The spectre of housing bubbles is starting to raise its head again, along with discussions about what the central banks, regulators and/or governments should do about it.
House prices have been going through the roof in China, there is talk about a bubble in Australia and yet over there prices keep rising, Canadian house-prices are looking “toppy” and in Riyadh they are insane.
It’s not as if there are no tools for combating a bubble. Many countries that did not have the option of controlling the money supply via the blunt instrument of interest rates (typically because their currency is tied to the US dollar), have used those tools successfully for decades. A good example is Hong Kong which controls both the supply of land, the maximum LTV that mortgages can we written at, and also the amount of exposure that the banking system can have to residential real-estate.
America (and other countries) that suffered from exploding house-price bubbles, had no excuse, the tools were there, just they chose not to use them.
The reason was mainly because there was no agreement on what is the correct or fundamental value of a house; even though some people were warning about potential problems with housing in USA and also in UK in 2003, there was no consensus. Alan Greenspan spoke of a “bit of froth” in 2005, but he had no courage in his convictions, and even as late as early 2008, just before the slump really started in UK, Gordon Brown was saying that housing in UK was “affordable”, and so there was no risk of catching the American disease.
Even now, after a cataclysmic wake-up call, there is no consensus about what is the “right” price for housing.
What is the right price?
At it’s simplest level (it gets more complicated), the “right” price for housing is that which allows the population to spend a fixed proportion of their income on “shelter”, that proportion is broadly well-known, it works out somewhere between 20% and 30% (and I’m not going to argue-the-toss about exactly where it is, although there is a broad consensus in that the calculation of the weighting of “shelter” in the CPI analysis provides good base-line).
So in a stable population, the expenditure by people per house should be a function of the aggregate income, whether that’s disposable or not is debatable, or for example nominal GDP. Personally I like nominal GDP because it is (a) checked three ways when it is measured (b) you require no further soul-searching about what element of “income” to consider.
Thus the “right” price “ought” to be a function of the “expenditure” divided by the number of houses.
And historically, wherever I’ve looked, it is, plot nominal GDP per house in USA against house prices (annually), going back 100 years, and you get a 95% R-Squared (that means that 95% of changes in house prices in USA since 1910 can be explained simply by changes in nominal GDP per house).
Put in disposable income per house and you get a pretty similar result (although not quite as good which is one reason I like nominal GDP).
A nice clean example of how that works is Hong Kong:
That’s a rough and ready chart I put together a few months ago, the “Equilibrium (Other Than Market Value) is nominal GDP divided by the number housing units and then multiplied by a constant to get it on the same scale.
So yes there was a bit of a bubble in 1997/8 and like all bubbles bursting there was an “overshoot” in 2003, with the 70% drop caused more by a decline in nominal GDP than by the bubble burst; and now, yes it looks like a bit of a bubble. And the government of Hong Kong, as we speak, is doing something about that (pushing up the LTV and releasing more land for building). Over there the situation is slightly complicated by the fact that a lot of the money going into housing is from China-proper, but that’s another story.
Of course that isn’t the whole story, because the annual expenditure (by owners) on their house, only gets you to price if you apply a discount factor, i.e. you do an income capitalization valuation.
That sounds simple and rough and ready it is.
The “price” ought to be the “income” (how much goes to the house” divided by a yield which is the cost of money (i.e. the relative cost of for example living in the house, as opposed to selling it, putting the money in the bank (or buying a long-term bond), getting the interest and sleeping in your car).
You can make it a lot more complicated than that, to do a proper income capitalization valuation you need to project the future “income” (ideally you travel into the future and do measurements – and if “income per house” is likely to go up in the future, then you need to account for that), plus you need to decide what discount factor to use.
There are plenty of choices; you could use the mortgage rate, the base rate, the 10-Year Treasury, the 30-Year treasury. Personally I have found that a function of the 30-Year Treasury gets the best result, but there’s not much in it and the algorithm is not linear (it’s an “S-Curve”), but that’s “technical”. Another complication is whether mortgages are fixed or adjustable, I suspect that the “”bounce” in house prices in UK had a lot to do with the fact that mortgages are adjustable measured from the base rate, so dropping the base rate raised the “fundamental”, whereas in USA where a lot of mortgages are fixed rate, the lowering of the base rate didn’t help.
