Wednesday, May 9, 2012

Comparing Canada's CMHC to the U.S. Fannie/Freddie

At the bursting of the U.S. housing bubble in 2006 Fannie/Freddie had $4.3 Trillion in mortgages held or securitized. GDP of the U.S. that year was $13.1 Trillion, for a ratio of GSE holdings/GDP of 33%.

North of the border is Canada where CMHC is bumping up against $600 Billion on mortgages ($540 as of Q3 2011). GDP last year was $1.75 Trillion, for a ratio between 31%-34%.

Those look pretty comparable. Scary comparable.

Fannie/Freddie have used between $190 and $641 Billion in bailout money (depending on how you want to measure it). They just booked a profit of $2.7 Billion this last quarter and Freddie a profit of $577 Million. Only 6 years after the crash began.

By the way, the reason it is hard to measure the bailout is the treasury bought the mortgage securities off of the GSE's books, but they didn't turn out to have zero value. Hence the discrepancy.

Where was I? Oh yeah, those two liabilities look awfully, frighteningly similar. If we assume the same ratio of losses (and given the other stark similarities of: total growth in house prices, consumer debt load, total growth in mortgage debt, this isn't a stretch) That would be $415 Billion (averaging the bailout range) in losses out of $4.3 Trillion in mortgages or 10%. So perhaps CMHC is looking at needing a $56 Billion bailout by the time the market reaches bottom.

A few other things. A lot of Canadian newspapers (especially!) but analysts too, oversimplify what happened in the U.S., to the point of misleading. So, some points I'd like to make here, because they become relevant, post-crash:

1) Fannie and Freddie were not the primary cause of the market bubble in the U.S. The Private Market was. Try on this chart from this UNC report

You will note that Fannie/Freddie were actually blocked from issuing many kinds of mortgages. They dropped significantly out of the market just as the bubble was taking hold. Private Issue Subprime took off. That is your market destroyer, price bubbler, evil dark lord of financial nightmares.

Another view of the same data by total amount, not just percent of market share.
Finally, who was issuing the crap, really? Turns out Fannie/Freddie were not allowed to deal in subprime, although because the executives in this frankensteinian, half privatized, half nudge-nudge wink-wink taxpayer backed monstrosity were losing out on insanely fat bonuses their banking buddies were all getting and were having a serious sad about that, they DID weaken their requirements and started accepting NEAR-prime in an attempt to compete with the insanity.

[That is another thing, the role of executive compensation in the bubble has never been addressed, which is a tragedy. It was a huge factor in the short term thinking of all parties involved in finance, from the heads of the GSEs right down to the loan brokers.]

But back to the garbage, er, delinquency rates. Massive trouble on all scales. But notice the bottom blue line. That is Fannie/Freddie's Overall Book. It barely topped 5% deliquency. Which is horrible! No doubt about it. But it is peanuts compared to Wall Street's fraud machine, which cranked out enough garbage to hit 30% with their subprime issuance. Even their Prime issuance is worse than Fannie/Freddie's overall book.

Back to the topic. Cast your eyes back to the first graph. Notice what happened to the two pink lines after the crash. THAT my friends is why CMHC will not die. It is going to prop up the market, it will BE the market, when everything private label is frozen. Yeah, irony is long dead.


another value investor said...

But, but... Cmhc was involved in 0% down / 40yr mortgage... Fannie/freddie didnt

jesse said...

We will eventually be able to look at CMHC's credit score and LTV distribution to get a sense of how "subprime" CMHC is. I don't think it's as egregious as what the US had; nonetheless high prices are high prices!

Ian Jose said...

Freddie Mac and Fannie Mae also had a cap of 417k for home mortgages. CMHC's cap was eliminated in 2003. Thus, CMHC's losses will likely be far greater than seen south of the border. CMHC is the root cause of the bubble in Canada, but has certainly had help from private institutions.

GG said...

All good points. Fannie/Freddie's books are/were potentially not as bad as CMHC's (we WILL find out) but in terms of scale of exposure relative to the size of the economy, they are virtually identical. Which has got to be a serious concern.

Realize also that Fannie/Freddie's collapse was backdropped by a global collapse. CMHC/Canada will likely go into crisis alone if they do go into crisis. That will help a lot. The global credit markets will not be frozen.

Fannie/Freddie's bailouts were probably higher because it was one of the few levers the government could pull quickly. That will be true for CMHC as well (Like the 40year 0% down mortgages. That in itself was a crisis lever being pulled.)

jesse said...

How is it true that having no cap means more losses? Only if the loan isn't backed by adequate income.

The nature of stated income on CMHC loans is something to consider, especially for those whose stated incomes are tied to real estate transactions and activity. I don't know how the GSEs handled stated loans.

