One clue that this is the case came in official figures last week showing a significant increase in the number of EU citizens leaving Britain last year. Is it really an accident that this has coincided with a steep decline in house price inflation? The Halifax house price index next week could show a drop in house price inflation to its lowest rate since 2013. Migration affects house prices.
This is by no means the only evidence. Tarun Ramadorai and Cristian Badarinza have shown that London house prices are driven in part by political uncertainty overseas. An increase in instability in, for example, Pakistan, leads to higher prices in areas where Pakistanis are disproportionately likely to live. This shows that London prices are driven by a global safe-haven demand. And to the extent that variations in London prices spread to the rest of the country, it follows that there's a nationwide effect.
There's another mechanism. House prices depend a lot on bond yields, which are of course determined largely globally. Since 1986 there has been a huge correlation (of minus 0.8) between 10-year index-linked gilt yields and the ratio of house prices to earnings for first-time buyers. Lower bond yields mean lower mortgage rates, which pushes up prices. Equally unsurprisingly, a lower discount rate should mean higher house prices.
Now, this isn't to say that the housing market is entirely globalised. The current slowdown is due in large part to domestic factors, such as higher taxes on buy-to-letters and a lack of affordability.
Nevertheless, this should alert us to some perhaps under-appreciated risks to house prices.
One would be a decline in foreign demand. This could happen simply if overseas buyers diversify away from London property, if they sense a less welcoming country post-Brexit, or if the government actually hits its migration target. These forces might be already under way: both Hometrack and the RICS report a decline in house price inflation in London.
Another risk is that if we see a global sell-off in bonds, house price-to-earnings ratios should fall; just as lower bond yields led to high multiples in the 2000s, so higher yields should mean lower multiples. The short-run correlation between house prices and bonds might be low (in part because house prices are slow to adjust) but the longer-term one is more significant.
I fear, though, that there's another problem here. In principle, low bond yields should be ambiguous for house prices. On the one hand, lower mortgage rates (and a lower discount rate on future rents) should push prices up. But on the other hand, bond yields are low in large part because markets expect slow growth in incomes, both here and in western economies generally - or, just as bad for house prices, another recession. And low growth in future incomes should depress prices now.
This raises a danger. House prices have priced in the good news of low bond yields, but not the bad. To put the same point more concretely, how will people on interest-only mortgages repay their loans if their incomes continue to stagnate?