There are other issues, for example in Hong Kong and London, foreigners like to buy houses, but their income is not measured (or captured) in the city; so that can push house prices higher than the “fundamental” which would apply if the foreigners didn’t have a nice little pied de terre on the Edgware Road.
Effectively what happens there is that the foreigners push up house prices (nice if you own one), typically more than the extra revenue that they bring in, and effectively close out some of the supply; so to account for that you need top work out how many foreigners own second-homes but declare their income elsewhere (if at all).
Another thing is taxation, in USA it made a lot of sense to buy a bigger house than you needed and pay a bigger mortgage because a big part of the interest you paid was tax-deductable (renters didn’t get that break), and also you didn’t pay capital gains tax (in the days when was a capital gain). That meant the “cost” of ownership was distorted, as was (effectively) the discount factor in the income capitalization valuation.
House-prices can also get distorted by rent-control, when a portion of the population are paying much to little for housing (compared to what they “should” be paying if the market was not manipulated),, the rest of the population (typically new-arrivals), ends up paying more, and house prices (or the residual) get artificially inflated, that was the main reason for the Dubai house-price bubble.
Rental properties also distort the market in other ways, for example in many European countries it is impossible to get a tennant out, if the property was rented without furniture - so that type of property is held back from the market leaving only "furnished" property which gets a premium. For that reason and others, rent to price comparisons can be meaningless.
Equally the rental market can under report the real economic dynamic, for example in USA leading up to the bubble, there were huge incentives put in place to penalise renters and reward ownership, intriguingly in that case the cost of shelter in the CPI calculation was benchmarked against the rental market, which resulted in a massive under-reporting of the inflationary effect of soaring house prices.
Another way that house-prices get out of synch with the “fundamentals” is by hoarding of potential building land (which is what is happening in Riyadh), or by distortions in supply of building land which is what happened (happens in UK).
The issue there is that incumbents know that if the supply of housing goes up in their little micro-economy, the price of their house will go down. It a simple “anti-trust” play (except it’s legal – and they come up with all sorts of reasons why not to allow building next-door…often environmental, you can get those to run for years), and the point is that once 70% or so of the voters own their own houses, the majority wants house prices to go on going up forever and forever.
Except in the long-run that’s often unsustainable, because it increases the cost of doing business, and acts as a break on economic growth.
It also bleeds the country dry as people realise they can sell their incredibly expensive house in UK (for what you get), and buy something three times nicer in Florida, Spain, France, Croatia, or Dubai, for half the price.
The decision that governments need to take in such situations, is whether they want to have a country full of very-very expensive houses, with the poor people living ten to a room (and charging a fortune for their “services”), with all of the spare income and potential of the country drained away bit by bit, leaving a wonderfully rich gated community, and no poor people and no old people (except rich ones), or not.
The Labour Government under the dream-team of Blair/Brown decided on the latter, even though they had promised to build affordable housing for the less wealthy (which would have pushed prices down). In the event many of the Members of Parliament elected to represent the best interest of the “unwashed rabble”, decided to represent the best interests of the incumbent house-owners, and many of them also made a fortune “buying and selling in the marketplace”, particularly Tony Blair.
With regard to the potential “bubbles” around the world, it’s very hard to work out the fundamental without a lot more data than is made available, particularly if you are looking at individual cities.
The reason I claim to be some sort of “expert” on USA house prices is simply that there is better data available than anywhere else, but even with that the model is crude.
The reason I claim to be an “expert” on Dubai is that I’ve been collecting data and building models for nearly twenty years, and even then, it takes easily twenty man-days to get hold of the data and where there isn’t any, build proxies, to update a model.
What I can’t understand is that the data exists, why doesn’t anyone do the analysis?
If you have the data, the other analysis is not hard, even if you argue about the details of the data, even using rough numbers it’s not hard to get within +/- 10% which is “”good enough for government work”. And there is no reason why you can’t do it on a city-by-city basis (I’ve got perfectly respectable models for Dubai that have pretty much worked for fifteen years plus or minus 10%).
You would have thought that given the unpleasantness of the fall-out from the recent housing boom, for all the money they are throwing at the problem, the guardians of the taxpayers money, might consider spending say 1% of what they spend on fire-fighting after the event, on prevention?
But of course, incumbents all make money out of house price bubbles, and if they control the political process, well…who wants to know?
Disclosure: "No Positions"