Housing Bubbles: Less Frothy but Europe is Behind
Wolfgang Muenchau’s article in the Financial Times, There is no Spanish siesta for the Eurozone, inspired me to update my post on housing bubbles around the world (really just Europe and the US). He argues that Spain’s bubble was much more extreme, and that the price adjustment is less mature compared to the others. I would add here that it’s European housing markets more broadly that look overvalued compared to that in the US, as measured by the price to rent ratio.
The chart below illustrates the housing bubble, as measured by the house price to rent ratio, in the US, Spain, the UK, and Ireland that is normalized to Q1 1997 and through Q1 2011. The price to rent ratio can be compared to a price to earnings or a price to dividend ratio in finance. It measures the relative value of the asset: the price of the asset (purchase price of a home) divided by its flow of fundamental value (rental income earned or the value of having a roof over your head). As the price-rent ratio falls, the market home values moves closer to fundamental value.
Spanning the years 2005 to Q4 2011 and indexed to 1997 Q1, home values peaked at roughly 1.7 times rent in the US, 1.8 times rent in Spain, and north of 2 time rent in Ireland and the UK. Since the peak, though, US home values have fallen to 1.0 times rent – a considerable reduction in asset prices toward fundamental value. In contrast, home values in Spain, the UK, and Ireland remain quite elevated to rents, 1.3 times, 1.6 times, and 1.4 times, respectively in Q4 2011. If 1.0 is deemed equilibrium, either home values in Spain, the UK, and Ireland must fall further and/or rents rise to normalize home values. That’s a tall order: rising rental values amid defficient and contracting domestic demand in Spain and possibly Ireland.
The UK has more of a fighting chance, given its relatively easy monetary policy, compared to Ireland and Spain, where more accommodative monetary policy is very lagged amid fiscal contraction. Without growth, though, default is probably the only answer left to normalize housing markets in Spain and Ireland.
12 Responses to “Housing Bubbles: Less Frothy but Europe is Behind”
Great post Rebecca – It would be interesting to see how Canada's housing situation looks.. as we are still waiting for a pull back in our markets.
Thanks Rebecca for this very interesting post. I was intrigued how this ratio would look for Australia so I constructed a chart with similar parameters. It's 1.7 at the end of December 2011 (higher than in EU countries in your chart but not too much, considering we only had modest price falls so far, less than 5% on country wide basis).
I posted the chart to my twitter account for easy reference, if of interest: https://twitter.com/#!/arekdrozda/status/18190333…
Would you mind posting the equation used please?
It's nothing fancy, exactly as Monica described below. Excel formula:
where column B is CPI (rent) and Column C is Establish House Price Index, 8 Capital cities weighted average (which in case of Australia is the official "national house price index").
Where were you when I was doing my dissertation in university? If I had seen this equation back then, you would have absolutely saved my life!
I am sorry, your ratios look bizarre. Even if normalized to to some base-date.
What do you put in the numerator: the price of a housing unit? In the denominator the annual rent for a comparable unit? And this ratio would only be 1.3 in the US today?! What are they in the city where you live. These ratios would be more like 12-15 today.
These are ratios of ratios. Some home cooking!
The ratio here in the chart says HPI / CPI, which should refer to House Price Index ( house price sales price change ) over CPI rent expenses change.
On the subject of “bubbles” I believe that annual “cost of buying” a property is more relevant measure than price alone (simple logic is that hardly anybody buys properties for cash and buyers first determine how much they can “afford” – ie how much they can borrow to be able to service the loan on an annual basis, plus some buffer in case of interest rate hikes). That is, although price directly influences costs, in itself it is of secondary importance. And since paying off loan principle can be considered a form of “saving”, only interest payments are of relevance.
The cost approach allows to draw some interesting conclusions since it brings together change in house prices AND change in interest rates (which fluctuate in Australia quite widely). More so, if contrasted with changes in incomes and rental costs. Anyway, in case of Australia, all 3 appeared to move in very similar proportions over the last 25 years. Here is a chart showing the relationships: https://twitter.com/#!/arekdrozda/media/slideshow…
Which brings me to the issue of selecting 1997 as a base year. In case of Australia, it was a “low buying cost” year (we were just coming out of the 91 recession) so, the rise in house prices in the following years may be exaggerated in relation to the rise in rental costs. In other words, prices had so much room to move because incomes and rents increased so disproportionately in the preceding decade. Another proof of irrational consumer behaviour…
the states where a minimum wage worker can afford an apartment:
Just out of intest Rebecca, what calculation are you using for rent in the denominator? Is it over an infinite time horizon? Also what sort (if any) discounting do you use e.g. monthly, continuous?
Are the rental markets in Europe more regulated so that the rents cannot rise as quickly as they have in the US, implying that part of this discrepancy is simply a more sclerotic rental market in Europe?
As I see things, ratio is the average rent divided by the average monthly cost to fund and operate a house (including tax, insurance, maintenance, opportunity cost of equity capital, etc.). There is no infinite series. Ignore principal payments.