11/10/2024
Changes to the stress test on the horizon? Maybe even scrapping it?
Peter Routledge and OSFI recently put out their semi-annual risk outlook. I've put together what I think are key points regarding how OSFI views mortgage risk:
1. However, there are two challenges with respect to the MQR. The first is that it really didn't stop a very substantial build-up in mortgages with very high loan-to-income ratios, which I define as 450% loan-to-income and greater.
2. ...during COVID that big block of very highly leveraged mortgages also happened to be weighted towards variable rate products that had fixed payment components to it, which we see as a manageable but serious risk concentration.
3. The second challenge is by virtue of the way it is implemented, which is at the institution-to-borrower level. It feels to the borrower like it's OSFI regulating them individually.
4. OSFI's job is to regulate financial institutions, not Canadians. But the MQR creates that perception and I think that’s a challenge to it.
The loan-to-income test is pretty simple. It says every quarter a bank or a lender can only lend 15% of their mortgages in that quarter to borrowers with loan-to-incomes higher than 450%. That is a very effective ceiling that stops the build of risk concentration.
5. When we looked at that and tested it out, we learned that a single 15% ceiling or 20% ceiling across the industry wouldn't work. We determined we would do it on a bespoke model. We would go institution by institution and calibrate it to those business models.
6. We're going to test it next year and if it works the way we want, and we'll probably have to tighten or loosen the bolts here and there, we expect it be a legitimate alternative or a legitimate complement to the MQR. We'll make that decision after we have a full year of testing to make sure if we do anything, we do it right.