The tool in question is the Sharpe ratio, named after a Nobel Prize-winning economist, William Sharpe, who also gave the world the capital asset pricing model. It is well known - you can find values for the Sharpe ratio of unit trusts and investment trusts on many financial web sites - yet it is poorly understood.
This means there is little intuitive feel for what constitutes an interesting Sharpe ratio in the way that there is for the ubiquitous price/earnings ratio. Spot two companies whose shares trade on earnings multiples of, say, six and 20 times and intuitively we know that one is lowly rated and the other highly rated. From that we can immediately start to ask relevant questions. Find comparable Sharpe ratios of, say, 0.1 and 0.2 and no touchstone springs to mind to prompt the thought that one fund manager might be underrated, while the other is pushing his luck. Nor does it help that respective Sharpe ratios often aren't comparable. But we'll come to that in a moment.
First, let's define it. Textbooks say the ratio is also known as the 'reward-to-variability' ratio, although a better description might be 'reward-to-worry' ratio. It quantifies how much anxiety, as measured by bouncing returns, an investor has to suffer for every unit of reward he garners. The reward is the return that a fund generates above the so-called risk-free rate, which is the yield on some form of government debt. The variability is the standard deviation of the fund's returns.
Take the table. Over the 28 years from 1984 to 2011, the average annual return from UK equities, as measured by the All-Share index, was almost 8 per cent. Over the same period, the risk-free return, as quantified by the yield on 10-year gilts, was 6.7 per cent. Immediately we can see that UK equities did not generate much excess return. Meanwhile, they endured a standard deviation of 16 per cent, which means that two-thirds of the time each annual return was somewhere between plus 24 per cent and minus 8 per cent. That translates into a Sharpe ratio of just under 0.1 (an excess return of 1.3 per cent divided by a standard deviation of 16 per cent). Is that any good?
| Stay Sharpe | ||||
| 1984-2011 | Equity returns (%) | Standard deviation (%) | Risk-free return (%) | Sharpe ratio |
| UK | 8.0 | 16.1 | 6.7 | 0.08 |
| USA | 9.6 | 15.4 | 4.1 | 0.36 |
As we are dealing in ratios - the relationship of one figure to another - there are no absolute values. So we need to make a comparison. Returns for US equities - as measured by the Dow Jones Industrial Average - over the same period make an obvious comparator.
In a sense, it's not much of a comparison because the performance of US equities is so obviously superior. Average returns are higher than the UK's, but come with lower standard deviation; while the risk-free rate is lower, too. Obviously that translates into a better Sharpe ratio: 0.36.
However, the ratios demonstrate a familiar problem with this particular tool - they are not comparable. I deliberately based the US risk-free returns on three-month Treasury bills, while the UK's are based on 10-year gilts. Which is the correct risk-free rate? It depends on your perspective. One view would say that the term of the risk-free asset should coincide with the holding period for the investment portfolio - five years, 10 years, whatever. Another would say that short-dated Treasury bills are the ultimate risk-free asset, so their returns should always be the benchmark. Bring the UK data into line with that of the US, by using UK T bills, and the Sharpe ratio improves to a fraction over 0.1, although it still lags far behind the US.
From this, two points arise. First, a Sharpe ratio will rarely be greater than 1.0. That should not be surprising given that the volatility of any equity portfolio is almost always much higher than its average return. That ratios are often expressed as 1.5 or 2.0 usually indicates that the decimal point has been shifted a place to the right for convenience.
Second, a Sharpe ratio of 0.3 or higher is exceptional. Indeed, for equity portfolios it's rare to see such ratios using data from recent years. For example, based on monthly returns, the ratio for the Bearbull Income Portfolio since January 2000 is just 0.1. If that sounds ordinary, consider that over the same period the All-Share index can't even muster a proper Sharpe ratio because its average monthly return is negative.
■
. You can adapt it for you own portfolios. Sure, the Sharpe ratio is hardly the ultimate tool for assessing fund performance, but it's a useful addition thanks largely to its intuitive simplicity. Odd, then, that it's so poorly understood.
