Join our community of smart investors

Forty years of disappointment

Equities have done badly for decades, relative to bonds
August 4, 2020

Everybody knows that bond yields in developed economies are close to all-time lows. What’s not so well-known is that this has important implications for how we think about equities.

A low bond yield means that we apply a low discount rate to future cashflows. Which in turn means that if other things are equal (a crucial caveat we’ll come to) then lower bond yields should raise share prices simply because shares are a claim upon future cashflows.

But shares haven’t risen as much as you’d expect in light of the fall in yields – and not just because they’ve done badly this year. The simplest way to think about this is in terms of perpetual bonds – ones that never mature. Their price is equal to their annual coupon divided by their yield. This implies that if the government had not redeemed them in 2015 their price would now be around 25 times higher than it was in 1977, when yields hit a record high. The All-Share index, however, is only 18 times higher than it was then.

My chart puts this another way. It shows that total returns on the All-Share index have been much the same as those on gilts since the mid-80s. If you’d bought at the right time you’d have done better in equities than bonds, but if you had bought at the wrong time you wouldn’t. Generally, there’s been no equity premium in the past 35 years.

This isn’t a quirk of the UK market. Jules Binsbergen at the University of Pennsylvania shows that much the same has been true of the better-performing US market. He has calculated that if we control for the different duration of cashflows then equities haven’t outperformed bonds since the 1970s.

To see why this is so weird, consider the basic difference between bonds and equities. Bonds pay a guaranteed coupon each year – which is why they are called fixed income investments. With equities, though, your income is not fixed: dividends are often cut in recessions, but increased in good times. This means equities are riskier than bonds. Which means they should pay higher returns over the long-run to compensate us for their greater risk.

But they haven’t done so. On average, for decades, investors have not been compensated for taking on the risk of dividend cuts. Which contradicts common sense.

So why have we missed out?

One possibility is that we never really needed such compensation. Dividends are in one sense safer than a fixed income because dividends rise with some types of inflation whereas all inflation erodes the real value of the income conventional gilts.

This theory, however, runs into a problem. It predicts that equities should have underperformed gilts in the 1980s and 1990s because we no longer needed so much protection from inflation as the latter fell. And it also predicts that equities should have outperformed since the late 1990s because inflation has been stable whereas recession risk has been considerable and we should have been rewarded for taking it. Both predictions, though, have been false. Something else must therefore be at work.

One possibility, says Professor Binsbergen, is that we’ve suffered a series of disappointments about dividend growth. The crises of 2008 and 2020 have taught us that deep recessions are more likely than we thought during the “Great Moderation” of 1992-2007. Hopes that the IT revolution would create a faster-growing economy have been proven wrong (so far?). And – partly thanks to falling profit rates – secular stagnation has intensified.

But there’s another possibility. It’s that shares have underperformed because investors have begun to demand a dividend risk premium that they didn’t in the 1970s and 1980s. This might be because they are now more aware than they used to be of the dangers of creative destruction. In the 1970s, we presumed that some companies would be around for life: Marks & Spencer, ICI, Cadburys and so on. Today, we are less naïve. Such destruction threatens not just dividend growth but companies’ very existence. In fact, Hendrik Bessembinder at Arizona State University shows that the typical share underperforms cash during its lifetime.  If investors have become more aware of this, they should have increased the risk premium they attach to dividends, causing share prices to fall to a level from which future returns will be high enough to compensate for their risk.

If this is the case, then future returns on equities should be better. They should also do better if the disappointments about growth were just bad luck and if we get more normal luck (relative to our low expectations) in future.

But there’s a more disturbing question raised by all this. If shares have done well since the 1970s only because bond yields and hence the discount rate have fallen, what will happen if or when yields rise?

If they do so because of better news about economic growth, equities should do OK. But if they rise for other reasons – say because of a diminution in the global savings glut – they might not.

Personally, I slightly favour the optimistic reading. But let’s be clear. There are huge uncertainties here, and if equities do well over the long term it will only be because these risks pay off. The idea that shares are safe in the longer-term is a dangerous one.