Join our community of smart investors

Cheap equities, expensive housing

Shares will probably outperform house prices in the next few years.
April 11, 2019

Equities are very cheap relative to housing.

My chart shows the point, showing the gap between the rental yield on housing and the dividend yield on the All-Share index. If we combine data from Homelet and the Nationwide, the rental yield is currently 5.3 per cent. That’s barely a percentage point above the dividend yield, which is close to the smallest gap we’ve seen since 2008.

Now, we’d expect houses to yield more than shares because there are costs to insuring and maintaining a house. The question is: why is the yield gap so small now?

One possibility is that it is a reasonable assessment of the growth prospects of the two assets. If dividend growth is likely to be lower or riskier in future than it has been in the past, then dividend yields should be high relative to housing yields.

One reason why this might be so lies in a big difference between houses and shares. A new house doesn’t render an old one obsolete. But a new product or company might destroy older ones: think of how smartphones devastated Nokia, or how online shopping is clobbering traditional retailers. Equities face creative destruction risk; houses do not.

But why should this risk be especially high now? In one sense, it shouldn’t be: western economies are less dynamic now than they were a few years ago. Instead, there’s another problem. It’s that the best corporate growth in future might come from outside the quoted sector – from companies that aren’t yet listed or which might not even exist yet – and by the time a company does list on the market it has enjoyed most of the growth it will see. Kathleen Kahle at the University of Arizona and Rene Stulz at Ohio State University have shown that in recent decades quoted companies have become bigger, older and less profitable than they were in the 1970s and 1980s. That means they are more likely to have gone ex-growth.

A second danger for shares relative to housing is distribution risk. We can think of shares as being a claim on future profits and housing as a claim on future wages. Anything that reduces expected profits but raises expected wages should therefore depress share prices relative to house prices. One possibility here is that near full employment might eventually reduce profit margins. Another possibility is political risk: if a Labour government nationalises utilities or regulates them more severely or (more arguably*) raises company tax, profits could fall relative to wages.

If these are the good reasons why equities might have a high yield relative to houses, there are a couple of less rational ones.

One is that house prices tend to be less flexible than share prices. If shares suffer bad news, prices can fall a lot, and quickly. If the housing market gets bad news, however, sellers often maintain their asking prices with the result that transactions dry up instead. That means that prices drift down much more slowly as sellers eventually face up to reality. This might be happening now. Simon Rubinsohn, chief economist at the Royal Institution of Chartered Surveyors (RICS) has warned that “a reluctance from some vendors to acknowledge the shift in the balance of power in the market will compound the difficulty in executing transactions." To the extent that this is the case, housing might now be overpriced.

There’s another problem. Rental yields have been much more sensitive than dividend yields to the fall in long-term real interest rates since the 1990s. Whereas the rental yield has halved since the mid-1990s (during which time 10-year index-linked yields have dropped from 4 per cent to -1.9 per cent) the dividend yield on the All-Share index hasn’t changed much.

This raises a worrying possibility. Perhaps the housing market has been irrational. It has celebrated the good news that borrowing costs have fallen. But it has ignored the dark side to this – that falling real yields are a sign of lower future economic growth. Sure, mortgage rates are lower now than they were in the 1990s. But a home buyer then could look forward to the real burden of his mortgage falling as his income rose. A borrower today has no such comfort. To the extent that this is the case, houses are over priced.

So, is the low gap between housing and dividend yields rational or not? We have a test here. If the gap is rational then it predicts differences in rent and dividend growth. If it is irrational, it predicts differences in price growth.

History suggests it’s the latter, to some extent. Since 1987 there has been a statistically significant correlation (of 0.32) between the yield gap and subsequent three-year changes in house prices relative to equities. A high gap (such as in the late 1990s) leads to house prices rising relative to shares, and a low gap (such as in 1989) leads to shares rising relative to house prices.

This correlation, however, is far from perfect – which suggests there is also some truth in the idea that the two markets might be rationally priced relative to each other. Nevertheless, it warns us that, for now at least, equities look like a better investment than housing.

*The issue here is one of incidence: do corporate taxes really reduce profits or do they instead cut employment and so, in effect, fall upon wages?