« Worrying about the Canadian housing market, Part 2: The effects of higher leverage | Main | Behavioural economics and the Lucas critique »


Feed You can follow this conversation by subscribing to the comment feed for this post.

I think the main risk (aside from financial sector meltdown) would be a sudden increase in savings demands due to household wealth declining as house values fall. Do you not see this as a risk?

That second graph -- Alberta vs ROC -- is very dramatic. But I also remember reading that the unemployment rate is one of the biggest drivers of mortgage defaults. I wonder how much of the Alberta/ROC difference is due to differences in recourse/non-recourse and how much is due to differences in employment? (Or maybe employment, rather than unemployment, since, anecdotally, I think Albertans who lost their jobs handed the keys to the bank then moved out of province back east?)

I just checked: the Alberta unemployment rate stayed below the national average throughout that period, although the increase in the rate - from 3.5% to 11.5% - was greater than the change in the national average (7% to 13%)

Stephen: thanks. I would say those numbers are *consistent with* (rather than *confirm*) my "people moved to Alberta in the 1970's to get jobs and when the jobs went a lot of them moved out again and abandoned the house and mortgage" story. Non-recourse and interprovincial labour migration working together.

(God we forget how bad the early 1980's were. Young people today.....)

A very enjoyable series of posts Stephen, thank you! About Alberta and its difference from the ROC. From what I recall of the early 1980s and what I know of the mid-2000s, in both periods Alberta enjoyed a housing boom which was correlated with a rise in oil prices, which of course was followed by corrections. I wonder if a sharp rise in housing prices somehow correlates with the above ROC rate of mortgage deliquiency in both periods. I can't recall what the numbers were for early 1980s, but the 2000s the rise in housing prices in Alberta did outpace the ROC. Could this sharp rise be an important part of the subsequent default rate?

Great work.

Also, as lending to households is almost wholly collateralized, a reduction in collateral values is the same as a tightening of credit constraints on the household sector, which will cut off credit to borrowers even if the balance sheets of banks are not impaired.

Excellent series, Stephen. I commend the CBA for contributing to this worthwhile Canadian initiative, which honestly helps them.

But still, when a Nobel Laureate in Economics and writer of a standard undergrad econ textbook says that we are in a Depression, yep, that word, then things are bad.


I also note that the effects, experiences and strategies to get out of a supply-side recession (a high inflation slump, as you remember in 1981) are very different from a demand-side depression with little or no inflation (in fact deflation, but a lot of it stayed out of CPI).

Not to argue much with recourse vs non-recourse, but the picture in the US is more muddled than the simplistic non-recourse versus recourse scenario. Florida is an example of a state where excesses and a stark rise in defaults can occur even with the added burden of recourse. There is variance between states on what types of mortgages are considered recourse, and further what type of recourse (for example, judicial). Even Alberta recourse is not without its nuance; certain mortgages are recourse.

My concern with much of the MSM debate regarding contrasting the Canadian situation to the US extends beyond the heterogeneity of the US situation aggregated into one tranche, it ignores that Canada's fate has similarities a few concerning ways, most notably in its debt levels. The problem, in my view, will be in the inability of Canada to react to positive real rates due to its debt encumbrances. The US is about halfway through cleansing itself of high debt levels and Canada will be caught out if it's not careful. In short, concentration on "subprime" and the foibles of the American mortgage market to justify why Canada gets a hall pass may be following the wrong ball.

Is there data on AB house prices going back to the '80's? The stories I hear from people who lived here in AB at the time is that there was a housing bust that coincided with an oil industry bust.

Also, keep in mind that CMHC backed loans are always recourse. Even in AB. If you default, the bank is made whole by CMHC, then CMHC sues to be made whole (if they aren't happy with a short sale or some other remedy that involves fewer legal fees).

The interesting comparison would be with Australia. We have much more intense housing bubbles than Canada that have subsided somewhat, but not burst. But household leverage is very high.

Lorenzo: I agree on the Australia comparison. I've been keeping half an eye on the Australian housing market, for clues about what might happen here.

Are Australian mortgages recourse?

Does Australia have anything similar to CMHC, a government-backed body that insures high loan to value mortgages?

As someone who just got approved for a mortgage and is hesitant to dip their toe into the Canadian housing market I would be very great full for mor posts like these ones. Very informative.

