Over the last six weeks, I have discussed the potential impact of volatility, correlation strategy and efficient frontier on portfolio planning. The aim was to devise a portfolio to at least give the highest possible returns for a given investor risk profile.
More than this, though, the combination of these concepts and strategies aims to (though, of course, does not always claim to be successful) combine generally low-correlated assets in such proportions as to yield investment returns much higher than the overall portfolio risk profile would suggest is possible.
All very good in theory, but is it possible in practice? Over the next few weeks I will look at features of different classes of assets and their respective sectors. We will examine, in particular, their respective past performance and expected future performance, reasons for any major changes along the way, past and expected future volatility, and the typical use of asset class within different portfolio profiles.
As the weeks progress,we will gradually build on the past and expected future cross-correlation of each combination of asset class so that we may conclude by suggesting a range of high-return/low correlation portfolios for use in a wide range of practical situations, including pension drawdown (as an excellent example of the working of these strategies in real life).
We will start out by looking at asset classes with cash. It would perhaps be better to call this “money on deposit”, and it may not be obvious at first sight that this category includes a wide range of deposit-based investment opportunities.
At the bottom end, we could include money on deposit in a current (that is, cheque)account which might, but usually does not, pay a relatively low amount of interest on credit balances.
Above these accounts, we can obviously identify bank deposit accounts and the wide range of building society accounts, paying a variety of rates on interest. It is most instructive to look above all of these retail deposit vehicles at the wholesale market – where the institutions deposit and lend money to each other. It is the wholesale rates on which the retail interest rates are ultimately based.
Wholesale deposit rates are available in a wide range of guises, typically dictated by the term over which the deposit money is tied up – ranging from one hour to, usually, no more than one year – and the amount invested.
An example of UK rates available on the money market earlier this month was Interbank sterling, which for seven days notice was 6 per cent, three months was 5.75 per cent and one year was 5.5 per cent.
You can see there appears to be relatively little difference between these rates but such difference as there is could be taken in part as an indicator that the market is anticipating fairly stable money market rates over the coming year or, if anything, falling slightly.
I have only quoted UK sterling money rates here – different rates apply for deposit money in other currencies. These are potentially attractive to UK investors wanting to seek higher interest rates abroad. It should be noted, though, that the potential interest rate gain or loss must be balanced against the prospective comparative strength in the currencies of the country of deposit and sterling.
Past and expected
Over the last 20 years, the average annual rate of return on deposit money, as measured by reference to the London Interbank seven-day rate, has been around 9.5 per cent a year. If this sounds higher than you might expect, remember the days 10 and 20 years ago when interest rates exceeded 15 per cent.
I am sure the vast majority of readers will anticipate that this annualised return will compare unfavourably when we look at the rate of return over the other asset classes we discuss over the next few weeks, but you never know.
It is extremely difficult to predict future returns from money on deposit, although some indication may be derived from the respective annualised redemption yields from gilts which have progressively longer redemption dates.
We will cover this point in more detail when we discuss gilts but, for now, suffice to suggest that these redemption yields indicate (in theory) whether the market expects interest rates to rise or fall over a stated period of years.
Past and future
The lowest annual return from deposit money over the last 20 years has been 4.9 per cent (in 1994) and the highest has been 18.6 (in 1980, with a later peak of 15.7 per cent in 1990). I have calculated that one standard deviation is 4.25 and two standard deviations are nine.
The figure for one standard deviation means that in 70 per cent of the years in question, the rate of return from cash has fallen no lower than 5.25 per cent (that is, the average return of 9.5 per cent less 4.25 per cent) and been no higher than 13.75 per cent (that is, the average return of 9.5 per cent plus 4.25 per cent).
Putting it another way, this means that in 30 per cent of the years in question (30 per cent of the 20-year period equals six years) the rate of return on cash has fallen below 5.25 per cent or risen above 13.75 per cent.
But this standard deviation figure does not tell us that those six years have been divided equally between particularly bad or good years.
In fact, the annual rate fell below 5.25 per cent on two occasions (1994 and 1999) and exceeded 13.75 per cent on four occasions (1980, 1981, 1989 and 1990).
Similarly, the figure for two standard deviations of nine indicates the rate has not fallen below 0.5 per cent nor risen above 18.5 per cent in 19 of the 20 years in question (that is, 95 per cent of the time). In fact, the upper limit was breached in 1980 (18.6 per cent), as we noted earlier.
If these figures have not made full and immediate sense to you could I suggest a slow re-read, as we will be developing this important point in future articles.
All good stuff, perhaps, but where does all this get us, apart from a mathematics-induced headache? Well, as the forthcoming comparative articles will indicate, it is all a question of relativity – that is, the volatility of cash relative to the other asset classes.
We will confirm in these articles that the detail does not really matter that much – it is simply a matter of comparing the standard deviations of one asset against another.
I would suggest that most readers will currently be thinking that, therefore, rates of return from deposit-based investments have very low volatilility.
This is obviously based on your preconceived ideas and so I would urge you to remember that standard deviations are only a useful tool when they are compared with the standard deviations of comparable or competitive assets, as we will see over the next few weeks.