Are many of us holding too many equities? I ask because equity weightings of 70 per cent or more seem common, but are hard to justify with conventional economics.
To see the point, let’s start from some work done back in 1969 by the Nobel laureate Robert Merton. He asked how an investor should divide his wealth between risky assets (such as equities or property) and safe ones such as cash or bonds. After some fiendish maths, he derived a simple equation. The proportion of wealth you invest in the risky assets, he showed, should be equal to the expected excess return on the risky asset over the safe one, divided by the product of two things: the variance of the risky asset, and a coefficient of your risk aversion.
Let’s see how this might support a high equity weighting. If we assume equities will give a real average annual return of 5 per cent (say, 1.5 per cent real growth and a 3.5 per cent yield) and cash a zero real return we have an expected excess return of 5 per cent or 0.05. Let’s say annualised volatility is 15 per cent. Squaring this gives us the variance of returns, which is 0.0225. If your coefficient of risk aversion is three, then we have an equity weight of 0.05/(0.0225 x 3), which is 74 per cent.
Now, I appreciate that a coefficient of risk aversion is econo-jargon you might be unfamiliar with. But we can reverse-engineer this equation to get a more familiar sense of risk. Our numbers imply that such a portfolio gives you around an 11 per cent chance of losing 10 per cent or more over a 12-month period. I suspect many of you would be relaxed about this, and do in fact have an equity weighting of around this much.
What you shouldn’t be relaxed about is that the numbers I’ve used so far are deeply questionable.
For one thing, an expected excess return of 5 per cent per year is dubious. Over the past 30 years you would only have made such a return if you had bought at the very bottom of the market. More commonly, we’ve had excess returns of around two percentage points.
Yes, there are good reasons to fear low or even negative returns on bonds. But some of these – such as rising interest rates and a reversal of quantitative easing at least in the US – are also reasons to be wary of equities.
If returns are one problem, risk is another. Yes, past volatility suggests that shares have been only moderately risky. But we’ve no assurance at all that the future will resemble the past. We face numerous risks: prolonged stagnation; a fall in the share of profits in GDP; creative destruction which benefits unquoted companies at the expense of quoted ones; a recession when central banks are short of monetary policy room; or rising populism. The fact that shares have overcome past dangers does not guarantee they’ll thrive in the face of future ones.
Plugging some other numbers into Merton’s equation gives us much lower equity weightings. A risk premium of two percentage points and a doubling of the variance (to account for the fact that we face what Richard Bookstaber calls radical uncertainty as well as ordinary volatility) gives us an equity weighting of only 15 per cent.
You might think this is absurdly low.
From one perspective it is, because many of us own other risky assets which are correlated with share prices. Property might well do as badly as shares in a recession – and it carries liquidity risk, too. If we count property as a risky asset (as we must) many of us have exposure to risky assets much greater than 15 per cent.
From another perspective, however, it’s not at all absurd. Many of you in fact do have such a low weighting because you have huge investments in safe or safeish bonds, even if you don’t know it.
One such bond is a defined benefit pension. If you’re a 60-year-old with a £20,000 inflation-linked pension income, you have in effect a bond worth £600,000. Even if your other financial assets have a high equity weighting, therefore, your overall equity weight is low. This is true for those of you still paying into a DB pension: in effect, you have a bond future.
Many of you have another bond - your human capital, or power to earn a salary. If your salary is uncorrelated with equity returns – say, because you are in a safe job – then you have an asset that diversifies equity risk. When share prices fall, you can top up your wealth from your savings. And for those falls which lead to higher subsequent returns your extra savings will make more money. In this sense, your human capital reduces your equity risk, which justifies holding more of them.
In truth, all of us have some kind of bond holding: the state pension, in effect, gives us all an index-linked bond. Which means our equity weightings aren’t as big as they seem.
Even with this large caveat, however, the advice that we should own so few equities seems disturbing and counter-intuitive.
Maybe, then, a little historical context will help. The idea that investors should be massively exposed to shares is a relatively recent one. We can perhaps date it to the 1950s when George Ross Goobey moved Imperial Tobacco’s pension fund from gilts into shares, thus beginning the cult of the equity.
That made great sense at a time when shares were undervalued, when the economy was growing well, when risks (we now know!) failed to materialise and when investors under-rated the damage that inflation would do to gilts. It’s not obvious, however, that those conditions will apply in future.
Which poses the question. Could it be that our high equity weightings are – like perhaps many ideas – a legacy of an era when they made sense, but that this era has passed?
