I have been checking almost daily for the Teranet-National Bank May house price data, and now it's out. Their national composite price index increased 0.72% from April to May. This confirms the less accurate and more anecdotal data we've been hearing about the Canadian housing market improving this Summer.
The big difference between the Teranet and other data sources is that the Teranet index is not skewed by changes in the composition of houses sold. For example, if one month the sales were mostly of upmarket houses in pricier areas, it would look like average house prices had increased, even if the price of every individual house stayed the same. Teranet gets round this by looking at repeat sales of the same house. Plus a lot of econometrics.
I am now more confident that the apparent bounce in house prices in June is also real. And if I am correct in my impression that first-time buyers are a bigger proportion than usual, and if first time buyers tend to want to buy smaller and cheaper houses than average, the composition effect would tend to bias reported average prices downward.
The one problem with Teranet (aside from slowness to report) is that it reports data on 6 big cities only. I don't know how much it matters that smaller towns and rural areas are excluded from the national index. And they can't do much about it, since their method is data-hungry, and the only data they can use is when the same house sells twice in a few years. (And I feel churlish even making this criticism.)
First derivative. Prices in Halifax, Montreal, Ottawa, and Toronto are all up in May over April. Prices in Calgary and Vancouver are down. The sample size doesn't warrant me saying this: but the ones that rose the most in the past few years are the ones still falling in May.
Second derivative. (A positive first derivative means prices are rising; a negative first derivative means prices are falling. A positive second derivative means prices are rising faster than before, or falling more slowly than before, or rising where previously they were falling.) Just from eyeballing the graphs, the second derivative is positive everywhere except Halifax and Calgary. In Halifax the turnaround came earlier, and prices are continuing to rise at about the same rate. In Calgary the second derivative is negative; prices are falling faster than before.
What's the deal in Calgary? The answer cannot be "things are bad in Calgary". We need to know why things got worse in May, and so much worse that they offset the upward effect of things getting better in May nationally.
Will recovery continue? I don't know, obviously. Lower interest rates, and generally improved conditions in financial markets (lower spreads etc.) are the most obvious causes of the recovery in house prices. I don't see interest rates staying this low forever, but they might stay lower than was previously thought normal, if global savings stay high. And I think global savings will stay high, mainly because some borrowers will have been scared by recent events. Fear of unemployment would make people less willing to borrow to buy a house, and unemployment may continue rising for a bit even if the Bank of Canada is right about the recession being over (so output stops falling). But it's the fear of future unemployment, not the current unemployment rate itself, that matters. My guess is that fear is diminishing, even if unemployment continues to rise for a bit. But if you think that house prices are still above fundamentals, and the bubble hasn't deflated yet, none of that will matter in the long run.
If I had to make a prediction (though I would really rather stay on the sidelines), I would guess that Canadian house prices nationally were a bit above fundamentals before the financial crisis, but the fall in prices, and the rise in fundamentals (with lower interest rates), have now brought prices and fundamentals roughly together. I don't expect any big changes in either direction going forward. Assume 2% inflation in house prices on average. This assumption will be false, of course. But I can't tell in which direction.
In Calgary new house construction may have been way above fundamental levels in recent years while prices were going crazy. I think it's more fundamental S&D issues for housing rather than general economic conditions. The question I have is why is the second derivative in Vancouver turning positive? And I wish they had a measure for Saskatoon or Regina, story there is usually a good benchmark to measure Calgary.
Posted by: Mark | July 29, 2009 at 04:15 PM
Mark: I don't find Vancouver's positive second derivative surprising. Most cities (and the composite index) had a positive second derivative. It's Calgary, with the negative second derivative, that's surprising. Despite the low interest rates, Calgary prices are falling even more quickly than before.
Posted by: Nick Rowe | July 29, 2009 at 05:15 PM
I live in AB. Some anecdotal observations:
1. Calgary had a big run-up in prices, so maybe it just has farther to fall?
2. Mid way through during the oil run-up (which really started almost 10 years ago) there was a general sense that Calgary had replaced Toronto as the center of the universe (esp. when their guy became PM) and people starting consuming like crazy and having lots of babies.
3. Calgary has become a head office town. Competitive consumption became the official sport.
4. Bankruptcies in AB are a problem:
http://www.bankruptcycanada.com/bankstats1.htm
Posted by: Patrick | July 30, 2009 at 12:39 AM
"I would guess that Canadian house prices nationally were a bit above fundamentals before the financial crisis, but the fall in prices, and the rise in fundamentals (with lower interest rates), have now brought prices and fundamentals roughly together."