The 90's Toronto housing bubble didn't seem to have a major lasting effect. If I remember, there was a ~30% decline post '90. Some friends that were buying "rental properties" (i.e. house flipping) finally took their losses a few years later. It would be interesting to see the defaults for Toronto during that time, but anecdotally it didn't seem to be a depression (although the poor general economy was problematic). You're right that CMHC will foot the bill nonetheless. But of course, this doesn't answer the other question, is housing in a bubble (i.e. you can still lose money on Nortel, even if it doesn't cause a depression!).

Some short notes:

IIRC, the spike in the number of non-performing mortgages in Alberta in the early 80s reflected the passage of the law making most Alberta mortgages non-recourse. (Alberta also got the double-whammy: first the 82 recession, then the collapse in oil prices 84-85.)

In addition to the above-cited work from UWO, the IMF staff also did some nice work comparing the US and Canadian housing markets. (Couldn't find the references, but I think it was both in the Article IV consultations in 2009 as well as in the working paper series that year.)

In the process of trying to find the above, I stumbled across the Dec. 2011 IMF Article IV Staff Report (http://www.imf.org/external/pubs/ft/scr/2011/cr11364.pdf) and the accompanying Selected Issues paper (http://www.imf.org/external/pubs/ft/scr/2011/cr11365.pdf) both of which have interesting material on the Canadian housing market. The former notes that house price-to-rent and house price-to-income ratios are at historic highs. It also talks about the need to ensure good governance and supervision of CMHC given its role in backstopping the financial sector.

The Issues paper is more fun, as it contains the IMF Staff's attempt to estimate a model of "fundamental" prices for Canadian housing. The punchline (see graph on p. 9) is about a 10% overvaluation in QC and ON, maybe 30% in BC, but AB may be undervalued(!) BTW, their model of fundamentals can't find a role for mortgage rates or borrowing capacity in influencing house prices; they instead find that employment, real income, commodity prices and immigration are key.

i think there is way too much emphasis on non-recourse and strategic defaults - the vast majority of homeowners pay their mortgage regardless for a variety of economic and non-economic reasons (like they like the neighborhood and schools) etc.

The vast majority of homeowners walk away because they *cant* pay not because they are unwilling and so-called recourse states in the US saw high defaults just the same.

The keys to avoiding a housing bubble are: lending standards (LTV, debt-income) and stabilizing nominal gdp.

Also, i strongly disagree with "Of course, another important main reason that we wouldn't see a US-style financial crisis is that banks are holding CMHC-insured mortgages."

no: in the US most mortgages are guaranteed by a government sponsored entity. Prime mortgages got into just as much trouble after monetary policy tightened.

overall, this post has too much of the "this time its different" flavor. Canada has seen real estate booms and busts before. delinquency rates will rise. home prices will drop. Banks may or may not be fine depending on how much capital they hold and how good the LTV ratios are, and whether the central bank allows NGDP to drop. It can happen anywhere, dont take any solace or false comfort whatsoever in the structure of the Canadian housing market.

As long as lending standards remain solid there is not that much to fear from finance and debt, but should resource prices collapse you will be in for a world of hurt. That is the bane of resource based economies.

Re: Lord's comment and those earlier that noted that the oil price drop coincided with housing value decline in Calgary.

Could we say that, in contrast to the US of 2007-2008, in Canada the economy is a risk to the housing market rather than the other way around?

That sounds right. In the US, people started falling behind in their payments a year before unemployment started going up.

Most US states have recourse (Table 1, page 32 http://www.fhfa.gov/webfiles/15051/website_ghent.pdf). Let's not sugarcoat how much fun default is for US borrowers. In Canada, even outside Alberta, borrowers with little other assets (stocks, bonds, other properties) can only have their actual home seized. Income is garnished for a fairly short period (less than 3 years) - even then leaving the bankrupt with a modest income to live on (perhaps they only had modest income to begin with...). I'm not saying this is fun, but nor was it fun in the US. The point is, "recourse to what?"

no: in the US most mortgages are guaranteed by a government sponsored entity. Prime mortgages got into just as much trouble after monetary policy tightened.

I invoke Warren Buffett here, he was a longtime investor in both Fannie Mae and Freddie Mac and his company also owns a Savings & Loan. His comments during the Savings & Loan scandal were acerbic.

The problem in the US was that Fannie Mae & Freddie Mac were not explicitly backed by the US Government, they had federal charters but did not legally have the backing of the full faith and credit of the US Government. They were private corporations with private shareholders. The market long treated them as being almost as good as Uncle Sam and probably backed by Uncle Sam, but there was no legal connection. Warren Buffett pointed this fact out repeatedly in public.

