What is OSFI Guideline B-12?

What it covers, and for whom

B-12 applies to banks (federal credit unions included), bank holding companies, and federally regulated trust and loan companies, on a consolidated basis and scaled to each institution’s size and complexity.

The risk is easy to state and hard to measure. A bank funds 5-year fixed mortgages with deposits that can reprice or leave at any time. When rates move, the income from the two sides moves differently, and the present value of the whole book shifts. B-12 asks institutions to measure that two ways: earnings (what happens to net interest income over the next year or so) and economic value (what happens to the value of everything on the books over its full life). It also asks them to watch credit spread risk in the banking book, a close relative.

What the 2027 version changed

OSFI’s backgrounder of 10 September 2026 names three changes from the 2019 guideline:

• New shock scenarios, aligned with the Basel Committee’s 2024 targeted amendments to its IRRBB standard.
• Clearer earnings expectations: a constant balance sheet, where maturing items roll over into the same positions, is the minimum for assessing earnings sensitivity. A dynamic view with planned new business is optional, and OSFI warns it can hide risk.
• Disclosure moved out: the detailed disclosure requirements left the guideline to avoid duplication and continue under OSFI’s Pillar 3 disclosure expectations.

It takes effect on 1 November 2026 for institutions with an October 31 year-end and on 1 January 2027 for those with a December 31 year-end.

The principles, in short

B-12 follows the Basel Committee’s principles, numbered the same way. In plain terms:

1. IRRBB is important for every institution and has to be identified, measured, monitored and controlled.
2. Senior management oversees the framework and the risk appetite, with regular independent review.
3. The appetite is stated in both economic value and earnings terms, with limits to match.
4. Measurement uses both kinds of measure, across a wide range of shock and stress scenarios.
5. Behavioural and modelling assumptions are understood, documented and tested.
6. Data and models are accurate, documented and independently validated.
7. Results and hedging go to senior management regularly, by level of consolidation and currency.
8. Exposure and practices are disclosed publicly, now through Pillar 3.
9. Capital for IRRBB is assessed in the ICAAP.
10. OSFI assesses each institution’s IRRBB and its management of it.
11. Institutions flagged as outliers can be required to cut exposure, add capital or improve risk management.

The shock scenarios

For economic value, B-12 prescribes six standardized scenarios: parallel up, parallel down, steepener (short rates down, long rates up), flattener (short rates up, long rates down), short rates up and short rates down. Earnings use two prescribed scenarios. Institutions also design their own scenarios and run historical, hypothetical and reverse stress tests, which are usually harsher.

The sizes come from rate history from January 2000 to December 2023 at the 99.9th percentile, rounded to the nearest 25 basis points. Post-shock rates have a lower bound of −75 basis points.

The outlier test

OSFI compares the largest loss of economic value under the six scenarios with 15% of Tier 1 capital. An institution over that line is an outlier, and OSFI may also run tests based on CET1 or earnings. An outlier can be required to reduce its exposure, raise capital, constrain its internal assumptions or fix its risk management within a set time.

Assumptions are where the judgement is

The numbers depend heavily on behaviour nobody controls: how long chequing and savings balances really stay (non-maturity deposits), how fast borrowers prepay fixed-rate mortgages, how many term depositors redeem early. B-12 expects these assumptions to be documented, varied by depositor and account type, tested against what actually happened, and reported to senior management with the results.

That makes an IRRBB system a model in OSFI’s sense. B-12 says institutions should fully review their IRRBB measurement models consistent with E-23, with validation independent of development, and internal audit reviewing the process. So the deposit model and the prepayment model belong in the same model inventory as the credit models, with their own risk rating, validation and monitoring. MINERVA keeps them there. See OSFI E-23 software.

Related questions

Sources

• OSFI — Guideline B-12, Interest Rate Risk Management (2027)
• OSFI — Backgrounder: Final Guideline B-12, Interest Rate Risk Management (10 September 2026)
• Basel Committee on Banking Supervision — Interest rate risk in the banking book

New answers are posted on MINERVA on LinkedIn. Follow along, or send the question you’d like answered next to minerva@minerva.studio.

A plain-language explanation, not legal advice. The regulator’s own text is the source; read it, and take advice on how it applies to your institution.

Read next

• What is OSFI’s CAR guideline?
• What are OSFI guidelines, and are they binding?
• What is model validation in banking?
• OSFI Guideline E-23, explained