So what are you really saying OSFI? Taking a look at their latest mortgage underwriting letter

Last week OSFI released a letter about mortgage underwriting expectations.  I’m going to walk through the important parts of the letter and also pose some questions that this letter evokes.

The main ideas conveyed in the letter are quite sensible; lenders need to improve their underwriting and increase their capital levels.  Further to this letter, OSFI suggested a few concrete proposals, the main one being that the eligibility of borrowers should be tested at mortgage rates 2% higher than current mortgage rates.

The letter, while a step in the right direction, does highlight that OSFI is aware of disturbing behaviour in the mortgage market.  Below is a breakdown of sections of the letter that I thought were important.

OSFI, like myself, is concerned about future loan losses (emphasis mine):

“Persistently low interest rates, record levels of household indebtedness, and rapid increases in house prices in certain areas of Canada (such as Greater Vancouver and Toronto), could generate significant loan losses if economic conditions deteriorate.”

OSFI wants better income verification (emphasis is mine):

“OSFI is aware of incidents where financial institutions have encountered misrepresentation of income and/or employment. Institutions should have adequate processes and controls in place to mitigate this type of risk.”

OSFI wants lenders to not rely on LTV alone when making mortgage loans (emphasis is mine):

“Lenders should not rely on collateral values as a replacement for income verification, especially in areas of Canada where house prices have been rising rapidly.”

OSFI is observing that alternative lenders have poorer underwriting practices compared to bigger banks.  (As an aside, alternative lenders, like HCG and EQB, actually have less loan loss reserves than the bigger banks). (emphasis is mine):

“OSFI has observed that underwriting practices for residential mortgages at or below the non-conforming threshold of 65 percent LTV are often not as strong as those for conforming mortgages, especially in regard to income verification.

The 65 percent LTV threshold used in OSFI Guideline B-20 should not be used as a demarcation point below which sound underwriting practices and borrower due diligence do not apply.”

They also emphasize that lenders should not treat all mortgages the same, as different loans have different risk characteristics (emphasis is mine):

“Certain mortgages inherently carry greater credit risk regardless of LTV. These include, for example, mortgages for properties purchased for investment purposes, or properties that are partly reliant on income from the property to service the loans. Institutions should identify residential mortgages exhibiting higher risk characteristics and adequately mitigate those risks through greater due diligence of the borrower and stronger internal controls.”

In general, OSFI is saying to lending institutions just be better at your job (emphasis is mine):

“It is important that lenders and mortgage insurers revisit their Residential Mortgage Underwriting Policy (RMUP) and Residential Mortgage Insurance Underwriting Plan (RMIUP), respectively, on a regular basis to ensure that there is strong alignment between their stated risk appetite (as approved by their Board of Directors) and their actual mortgage/mortgage insurance underwriting and risk management practices.”

I want to say I applaud OSFI for trying to make some well needed adjustments.  But it is hard to read this letter and not think that OSFI is not doing enough.  They are clearly aware of some very poor behaviour.  So here are my lingering questions.

It is obvious that OSFI knows that some lenders are underwriting really risky loans.  Who are these lenders?  I think investors and taxpayers have the right to know.  

OSFI really seems to be emphasizing the basics in lending. Why do lenders need to be reminded of the basics?  How bad is this income verification and reliance on LTV problem?

Canadian lenders have low levels of loss reserves.  If some institutions are engaging in riskier lending it should be reflected in their loss reserves, but looking at the financials it isn’t obvious to me who they are.  If OSFI has identified poor underwriting institutions, shouldn’t these companies be forced to increase loss reserves?

It is great that OSFI wants lenders to allocate more capital against their loans, but none of these changes are happening now or even retroactively.  If they know that poor lending has occurred, don’t they want more capital against those loans now?  Why not force lenders known to be doing a poor job to raise capital?

Rather than conserve capital, we are seeing the opposite from the publicly traded alternative lenders and mortgage insurers.  Here are a few examples:

Equitable Group (EQB): increased dividend by 4.5%

Genworth Canada (MIC): is buying back stock and it increased its dividend.

MCAN Mortgage Corp (MKP): MCAN increased its dividend by 7%

Laurentian Bank (LB): increased its dividend (although the increase was small)

I should note that all of these banks could be doing a fine job.  They may not be the ones OSFI is referring to in their letter.  However the truth is we don’t know, and I think we should.


Disclosure: I am short EQB, MIC and MKP



  • Muchas gracias. ?Como puedo iniciar sesion?

  • Good post. OSFI’s proposed changes have received a lot of attention. Most of it focused on how severely this will impact lending (and ultimately house prices). Based on this, do you think OSFI waters down these changes? Or waits to see impact of Toronto’s slowing market before implementing.

    Also, how significant is today’s 25bps rate increase? I’m thinking it’s marginal, primarily affecting first time buyers (those renewing their 5 year fixed are still getting a lower rate) and those with variable rate mortgages. HELOCs as well but the dollar figures on those are much lower. I think this hurts Alberta more than people realize though.

    But another 25bps in October (with a corresponding increase to the 5yr) would start to really make an impact, especially if Poloz guides to another increase in early 2018. If you have a different take would be interested in your thoughts.

    I enjoy your thoughtful posts. Thanks.


Leave a comment