According to Homelet, average rents in May were £919 a month. With the Nationwide reporting that average house prices were £213,618 then, this implies that average rental yields were just under 5.2 per cent at an annualised rate.
By combining the Nationwide’s house price index with the ONS’s data on historic rents, we can estimate past rental yields. My chart does this. It tells us that yields are now close to their lowest level since early 2008. And we all know what happened soon afterwards then.
There is, however, a reason for low rental yields. It’s that yields on safe assets are low, a fact that should reduce yields on other assets. Since 1987 there has been a strong correlation (of 0.66 in quarterly data) between my measure of the rental yield and the 10-year index-linked gilt yield. One reason for this is that low interest rates encourage people to take out bigger mortgages, which pushes up house prices and so depresses yields. Also, low long-term interest rates mean that future incomes are discounted less heavily, which means that any asset offering future income – such as a house – becomes more valuable. It is for these reasons that many economists believe that the main reason for high house prices is not so much a lack of supply as the fact that interest rates are low.
From this perspective, what’s surprising is how high rental yields are. They are now 6.8 percentage points higher than index-linked yields, compared to a post-1987 average gap of 5.4 percentage points.
You can read this as a lack of confidence in housing. By definition, the gap between the two yields should be equal to the risk premium on housing over gilts, minus expected house price growth. A high yield gap therefore signals either the perception that housing is unusually risky and/or a belief that future capital appreciation will be low.
If this suggests that housing is cheap – albeit perhaps for good reasons – another indicator suggests they are expensive. This is the gap between the rental yield and the dividend yield on the All-Share index.
Rental yields have been above dividend yields for at least the past 30 years. This is reasonable. It reflects the fact that houses are harder to sell than equities, especially in recessions, and so they should carry a liquidity risk premium. Also, the prices of individual houses might well be riskier than those of diversified baskets of shares, which also justifies a higher yield. Plus, of course, there are costs of home ownership: insurance, maintenance, dealing with tenants and such like.
The gap between the two yields, however, is unusually low now. It’s 1.4 percentage points compared with a post-1987 average of 3.7 percentage points. This tells us that sentiment towards housing is unusually positive compared to sentiment towards shares.
This isn’t wholly absurd. There are reasons to worry about the sustainability of share prices. High foreign buying of US equities and a high ratio of global prices to the money stock are both lead indicators of a fall in prices. Also, it’s not inconceivable that incomes will shift in coming years from profits to wages – either as an effect of high employment or because of a future Labour government’s policies. If so, house prices might do well relative to shares (you can think of house prices as a claim on future labour income whereas share prices are a claim upon future profits).
That, however, might be too optimistic an interpretation. In the past, low rental yields relative to dividend yields have been a predictor of weak house prices. Since 1987 there has been a strong correlation between the rental-dividend yield gap and the subsequent three-year change in house prices: it has been 0.6. If this relationship continues to hold, it points to house prices rising only around 1 per cent a year over the next three years, which means they’ll probably fall in real terms.
So yes, those of you who are long of housing should worry that its yields are low.
