Residential Mortgage Underwriting Practices and Procedures Guideline (B-20)

Sound Mortgage Underwriting Contributes to Financial Stability

The safety and stability of federally regulated financial institutions is fundamental to the ongoing health of Canada’s financial system and its economy. OSFI contributes to this stability by fulfilling its mandate to protect the interests of depositors and other creditors of financial institutions. It does this by setting standards that improve banks’ resilience, both under normal conditions and in the event of a financial downturn.

Lenders subject to OSFI supervision hold nearly 80 per cent of all residential mortgages issued in Canada, and residential mortgage loans account for almost 30 per cent of the total assets held by these lenders. Sound mortgage underwriting practices reduce risks to the financial system, and to Canadians who entrust their savings to Canada’s financial institutions.

Sound mortgage underwriting practices require lenders to assess a borrower’s ability to pay their loan under a variety of conditions. A lender should consider potential changes to a borrower’s income and expenses, as well as changes to the market environment including the valuation of the property that is being mortgaged.

This information sheet is updated periodically to provide information about the effectiveness of Guideline B-20 and statistics about some of the areas that we continue to monitor.

History of Guideline B-20

In 2012, OSFI introduced its Residential Mortgage Underwriting Practices and Procedures Guideline (B-20) to set out expectations for strong residential mortgage underwriting for federally regulated lenders. The original Guideline was mostly principles-based and included an expectation that lenders would stress test borrowers for adverse conditions.

In 2016, OSFI reminded lenders of its underwriting expectations in the form of a public letter. When OSFI continued to see examples of relaxed mortgage underwriting at some lenders, it issued an update to Guideline B-20 in 2017 that came into effect in January 2018.

The updated Guideline B-20 clarified and strengthened expectations to address what OSFI saw as increasing risks in an environment of historically low interest rates, high levels of consumer debt and housing imbalances.

The revised Guideline B-20 includes:

  • measures requiring financial institutions to apply greater rigor when assessing a borrower’s ability to repay their mortgage loans, including when verifying a borrower’s employment status and income history
  • a revised minimum qualifying rate (stress test) that requires lenders to confirm borrowers would be able to continue repaying their mortgage loans if faced with a sudden change to their circumstances (income loss, increased interest rates, additional expenses, etc.). The minimum qualifying rate is the greater of the contractual mortgage rate plus two percentage points, or the five-year benchmark rate published by the Bank of Canada
  • a requirement for lenders to place more scrutiny on property valuations, establish dynamic loan-to-value limits that reflect the risk of specific properties and markets, and update these limits and practices as housing markets and the economic environment evolve.
  • The extraordinary circumstances at the beginning of the COVID-19 pandemic required OSFI to make a series of decisions intended to focus on continued resilience, the delivery of government programs and operational pressures on institutions we oversee. This included pausing policy consultations, and special temporary capital treatment of deferred payments.
  • OSFI continues to assess policy measures against the principles and considerations laid out in a speech delivered in September 2020.

Guideline B-20 Impact on Uninsured Mortgages

The following data, similar to the note issued in January 2020, represents uninsured mortgages from 20 federally regulated mortgage lenders. These lenders hold more than 99 per cent (by $) of all mortgages subject to OSFI regulation and supervision. The charts below include data up to and including November 2020.

Mortgage underwriting quality warrants continued attention:

When Guideline B-20 revisions were introduced, lenders made changes that reduced the proportion of mortgages approved for the most highly indebted or over-leveraged borrowers (e.g. mortgage loans that exceed 450% of a borrower’s income.)