When all hell broke loose in 2008 Fannie & Freddie went bankrupt and were placed in "conservatorship", the market called the implicit government backing and it turned out to be explicit. Fannie & Freddie were bailed out by the US Treasury. Fannie & Freddie turned out to be off-balance sheet operations and obligations of the US Government.

That isn't the same in Canada. CMHC is and has always been a Crown Corporation, it has one shareholder, the Government of Canada. Its debts have always been obligations of the Government of Canada and its paper has traded on this fact honestly. CMHC's obligations and financial results have always consolidated with those of the Government of Canada via the Department of Finance. CHMC's government backing has long been explicit and clear, it has never been off the Federal Government's balance sheet. The failure mechanism that would transfer defaulted mortgages to the Government of Canada has long been clear. It went a long way to keeping lending standards up, through the interests of the banks, CMHC and Finance.

On the subject of Alberta recourse - when my mortgage was issued, I was told that CMHC-insured mortgages (being pretty much all high leverage and therefore at-risk mortgages) are subject to recourse. I believe this is because CMHC mortgages are governed by the federal National Housing Act, rather than the provincial Law of Property Act.

I don't know if this has changed since the early '80s, but I would guess not, and that recourse isn't the driving factor behind Alberta exceptionalism in your graph. As noted in other comments, the number of employed Albertans becoming unemployed was substantially higher than average at the time. It would also be relevant to see what happened to national home prices at the time, since my guess would be the price decline was much more precipitous in Alberta as well.

Mortgages given to corporations (including cases where a natural person has later assumed the mortgage) are also full recourse.

Oil sands pipelines are being built with 100 year Bonds. A scaleable alternative will be avaiable in one to 1.5 decades. Hate to be repetitive but (evil) idiots will thank me later...

On non-recourse mortgages in Alberta, note the following:

Important Alberta provisions that were found intra uires or not disallowed during those years included amendments to the Judicature Act giving the Supreme Court the power to grant a stay of exe~utiont;h~e~ enactment of a redemption period of one year (but with power in the court to extend or decrease the time);54the rendering void of attornment
clause^;^ and as perhaps the most permanent legacy, the restriction (in 1939)of the lender's right to the land itself and the prohibition against an
action on a covenant for payment contained in the mortgage.56The wheel had turned full circle from World War I.

The above is fromt his reference:

Mortgages in Alberta have been non-recourse long before the 80's.

"What strikes me about that graph is that even when mortgage payments were increasing by more than 50%, more than 99% of homeowners with recourse mortgages did whatever they had to do to continue making their payments."

Nothing striking in that chart, as this one explains why homeowners continued to pay their mortgages during that time. http://postimage.org/image/eujhtk16f/

As rates went up—so did interest on savings and bonds—keeping purchasing power in equilibrium. Today's situation is much different being that homeowners are trying to pay down their mortgage with a currency that is being devalued by the BoC.

Reference points are 1930s and 1940s, not 1980s.

i think you are missing the point. If Canadian home prices were to drop 30% and unemployment rise to 10%, then many bank assets would be toxic. Some of them (mortgages) might be insured by the government (which the taxpayers would subsequently have to bail out), but in no way does that ensure the overall solvency of the banks - that depends on what else they own and how much capital they hold. The best prescription to prevent this is: 1) to prevent unemployment from rising so far to begin with (i.e. stabilize nominal gdp), because delinquencies only exacerbate the home price decline; ensure sound lending standards with high ltvs; 3)ensure adequate bank capital to withstand a potentially severe downturn.

Unemployment peaked in Canada at 8.5% this cycle. Mortgage insurance in Canada means that defaulted mortgages are a liability of CMHC/the Government of Canada, not banks. Further, bank experience going back to the 1950's shows that default risk is nearly entirely concentrated in those borrowers with LTV ratios above 80%.

Second, unlike the US in Canada mortgage interest rates float during the amortization of the loan, they are not fixed for 30 years, they are only fixed to five years generally. Then the rate is reset at the prevailing market rate. Every mortgage borrower in Canada knows this. The rate resets which caused so much trouble in the US are not a problem here because that is the way the industry operates and everyone knows it.

i think you are missing the point. If Canadian home prices were to drop 30% and unemployment rise to 10%, then many bank assets would be toxic. Some of them (mortgages) might be insured by the government (which the taxpayers would subsequently have to bail out), but in no way does that ensure the overall solvency of the banks - that depends on what else they own and how much capital they hold.

Nonsense. That happened in the 1980's and the banks were fine. Canadian banks have much, much better loan structure practices than US banks. They also have less maturity transformation (see mortgages). Mortgages form 60% of the chartered banks balance sheets, either of very prime quality or insured by the full faith and credit of the Government of Canada. There is very low possibility of capital erosion here because the default suspects are replaced with government money.

