Blessed by the Potato

If you had asked me in 2011 or so, I would have gone on at length about how critical CMHC (and Genworth) mortgage insurance was to fueling the housing bubble in Toronto and Vancouver. Heck, you didn’t even have to ask me, I would have told you anyway. Having to save up a downpayment helps to make the system more robust, and acts as a brake when house prices start to rocket higher as people can’t save fast enough and get priced out. Mortgage insurance circumvents that, and lots of people were buying with minimal downpayments, thus it had to have been a big causal factor in the bubble.

In 2012 some reforms were announced, including that mortgage insurance would be capped at $1M house value. It was no 10% downpayment minimum like I was hoping for, but I thought that would help cool the market by providing an upper limit to the insanity (at that point the average Vancouver detached house had already crossed $1M, and the average Toronto one was a bit shy, but many neighbourhoods were over).

For a brief while you could see the strange behaviour as lots of houses that may have fetched prices ranging anywhere from $800k to $1.1M were all compressed into a narrow band near $990k. I thought a few more months of that and a few people with downpayments would push a bunch over $1M, and those without downpayments would have to exit the market, and the correction would start. I was so wrong. Before you could even take MLS screenshots and write a blog post on the strangeness, prices started rocketing higher again, and $1M was no kind of barrier at all.

And then the truly puzzling stats started coming in: Canada’s two most bubbliest cities had the highest downpayments. Roughly two-thirds of buyers across the country were tapping CMHC, but it was Toronto and Vancouver that were pulling the average downpayment up.

The bank of mom & dad was by and large saving the day. CMHC insurance became but a minor factor in Toronto and Vancouver’s housing bubble.

This is troubling because it meant that things had gotten away from the government’s ability to control via CMHC reform unless they went nuclear (i.e. 10-15% downpayments, not this wimpy sliding scale tweaking stuff — more on this below). It’s also troubling because many a First National Bank of Mom and Dad gets its financing from HELOCs, and that’s very pro-cyclical — it’s easy to get a HELOC while prices are rising (and indeed, you can fool yourself into thinking you have to and that it’s good for your kid to do so), but that source gets turned off when the bubble bursts, making corrections worse. It also masks the vulnerabilities in the system, making it look like we have a bunch of borrowers taking out ~20% of their equity and some buying new with ~20% down when really it’s more like a bunch with paid-off houses and a bunch with nothing (the total equity may be the same in both cases, but the latter group is much more likely to blow up).

[And to add, the other answer is “foreign money” which may be a bigger component of the market than I thought, but still doesn’t change the answer as to whether you should buy or rent — as we’re now seeing]

So that brings us to today, with some new rules from the Finance Minister. What’s interesting here is that a lot of the previous rule-tightening moves for CHMC didn’t apply to insurance that the banks took out on mortgages with over 20% down.

Aside: Why would they do that? So that they can securitize the loans. From the data I can find, roughly a third of all mortgages issued in Canada end up insured and securitized.

Now, loans used for portfolio insurance must meet the same criteria, plus the new criteria, particularly the closing the 5-year filter on qualifying rates.

Aside: a while ago the rules were changed to try to be more conservative so borrowers had to qualify at a higher rate — based on posted rates — than what they were to pay, to ensure that buyers had the financial flexibility to take on higher interest rates. However, if you went for a 5-year term you didn’t have to go through this check and could just use your contracted rate. This pushed many people into getting 5-year fixed terms (vs. variable-rate mortgages), and created incentive for the banks to make their 5-year fixed discounted rates their most intense point of competition.

So now there’s going to be an effective $1M cap on securitized mortgages (it’s possible to securitize and sell mortgages without insuring them but that largely doesn’t happen because Bad Things Happened and the market for that kind of product is dead — lookup ABCP). It will likely also spell the end of longer-than-25-year amortizations (which could still be had for people with more than 20% down).

That means any bank making a loan in Toronto or Vancouver where so many places go for over $1M is going to have to keep that jumbo loan on their books. No more moral hazard from passing it off in a securitization.

We’ll see how these changes affect the market in the coming months and years. Maybe this less-obvious change will have big effects. Maybe the market is already rolling over so it won’t matter. Maybe the meme is broken.

For the qualifying rates, the difference can be somewhat meaningful for those who are stretching to the limit. If you make $100,000/yr and can borrow up to a GDS of 39%, that means your maximum monthly payment for the mortgage and a few other costs (tax, heat) is $39,000/yr ($3,250/mo). If you take off say $400/mo for heat and property tax, that leaves you with a maximum mortgage payment of $2,850 on that income. With the 5-year filter you can use the actual rock-bottom rate of 2.5% to qualify, letting you borrow about $630k. At the qualifying rate of 4.6%, that takes you down to $510k — a fair bit less room to reach.

There are few ways the banks can respond to this. They can reign in lending (clearly the intended approach). They can say fuck it and put the pedal to the metal and just keep all the loans on their books. For the qualifying rate, that comes from the “posted” rates the banks put up. It’s not likely that they will lower the posted rates to circumvent the stress tests (those higher rates help them rake in interest rate differential fees when people break their mortgages and it’s low-hanging fruit to get people to renew at higher-than-market rates if they don’t shop around), but that is an option. Of course in that case the government can just specify a qualifying rate.

Also today Preet released his interview with Ben Rabidoux which is a good listen that touches on many of these issues (though it was recorded months ago).

An important topic they discuss is the rise of private mortgages (aka shadow banking). Now, an incredibly, impossibly stupid thing has been allowed to happen: people are expected to put no-condition offers on places in bidding wars in our hot housing markets. So every now and then (though surprisingly less often than I would have thought) someone can’t actually get prime financing like they planned, so they get a private mortgage to cover the gap. Or someone will get into trouble, and tap the private mortgage market.

Every time the government tries to tweak prime banking and CMHC just a little bit it does not work to cool the market, because the changes are too marginal. Instead they push a few more people into borrowing from mom & dad or private lenders, and normalize that whole system and alternate ways of qualifying. This approach is doing less to cool the market than it is to build the shadow banking infrastructure. That’s part of why¹ I think we shouldn’t keep trying a series of small nudges to CMHC and lending criteria, and letting the system build resistance to regulation. We’ve got to rip the band-aid off and punch the speculators in the nose.

1. The other part being that I am actually Ra’s al Ghul and believe that the bubble can’t be saved so we should just get on with purging the city.