There is a line of thinking that a single core asset allocation portfolio (or perhaps a small number of standard portfolios) may be appropriate to the vast majority of investors.
This view is based on a concept of diversification between assets whose performance is very lowly correlated with each other, that is, their investment performance follows different trends.
To the extent that this approach might have some merit, the majority of investors would therefore be advised to follow the same or very similar asset allocation decisions. Any required increase or reduction in risk from this model portfolio might be achieved for each investor by selecting appropriate individual funds with high standard deviations (to increase volatility within the fund), low correlations (to reduce portfolio risk) or standard deviations close to each fund's sector average (to provide an average overall portfolio risk).
As an example, I recently noted the standard deviations of the main asset classes over a five-year period as being:
It is no surprise that cash has represented the asset class with the lowest volatility but some people might be a little surprised at the very low volatility of commercial property, especially against gilts. Moreover, it can be noted that mixed managed funds had not over this five-year period produced a significantly lower level of volatility than a pure equity fund and even defensive managed funds generated volatility around double that of commercial investment property. Even at this level of standard deviations there are some valuable messages for investment advisers and clients.
Driving our studies deeper than this level can yield even more valuable messages. If we look at the standard deviations of funds within the commercial investment property sector, of which I have noted five in the table above right (being the two funds with the lowest standard deviations, the two with the highest and one with a standard deviation broadly in line with the average), we can derive much guidance about fund selection of clients in different circumstances.
We know that the property sector as a whole has experienced very low volatility over the last five years in comparison with the other major asset classes and thereby represents a valuable low-risk component of many portfolios (if past volatility can be taken as a guide to the future which, in this case, we would suggest it probably can).
If property is included in a portfolio in which it is desired to reduce volatility even further, an adviser might be prompted to give particular consideration – all other things being equal, especially fund performance and management styles – to funds A and B, both of which have historical standard deviations around one-quarter the level of the sector average.
Conversely, if a client's circumstances indicate that higher volatility can be tolerated in the hope of achieving above-sector-average short-term gains, then perhaps more attention can be directed towards funds C and D.
Fund E's standard deviation more or less replicates the sector average so it might be considered in most circumstances but particularly those in which the portfolio's aim is to achieve a risk profile which broadly replicates the average volatility for each asset class.
Similar comparisons can easily and quickly be made in other asset classes and sectors, for example, UK and overseas equities and fixed-interest gilts and corporate bonds, with similar significant differences between funds.
This aspect of fund selection should be a crucial part of an adviser's portfolio planning strategy to enable model portfolios to be adjusted according to the particular needs of each client. For example, clients with a very high tolerance to risk in a particular financial planning strategy – perhaps those aiming for above-average investment returns in the short to medium term – might favour funds with high standard deviations as, all other things being equal, these funds will tend to alternately show gains and losses well above the sector average. Investment returns from these funds will therefore tend to appear in either the first or fourth quartiles in their sector.
Of course, consideration of standard deviations is not the only or even the major factor in fund selection. Investment performance and management style, among other aspects, must continue to play a big part. However, a clear understanding of the meaning, uses and implications of standard deviation statistics can greatly assist an adviser in structuring a portfolio which is truly appropriate to the risk profile of the client's needs.
It must always be remembered that published standard deviations for asset classes, sectors and funds indicate historical volatility and should not be interpreted as a definitive guide to future volatility.
However, a study of the past volatility of funds and sectors can often be an extremely useful first stage in a projection of future volatility, as the above property statistics can demonstrate.
As we identified earlier, the investment property asset class has an historical standard deviation less than a quarter that of UK equities. If we can identify whether there is a logical reason for this lower volatility, we can hope to establish or justifiably speculate whether it is likely to continue in to the future.
Although anything other than a very brief summary is beyond the scope of this article, it can easily be identified that by far the biggest element of returns from investment property is rental income, with capital growth representing a much smaller part. As many leases of investment property are established over a long period (20 years or more is not uncommon) and frequently with periodical rent reviews which are stipulated never to be reviewable downwards, rental income then becomes very predictable, with no downside and steady if unspectacular upside leading to very low overall volatility in total returns from this asset class.
Thus, the low standard deviation for property has been the catalyst for further consideration of the nature of investment returns, which can be interpreted as indicating that the low volatility of investment property as an asset class could be expected to continue.
Similarly, we can perhaps anticipate that there is a reason for the very low standard deviations of funds A and B and the very high standard deviations of funds C and D, which might indicate the likelihood of these features continuing into the future. In fact, a little more research reveals that funds A and B invest directly in investment property, with by far the biggest part of their returns coming from rental income over long-term leases – exactly the attribute we identified for the property sector generally.
On the other hand, further investigation reveals that funds C and D invest in the shares in property companies rather than directly in investment property itself. This being the case, the price behaviour, including volatility, of these latter two funds more closely matches equities than the property sector, hence standard deviations that are more akin to equities.
Although arguably a little simplistic, this type of research – starting from identification of significant differences in standard deviations – can justifiably suggest that the differences in historical volatility can be expected to continue in the future.
No conclusion should be reached about likely future investment returns, simply that certain asset classes, sectors and funds may be more appropriately used in future for higheror lower-risk situations.