It seems that every week someone is telling CMHC how to run its business. The latest advice comes from Ben Rabidoux, a well-known CMHC critic.
In the Globe and Mail last week, he suggested four changes to the country’s top mortgage default insurer. (Here is the story)
In some respects, he’s right on the money. In others, well, here’s another take…
Ben’s Recommendation #1: Increase income documentation requirements on insured mortgages
The argument here is that CMHC asks for “far less” documentation than U.S. lenders/insurers. On a typical insured purchase with salaried income, the key documents required by American lenders include:
a credit report
verification of down payment
recent pay stub(s)
purchase & sale agreement(s)
U.S. documentation not mandatory in Canada:
two years’ worth of W2 forms (i.e., Wage and Tax Statements)
verification of other assets via two months of bank/investment statements
verification of timely rent/mortgage payments (usually only for first-time buyers)
Canadian documentation not mandatory in the U.S.:
There are exceptions of course, but two of the key differences are that most Canadian lenders don’t make tax documents and proof of assets (not related to the down payment) mandatory for salaried applicants.
In Canada, salaried income is validated in other ways and employers generally collect taxes at the source. Therefore, requesting tax documents from salaried borrowers would primarily serve to discourage income misrepresentation. But, if someone is willing to fudge a job letter or pay stub, they’re just one more bad decision from faking their tax documents.
As for proof of assets, Canadian lenders ask for it whenever it’s required to support an application. When other loan characteristics are strong, proof of assets generally becomes non-essential because lenders have other ways to judge solvency.
The argument for more documentation rests primarily on two points:
It reduces fraud
It reduces lenders’ incentives to look the other way when underwriting (a moral hazard in the case of insured mortgages).
But these arguments ignore the realities of today’s risk-averse mortgage environment. Specifically:
While “soft fraud” (i.e., document fraud for shelter) is all too common, defaults linked to this type of fraud are still minuscule. One non-bank lender we spoke with said it could count the number it sees on one hand. Hardcore fraud (e.g., organized criminals recruiting straw borrowers) remains a multi-million dollar problem, but again, the total amount of lender losses are too small to justify hassling millions of Canadians with excessive paperwork.
Regardless of whether a mortgage is insured, Canadian lenders have considerable economic incentive to underwrite diligently. (See this)
In the Guideline B-20 era, bank audits are not a rarity. And OSFI regulators are hyper-sensitive to loan documentation quality. (It’s notable that banks finance the vast majority of insured mortgages in this country. And numerous non-bank lenders also rely on bank funding.)
CMHC reviews individual lenders at least every six months to survey performance and track key metrics. Loan quality is monitored monthly. Arrears, negative credit score trends, and the rate that emili refers applications to live underwriters are just a few of the early warning signs CMHC monitors. On-site physical audits take place when CMHC feels that lender performance may be outside of its parameters.
When insurers analyze delinquencies, they compare lenders against one another. (“Early term delinquencies” get additional scrutiny.) No lender wants to rank near the bottom and all lenders are required to maintain full audit trails of their underwriting and documentation reviews.
Insurers can and do deny claims. But their ultimate power is restricting lenders who don’t heed their policies, which can be exceptionally costly for a lender. It can also affect that lender’s ability to obtain competitive mortgage funding going forward.
Underwriters—who generally prefer to remain employed—are held accountable for preventable defaults. Documents and underwriter adjudication are closely scrutinized when insurance claims are made.
Non-bank lenders are routinely audited by their funders, who don’t hesitate to come down hard on lenders who underwrite recklessly. Losing a key investor is a nightmare for non-bank lenders.
When it comes to document fraud, lenders know the stakes. They are continually getting better at fighting it. This is in addition to sophisticated fraud detection algorithms that CMHC runs on every application.
The point of all this is simple. Ample checks and balances are in place to ensure that bad documentation never triggers a Canadian mortgage meltdown. Squeezing more paperwork out of consumers would make the approval process unnecessarily onerous and not lead to any meaningful reduction in default rates…or taxpayer risk.
Recommendation #2: Reinstate the regional mortgage cap
CMHC’s public policy mandate is to “ensure that mortgage financing is available for a range of housing choices.” CMHC is tasked with broadening housing options for all Canadians, not just Canadians at a certain income level. Limiting home buyers to purchasing no more than an average-priced home in a given area is a purely arbitrary decision. The economic and social repercussions of such policy haven’t even been contemplated in Mr. Rabidoux’s treatise, and could easily outweigh the modest risk reduction.
Let’s also not forget: The price of a home being insured, by itself, is almost inconsequential when a mortgage is correctly underwritten. For example, a borrower buying a $200,000 home with a 620 beacon, 42% TDS and 5% down payment can pose a greater expected loss to a lender than a borrower buying a house at double the price with an 800 Beacon, 32% TDS and 10% down.
That said, most would agree that there should be some insurable loan limits. It’s probably not good public policy to insure purchases at double the local price. The question is, with urban sprawl, how do we define the “regional area,” such that communities outside the fringe are not penalized?
And where do you draw the line on price? Do we not insure homes that are 20% above the “average”? 30% above average? One standard deviation above the average?
This is an issue in dire need of perspective. Only 1 in 25 high-ratio mortgages is over $550,000 to begin with. Capping loan amounts further would only marginally alter CMHC’s portfolio risk. But more importantly, doing so would penalize working families. We’re talking about husbands and wives making $60,000 to $75,000 each, for example—not exactly top one-percenters.
Recommendation #3: Eliminate the second home program
This one is much harder to defend. Intuitively, it’s an unnecessary risk to back people who buy second homes with only 5% down. Moreover, as Rabidoux correctly notes, this program is abused.
CMHC’s Second Homes product is such a small part of its portfolio it would not make major waves if it were eliminated. The most current data we know of is that 9% of Canadians own second homes, but the majority could put down 20% or more.
Recommendation #4: Increase transparency and oversight
Rabidoux is bang-on here. CMHC refuses to provide basic portfolio information like average down payment and average loan size on high-ratio mortgages. In this author’s view, there is no justification for withholding such information from the taxpayers who back CMHC. None. The excuse of shrouding it for competitive reasons doesn’t hold water (since insurers already know the general metrics of their competitors).