It is also a mandate of the OSFI (bank regulator) that bank's investment activities cannot rise to such a level as to seriously jeopardize a bank's financial health. It is also the practice in Canada for banks to have the highest Tier 1 capital ratios in the world (12%) and the ratio was not lowered to cause a boom in the early 2000's. No balance sheet erosion, no balance sheet recession.

The US problem was not with insured mortgages, it was with junky mortgage-based derivatives held by large financial institutions. When these derivatives lost value, all hell broke loose. In Canada, the OSFI won't let banks hold or issue those junk derivatives, the capital ratios of Canadian banks are very high to absorb losses and the mortgage market in Canada is concentrated in highly regulated banks who didn't make bad loans (The Scotia's flaky 100% mortgage was withdrawn after the OSFI thought it a bad idea).

Even better, it is still the practice in Canada for banks to keep the mortgages they issue on their own books. The mortgage-based security sector in Canada is very small.

"The US problem was not with insured mortgages, ..." You're right that the problem did not start with insured mortgages. But when the mortgage insurers go bankrupt, is it still reasonable to claim that there was no problem with insured mortgages?

"It is also a mandate of the OSFI (bank regulator) that bank's investment activities cannot rise to such a level as to seriously jeopardize a bank's financial health."
Mandates are funny things; the Fed has a mandate to maintain full employment, for example. It is often useful to look beyond mandates and consider the success in executing the mandate.

The derivative security market in Canada was ramping up nicely until the crisis hit. Anyone remember ABCP? Anyone remember the lawsuits brought against chartered banks for selling complex derivative products to customers as "safe" or "AAA" fix-income products? Carney was instrumental in negotiating solutions to that crisis, which involved both important haircuts for investors and buybacks from many chartered banks.

My point is that Canada was not sheltered from the global drive to introduce complex derivatives that were (1) barely regulated, (2) poorly understood, (3) easily abused, (4) highly profitable for a while for some market leaders. When they blew up, we happily discovered that they had not yet grown big enough to wreck our economy. There are still many keen to ensure a fertile environment for "innovative financial products" and blunt any attempt to regulate them.

Yep. If the US had managed to stagger on for a couple more years before things blew up, our recession may have not been as mild as it turned out to be.

Yep. If the US had managed to stagger on for a couple more years before things blew up, our recession may have not been as mild as it turned out to be.

double that.

These were really interesting postings by Stephen Gordon.
Regarding the C.D. Howe Commentary by Jim Magee, I was not quite as impressed with it as Stephen Gordon. Like just about everyone in North America, Mr. Magee completely ignores the important issue of what inflation indicator the central bank is monitoring. Three countries in the Anglo-Saxon world had housing bubbles in the second half of the nought decade. Two of them, Ireland and the United Kingdom had central banks that were monitoring consumer price series that had no component for owner-occupied housing (OOH), while the United States used the personal expenditure price index with an imputed rentapproach to OOH, which is quite useless for protecting an economy from housing bubbles. Australia and New Zealand both have CPIs with a net acquisitions approach to OOH, which is the best approach. Neither had a housing bubble, although both economies are vulnerable to the kind of strong inflows of foreign funds that can stimulate asset bubbles. The Bank of Canada targets the CPI All-items, a quite inferior inflation indicator, but at least its core CPI measure excludes mortgage interest. The core measure has sort of a poor man’s approximation of a net acquisitions approach to OOH, since mortgage interest is out of it, but current house prices are included via the housing depreciation component. So Bank of Canada monetary policy should be more sensitive to upsurges in housing prices than US Fed monetary policy, although in Canada’s case it seems that the Department of Finance under Jim Flaherty has done all the heavy lifting to protect us from a housing bubble and Mark Carney has been more or less an idle bystander.
I noticed that Magee’s chart of American and Canadian housing prices shows the excellent Teranet Natonal Bank housing price index (HPI), which was modeled on the Case-Shiller HPI that is used for the US in the chart. It shows just what a shoddy measure of inflation the Bank of Canada is prepared to put up with that the Teranet-National Bank HPI is not used anywhere in the series it uses to measure Canadian inflation: only the new housing price index data are used to measure the CPI for owned accommodation. This is true both of the mortgage interest cost component, even though most mortgage interest is paid on existing homes, and of the component for real estate commissions, even though virtually all real estate commissions are paid on sales of existing homes.

The comments to this entry are closed.

Search this site

  • Google

Blog powered by Typepad