I'm not sure we've seen any rise in fundamentals. Incomes are down, unemployment is up, debt levels are up, bankruptcies are up, there's downward pressure on rents, supply continues to be at or near record levels (although starts have come down, which will help in a few years) etc. Not exactly what most people would call improving fundamentals in the housing market.
At any rate, Vancouver (and B.C. in general) is still way overpriced by just about any metric you can think of (even with interest rates as low as its possible for them to get, it's still far more advantageous financially to rent in most cases - unless you assume significant further real price appreciation, of course!), so if the country is balanced as a whole, there must be some part of it that is underpriced...
Posted by: Declan | July 30, 2009 at 01:10 AM
Declan: Here's my post from back in November trying to get a fix on house prices compared to fundamentals: http://worthwhile.typepad.com/worthwhile_canadian_initi/2008/11/are-canadian-houses-over-priced-a-revised-estimate.html
I like to look at (annual rents minus taxes minus maintenance)/price, to get a real rate of return on owning a house vs. renting, add the expected rate of inflation on rents, then compare that to mortgage interest rates (if you are borrowing all the money) or after tax rates of interest (if you have a substantial downpayment).
The variable that really matters in that calculation is the rate of interest (the results are really sensitive to what you assume). I think interest rates will be lower over the next 10 years than they were over the past 10 years.
Also, with annual rent/price ratios around 5% at the peak, the roughly 10% drop in house prices since the peak would increase the rate of return by 0.5% points, to 5.5%.
So it's just not obvious to me that house prices are above fundamentals.
My best guess is that owning a house is not a good deal if you have to finance nearly all of it by borrowing, but is a good deal if you can finance a lot of it out of your own savings (because you don't pay tax on the implied rental income).
What's your sense of rent/price ratios?
Posted by: Nick Rowe | July 30, 2009 at 08:57 AM
Patrick: but it's still surprising that there's no sign of the effect of lower interest rates and improving financial markets on Calgary pricing. Maybe you could argue that those effects you cite would have made the second derivative even more negative in Calgary?
Posted by: Nick Rowe | July 30, 2009 at 09:01 AM
I had to think hard about what I'd argue and didn't really come up with anything coherent. Animal spirits trumping cheap money is about the only thing that comes to mind. There was a huge boom in Calgary, people had a very rosy outlook and consumed like crazy. Lots of people moved to the city. Now there is some unemployment, lots of BKs, oil prices are down, and oil companies aren't spending like drunken sailors. So maybe it's animal spirits. People are uncertain about Calgary's prospects so people aren't moving to the city, nor are they upgrading. And lots of people are in over their heads so they are trying to sell.
Posted by: Patrick | July 30, 2009 at 09:41 AM
Nick: I have to support Patrick in some sense in that it's hard to argue from a price/rent context. Housing in Calgary over the last few years has been predicated on strong in-migration due to booming employment and income growth where demand far outstripped supply and pushed prices higher, further expanding housing starts. With the economy stagnating there is likely to be a halt to in-migration and perhaps even out-migration since an unemployed worker in Calgary who is originally from Windsor might be happier being unemployed closer to home. Relative to other cities there is probably an oversupply of housing in Calgary and this may be holding down the second derivative.
The only city that seems to have matched Calgary in terms of frothiness was Vancouver. With lumber and nat gas in the dumps and Olympic construction winding down I'd expect some further price corrections. Many people have used zoned-land argument for Vancouver vs Calgary but this point just makes me more worried, a la San Fran vs. Atlanta.
Posted by: Mark | July 30, 2009 at 12:47 PM
"What's your sense of rent/price ratios?"
Well Forbes, in ranking Vancouver as the 6th most overpriced real estate market in the world (behind Monaco, Rome, Paris, Madrid and L.A. - I suspect we have since passed L.A.) back in 2007, estimated the rental yield at 3.19%, commenting that it was 'one of the lowest rental yields of any city measured' [in the entire world!]
The yield was probably below 3 at the peak and the price drop we've had since then has probably put it back about where it was when Forbes did their measuring. That's for downtown, I suspect the yield is worse in the 'burbs and in Victoria, Kelowna, etc.
