerous.
Matching something as complex as a human being with something as complex as a risk-rated investment portfolio is, not surprisingly, a complex process. Risk-profiling tools are a helpful way to guide advisers towards suitable recommendations for their clients. But it is important to be aware of their limitations, and advisers must understand that the methodology behind risk-profiling tools varies.
The first point advisers need to bear in mind is that there is no genuine industry standard way of approaching risk. Yes, there is the European Securities and Markets Authority (ESMA) benchmark, the synthetic risk and reward indicator (SRRI), which is required to be displayed on all key investor information documents (KIIDs) for Ucits funds. The SRRI is based on the volatility of the fund using weekly or monthly returns over the previous five years.
I won’t dwell on the fact that this approach means any volatility that predates the massive upheavals between 2008 and 2010 is now being overlooked. But it is worth remembering ESMA requires funds to be rated on a scale from 1 to 7. Not only does nobody in the UK really use this methodology, we don’t even use the ESMA numbering system, with most tools evaluating advisers’ clients on a risk scale of either 1 to 5 or 1 to 10.
So already we have the challenge involved in translating between an odd and an even numbering system - what might be a 4 as far as ESMA is concerned could be either a 5 or a 6 on a scale of 1 to 10.
Advisers also have to bear in mind that different systems have different ways of mapping the way they assess where the individual sits on their risk profile, and then how that is translated through to a portfolio. So an investor at the high risk end of the spectrum, be they a 5, 7 or 10, depending on the methodology used, could be pointed towards different asset recommendations because the interpretation of ‘very high risk’ will vary from system to system.
Going beyond whatever system is used, it is crucial to remember that there is more to the picture than a single number. Going back to the BMI analogy, if you take France’s Uini Atonio, the biggest player at this year’s rugby world cup. At 22 stone 12 lb and 6ft 6 inches, he has a BMI of 36.9, way beyond the 18 to 25 bracket deemed to be healthy, which puts him squarely in the obese category. It is clearly ridiculous that a sportsman operating at his level should be judged as overweight. Mo Farah on the other hand clearly looks fit as a fiddle and not surprisingly comes well within the healthy range with his 21.1 BMI.
Other risks exist that cannot be measured successfully with a single number - counterparty risk is an obvious example, as are liquidity risk, the age and profession of the investor, his or her objectives and the term over which the investment is being made. All these factors, as well as the number generated by the risk profiling tool, will influence what is suitable for them.
Counterparty risk is a particularly complex factor, and does not lend itself easily to measurement and gradation. You can look at what has happened to a particular class of assets in the past, but comparing that with what will happen in the future is very difficult. A type of investment that may have been very risky in the past may be less risky in the future if, for example, the way it is regulated has been beefed up following a misselling scandal. Similarly the introduction of charge caps on a type of product could make it riskier, because the provider has to operate with less margin.
However good we think our solutions are, there is much more to risk than the number created by a risk-profiling tool. Yes these tools are helpful, but they do not tell us the entire story.
Andrew Storey is technical sales director at eValue