  • This metric peaked at 20% prior to the implementation of Guideline B-20 changes in 2018 and remained low over the following year and a half.
  • In the second half of 2019, there was renewed growth in the exposure to these highly indebted borrowers that has accelerated further in 2020.
  • Of note, the debt servicing capacity of highly indebted borrowers has actually improved, largely driven by historically low mortgage rates. (See Chart 1)
Text version - Chart 1
Average TDSR / Proportion with LTI > 450% Trends for Unsinsured Mortgage Originations in Canada
PeriodAverage TDS% LTI > 4.5x
2014 Q144.13%12.29%
2014 Q243.82%11.93%
2014 Q343.32%13.22%
2014 Q443.23%13.57%
2015 Q142.87%13.95%
2015 Q241.86%15.58%
2015 Q342.41%16.91%
2015 Q442.78%16.65%
2016 Q143.39%16.84%
2016 Q242.77%16.92%
2016 Q342.51%18.30%
2016 Q442.53%18.61%
2017 Q142.57%19.33%
2017 Q242.32%19.51%
2017 Q341.56%20.03%
2017 Q441.55%19.18%
2018 Q143.14%17.36%
2018 Q243.98%13.65%
2018 Q343.65%14.45%
2018 Q444.56%14.06%
2019 Q145.99%14.25%
2019 Q245.25%14.57%
2019 Q344.41%16.43%
2019 Q445.03%17.49%
2020 Q145.35%18.26%
2020 Q244.29%17.99%
2020 Q344.09%20.00%
2020 Q443.57%22.68%
  • Another metric that OSFI monitors to assess underwriting quality is the proportion of new uninsured mortgage originations with loan to value ratios in excess of 75%. These are known as high ratio mortgages, meaning there is less equity at origination than a conventional mortgage. While this measure peaked prior to the implementation of Guideline B-20, it has been increasing, especially over the past the year, although it remains below its historic peak levels. (See Chart 2)

OSFI will continue to monitor the performance of these metrics, along with a number of other measures to assess the effectiveness of Guideline B-20.

Text version - Chart 2
Uninsured mortgage originations, proportion of mortgages with LTV > 75% by $
Period% > 75%
2014 Q151.1%
2014 Q253.3%
2014 Q351.9%
2014 Q450.6%
2015 Q148.8%
2015 Q249.9%
2015 Q349.9%
2015 Q448.5%
2016 Q145.6%
2016 Q245.9%
2016 Q345.0%
2016 Q445.8%
2017 Q145.6%
2017 Q245.4%
2017 Q345.6%
2017 Q446.7%
2018 Q145.9%
2018 Q246.6%
2018 Q346.3%
2018 Q447.1%
2019 Q146.7%
2019 Q248.8%
2019 Q348.2%
2019 Q447.9%
2020 Q147.6%
2020 Q248.5%
2020 Q350.3%
2020 Q449.8%

Amortization periods not extending:

Insured mortgages have a maximum amortization period of 25 years, while uninsured mortgages subject to Guideline B-20 can be paid over a longer period.

  • Uninsured mortgage borrowers do not appear to be extending amortization periods to pass the stress test requirement. (See Chart 3)

OSFI will continue to monitor the Canadian real estate market, the impacts of COVID on market dynamics, as well as lender practices, particularly in the area of income verification and collateral values. OSFI will be proactive with lenders when it identifies areas requiring attention.

Text version - Chart 3
Mortgages with amortizations greater than 25 years
Date% >25 yrs
2015 Q157%
2015 Q257%
2015 Q360%
2015 Q461%
2016 Q163%
2016 Q264%
2016 Q366%
2016 Q466%
2017 Q157%
2017 Q251%
2017 Q352%
2017 Q452%
2018 Q152%
2018 Q246%
2018 Q348%
2018 Q451%
2019 Q151%
2019 Q248%
2019 Q350%
2019 Q452%
2020 Q150%
2020 Q249%
2020 Q350%
2020 Q452%

Rates at renewal are stable:

OSFI’s monitoring of the mortgage market includes rates at renewals. Contrary to some speculation when the revised Guideline B-20 was introduced in 2018, there has been no significant spread between the rates offered at origination and those at renewal.

  • Data from OSFI regulated lenders shows that following the introduction of the revised guideline, the difference between renewal and new mortgage rates for uninsured five-year fixed and variable rate mortgages has remained largely unchanged. (See Chart 4)
Text version - Chart 4
Renewal / new mortgage rate difference for five-year mortgages
DateRen-Origin
2015 Q10.02%
2015 Q20.11%
2015 Q30.20%
2015 Q40.06%
2016 Q10.14%
2016 Q20.17%
2016 Q30.17%
2016 Q40.10%
2017 Q10.04%
2017 Q20.00%
2017 Q30.07%
2017 Q40.10%
2018 Q10.07%
2018 Q2-0.06%
2018 Q30.11%
2018 Q40.16%
2019 Q10.05%
2019 Q20.15%
2019 Q30.10%
2019 Q40.10%
2020 Q10.09%
2020 Q20.11%
2020 Q30.05%
2020 Q40.06%

Associated Links

Guideline B-20: Residential Mortgage Underwriting Practices and Procedures

Speech: Sound Mortgage Underwriting: Foundation for Stability - 2020

Speech: Stressing the Stress Test: The Importance of Strong Mortgage Underwriting - 2019

Speech: Guideline B-20: Preparing for the Unexpected - 2019