Rental yield is a decent measure, but not really sufficient on its own. Supply and demand is really the underlying, determining factor. One of the things you have to be careful of is a bubble creating correlation between house prices and rents. For example, if the two biggest industries in town are residential construction and mortgage broker, rents may fall along with house prices. To the extent that local spending is supported by the wealth effect of high house prices (either perceived or real via HELOC's) a downturn in prices will again put downward pressure on rents. In addition, if units are being held empty based on hope of appreciation, a sustained downturn may cause these units to be put back into the rental pool, again putting downward pressure on rents and prices. Finally, to the extent that falling prices happen along with economic downturn, household formation will fall (people move in together to save money) putting further downward pressure on prices and rents.
For example, the New York times had an article today on Miami (a very similar market to Vancouver - port city, constrained land situation, history of bubbly property market, lots of drug money and laundering, similar median income) and how as prices per sq. ft have dropped to 200 sq. ft. (less than 1/3 of the equivalent price in Vancouver!), rents have also dropped by 15-20%.
Vancouver didn't have quite the same crazy building spree as they did in Miami but still, there was a fair bit here too.
---
Also, with respect to your calculation, the thing about mortgage rates and rent appreciation is that you should be able to factor inflation out of both of them and ignore that part, leaving just the premium on the mortgage for default risk, interest rate risk and servicing costs/overhead.
Posted by: Declan | July 30, 2009 at 10:06 PM
Declan: Agreed: we could subtract inflation from both interest rates and rent price increases, so we compare the real rate of return on housing to the real rate of interest.
The way I usually think about "supply and demand" affecting the fundamental value is this: the (flow) supply of new houses built augments the stock supply of houses available to be lived in, together with the demand for some where to live, jointly determine the time-path of equilibrium rents. That time path of rents, and the rate of interest, jointly determine the fundamental value of house prices.
3% sounds low for a rent/price ratio. And all my instincts tell me that means prices are above fundamentals.
But what really bothers me is that I still cannot make a cast iron argument that prices are above fundamentals. Suppose nominal interest rates stay at 5% (they might). And inflation is 2%. That means real interest rates are 3% too, so prices are exactly at fundamentals.
Now, the 3% rent/price ratio ignores municipal taxes and maintenance. But on the other side there's income tax on interest income. And it's not impossible that rising real incomes and rising population might mean rising real rents in the most desirable locations. Say 1% growth in real rents.
Then there's risk. But my "born with a short position" post argues that the equilibrium rate of return on housing should actually be lower than on perfectly safe assets, since it covers your short position, and so hedges risk of rental uncertainty.
So even though I agree that Vancouver house prices are probably above fundamentals, I can't prove it.
I draw the following implications from this:
1. At least you and I know in principle how to calculate a fundamental value; but most people, even educated people, don't. So they couldn't calculate it even if they had perfect foresight on rents and interest rates. So it's not at all surprising that prices should wander away from fundamentals.
2. I don't see how anyone, even smart economists at central banks, could really identify a bubble with any degree of certainty, unless prices were very very different from fundamentals.
Dark forces of time and ignorance etc.. Depressing how ignorant we are.
Posted by: Nick Rowe | July 31, 2009 at 06:27 AM
>>"supply and demand" affecting the fundamental value
Supply and demand is not fundamental. During the NASDAQ bubble, demand outstripped supply!
>>Suppose nominal interest rates stay at 5% (they might). And inflation is 2%. That means real interest rates are 3% too, so prices are exactly at fundamentals."
If you assume inflation is even across the board.... then yes. With deleveraging, I would argue things that you buy with debt will deflate, regardless of the CPI.
>>I don't see how anyone, even smart economists at central banks, could really identify a bubble with any degree of certainty, unless prices were very very different from fundamentals.
it's more like they need to have bubbles to prop up the economy.... We've been living beyond our means for the last 20 years or so -- if you look at the debt to income levels
>>born with a short position
I don't think buying a house flattens out your position unless you buy the house right out with cash. Even then, you have risk in property taxes, utilities, repairs and maintenance (above and beyond your mortgage interest). How about those leaky condos, eh? With the law favoring the tenant (and the bubbly RE market), I would argue that rent is less risk than owning.
If housing is supported by fundamentals, then why have the commercial banks NOT increased their exposure to residential mortgages from 2007 to 2009. Instead, they chose to securitized most of the mortgages with guaranteeds from CMHC and gov't of Canada!
http://1.bp.blogspot.com/_0YOsyi5WbLY/SmTHCjXQhFI/AAAAAAAAAGM/SAJ6_qxTWGE/s1600-h/Mortgage+Credit+Outstanding+(June09).Bmp
Posted by: mike | August 02, 2009 at 10:41 AM