The biggest point I see GG is that the thought that CMHC is going to be truly privatized any time soon is unrealistic. The government has shouldered the liability and will need to maintain that for all existing loans. Further, when prices drop the privates will be backing away from, not increasing, their MI portfolios and CMHC will be last man standing.

Last, and not least, CMHC has provided a direct lever into driving fiscal stimulus into the Canadian economy. To cast CMHC into private hands requires the government find other transmission means for future economic downturns. Turning to Australia's model sounds nice because it uses the word "private" but we don't yet know how much the government will need to intervene if prices in Australia head into the toilet (say if China's GDP hits the skids). My bet is having a path forward for CMHC privatization will be a long process and will be coming into its advent, luckily, just at the time when Australia's model is being stress tested in situ. Fun times!

GG said...

The single biggest default rate in the U.S. is Jumbo Prime. That's in excess of $1 million. No government involvement, no backing, just crazy banks throwing money out the door. The default rate is 40%. Measurably higher than subprime.

In the same area where a tight housing supply will keep a floor under any livable SFH, a home that a wealthy investor dropped 15 million on could end up at 3-4 million. Now, did that investor (foreign or otherwise) really pay all cash and if not, why stick around to pay the mortgage on a near total loss? Especially if you have duel citizenship. (Maybe this is just a bias from how easily people weasel out in the U.S. and how in droves they have decided to.) I just wonder if these kinds of losses might be lurking in CMHCs books.

I brought it up only as a major difference between CMHC and F/F's portfolios. It's an unknown, but it could be a significant one. Could also be nothing.

Ian Jose said...

No lending cap means that the very first loss is experienced by CMHC. In the States, buyers would have a primary mortgage backed by Fannie Mae et al and then a second mortgage with a higher interest rate. The holder of the second mortage experienced the first loss. If a US750k Detroit house sold short for greater than 417k then Fannie Mae et al would experience no loss.

Privatizing CMHC mortgage insurance is not realistic. It has an enormous negative value so no one would buy it. If the Canadian government repudiated the insurance then they would still have to bail out their own banks. The cost to the Canadian tax payer is assured.

GG said...

Ian, your example must be some kind of split mortgage? And the lien above the cap would have been a private lender (very private). Those did happen, but they weren't common and I don't think they were technically allowable, or they were fudged a lot to pass through the system.

If your mortgage is over the F/F cap, F/F are simply unable to purchase your mortgage from your bank. Hence the 1.5% interest premium for the bank to find other funding.

But yes, as you say, CMHC takes the first loss and there is no limit on that. Not good.

jesse said...

"No lending cap means that the very first loss is experienced by CMHC."

Um I think first loss is the borrower, CMHC is next in line. Maybe I don't understand what "first loss" means.

jesse said...

"I brought it up only as a major difference between CMHC and F/F's portfolios."

Yes it's an important difference, I saw some data on high and low ratio loan insurance under force per capita by province. BC, Ontario, and Alberta are all higher than the other provinces on the low ratio segment.

The other nice little nugget CMHC has is its appraisal process. Don't look too much into it if you just ate. It could be nothing, as you say, but there should be concerns with appraisals based solely on recent market values. I remember one guy online claiming to be an appraiser tried to justify appraisal methods and how appraising residential was very difficult because there was no rental equivalence, so had to rely more on market pricing. Hilarious stuff, and is the stuff CMHC gets in the bag if TSHTF.

Ian Jose said...

Second mortgages are quite common in the US, becuase of the 417k cap. The cap was increased in some locations during the crisis, e.g. CA went to 750k. Some of my friends rushed to pay off the 2nd mortgage because of the higher interest rate. The rate is significantly higher, because the private institution has the greatest risk of loss.

In a sense, CMHC won't lose any money at all _if_ they can recoup losses from the borrower. I read a story two year ago where a borrower defaulted on a loan ten years earlier and received notice that CMHC expected him to pay up. However, until the borrower pays then the Canadian tax payers, through CMHC, make the bank whole.

Americans abandoned underwater homes in droves. Will be interesting to see how Canadian react in the same circumstances.

GG said...

Ian, yes, 2nd mortgages (HELOCs) are/were extremely common in the U.S., but those were post-purchase, equity extraction. And yes, it was common for the total lien on a property to exceed the 417k original F/F backed mortgage and F/F got to dip first into the short sale/foreclosure auction money to make themselves whole. That I agree with, but that's not the way the original purchase money was organized, in those two pieces, without working illegitimately around the system.

As soon as the market really bubbled, the F/F cap rules cut them out of the market. Until post crash when the caps were raised, that is. (It was lowered again in 2010). F/F were much less of an encourager of risk taking in the market because of the cap. And that saved taxpayers a bundle.

Raising the CMHC cap will probably turn out to be one of the dumber moves made in the last decade.