The treatment and assessment of investment risk is fundamental in determining the suitability of investment products for specific clients and, perhaps as importantly, the outcome of client complaints.
The FSA, Financial Omb-udsman and the FSCS appear to consider that a low-risk inv-estment is one where the scale of potential losses is limited. While they would not necessarily agree, their actions generally show that it is such a definition that holds weight in their thinking and decision-making.
However, not only is this definition incorrect in both formal and common usage but is also largely meaningless. An alternative definition, as shown, is not just more relevant but wholly so. Using a correct and sustainable definition has a material effect upon the concept of misselling.
Forget for the moment the approach of the regulator. In common usage, the risk of an event occurring is the measure of its likelihood. The scale of the event is not relevant.
For example, it is well known that the safest form of transport is air travel. We accept that flying is a low-risk form of transport. In other words, the chance of death or injury from flying is very low. This is despite the fact that, if your plane does crash, the likelihood of death is very high.
So, while the risk of an extreme outcome is high if the plane crashes, nonetheless, the risk from air travel is low. In other words, the risk of dying is only of concern if the plane crashes in the first place. To determine the true risk, one assesses both the likelihood of the plane crashing and the likely damage caused if it does. Despite the high possibility of death in a crash, the overall risk remains very low because planes crash very rarely.
It would seem, therefore, that the FSA might be misinterpreting risk. In the cases of precipice bonds and split-capital investment trusts, the regulator has concentrated on the fact that many investors lost a large proportion of their investment (that is, died, to use the aeroplane example) rather than considered whether it was likely that these losses would be incurred.
It seems, at first sight, that the FSA is in error. However, in its role as “protector of the ignorant”, it may be reasonable for the FSA to suggest that private investors would consider a large proportionate loss, no matter how it occurred, as evidence of a higher-risk investment.
Returning to the flying example, a high proportionate loss on the investment is equivalent to severe injury or death. Yet it has already been shown that the overall risk of flying was low.
In determining the risk of any event – whether death or suffering investment losses – one cannot select individual parts of the equation. It is the effect of all the parts in combination that clearly matters.
It is certainly the case that many investments, commonly agreed to be low risk, run the risk of total or significant loss. Building societies and insurance companies can fail and even government bonds may default in extreme times – a fact not lost on Latin American investors.
The fact is that these investments are, almost universally, considered low risk because the likelihood of failure is low despite the possibility of total loss.
When considered in this way, we can accept that the true investment risk of a product is properly reflected neither in the potential scale of losses nor in the likelihood of losses per se but in the combined effect of the two.
Application to investments
So, it seems the FSA and the Ombudsman are flawed in their thinking if they deter-mine that an investment cannot be low risk merely because investors have lost a substantial proportion of their investment.
It is entirely relevant to consider what was the likelihood of investors losing money. Not whether they have done or how much.
Another example may help. More people are concerned about dying in a plane crash than from a meteor impact. In modern times, there are no known cases of anyone, anywhere, having died from a meteor strike. Yet, statistically, there is a far greater chance of this happening to you than from dying in a plane crash.
In the case of a meteor impact, the risks are reversed, to a degree. The risk of a large meteor reaching the ground is very low. Yet, if one does, the likelihood of death is much greater.
While, overall, the risk of death by meteor remains very low and much lower than that from a plane crash, the risk if it does happen is considerably greater due to the likely widespread destruction.
So, why is it that we worry more about a plane crash than a meteor impact? It is because we have experience of air crashes (at least through the media) and have never seen the effect of a meteor strike. In other words, people worry about the likelihood of an event occurring much more than its impact if it does occur.
Therefore, it is shown that, while the real risk is a combination of the likelihood of an event and its scale, it is usual for people to be concerned only about the likelihood of the event occurring.
This is in complete contrast to the stance taken by the regulator, which has demonstrated a fondness for the only remaining and flawed option – that risk is determined by the scale of losses actually inc-urred, requiring a calculation that can only be performed with the benefit of hindsight.
Does the average investor ever really believe that he has a higher-risk investment if the chance of failure is very low? After all, he already accepts the risk of loss in the case of deposits because he does not expect it to happen.
If the chance of failure is very low, he sees the investment as low risk, irrespective of the scale of potential losses (except, perhaps, if losses are not limited to the original investment amount).
How risky are, for example, split-capital investment trusts and precipice bonds? Well, the answer clearly lies in a statistical analysis of the particular investment. Taking a precipice bond, let us say the bond pro-mised to pay a guaranteed return unless a particular index fell by more than a certain amount within the product's lifetime.
Determining the risk of failure can best be done by looking at the relevant index, or other fair comparisons if a longer timescale is reasonably required, and calculating the historical likelihood of a fall of the required amount within the product timescale.
Clearly, such calculations are likely to give markedly differing results for different products, resulting in differing risk categorisations for apparently similar products.
To a degree, this is a fea-ture of the “art” of risk assessment over the science. Risk assessment can only ever really show what the risks have been historically and it is then usual (and probably reasonable) to assume that, while the past may not be a guide to the future, it is the only one we have got.
For example, if one particular equity index fell by 25 per cent once in its history and another fell by only 24 per cent, the choice of index will have a significant relative impact upon the apparent risk of products linked to each index if a fall of 25 per cent is required to cause underperformance.
So, the best that we can do is to determine, based upon past performance, the likelihood of a particular investment failing to meet expectations. Such calculations will throw up some unexpected results and will go a long way to show that advisers' and providers' views of product risk were often correct at the time of sale, despite the FSA's stance that, if the investor has lost money, the product cannot have been low risk.
If the likelihood of failure of an investment was low, the fact that big losses were made does not make the investment higher risk. Low risks do happen, just less often. A proper analysis of the true risks of a product should save advisers and providers from paying out on many misselling claims.