Changes to the Capital Adequacy Requirements (CAR) Guideline
Information
Table of contents
OSFI is releasing the final version of the CAR Guideline for implementation in Q1 2019Footnote 1. The main revisions relate to the domestic implementation of the standardized approach to counterparty credit risk (SA-CCR), capital requirements for bank exposures to central counterparties (CCPs) and the securitization framework.
We have also clarified the capital treatment for right-of-use assets resulting from the adoption of IFRS 16 beginning January 1, 2019. This treatment will become effective for institutions upon their adoption of IFRS 16.
In addition to the above changes, the guideline includes the changes to the capital floor that were communicated to industry in January 2018. OSFI has also removed the Credit Valuation Adjustment (CVA) phase-in and other transitional arrangements that conclude at the end of 2018. Further, the guideline now lists Kroll Bond Rating Agency Inc. as an eligible external credit assessment institution (ECAI) for capital purposes, which became effective on July 6, 2018. Finally, we have provided clarifications throughout the Guideline in response to questions received from the industry as part of our regular annual updates.
The attached table in Annex 1 summarizes comments received and provides an explanation of how the comments have been addressed in the Guideline. We thank those who participated in the consultation process.
Questions concerning these changes can be sent to Catherine Girouard, Director, Capital Division by email at catherine.girouard@osfi-bsif.gc.ca.
Yours truly,
Carolyn Rogers
Assistant Superintendent
Regulation Sector
Annex 1: Summary of Comments Received and OSFI Response
| Comment | OSFI Response |
|---|---|
| Chapter 3 – Treatment of Public Sector Entities (PSEs) | |
| As per paragraph 10(ii), institutions should look to the host government to confirm whether an entity is a PSE in competition with the private sector. However, in practice, this is difficult, if not impossible to do. As such, banks have typically used the list published under OSFI's implementation of Basel I, which has since been removed from the CAR Guideline. Can OSFI modify the language to provide banks with flexibility in determining whether a PSE is in competition with the private sector? | OSFI recognizes the difficulty in consulting a host government to determine if a PSE is in competition with the private sector given that this information is typically not available in the public realm. As such, we have modified the language in paragraph 10 to allow banks to determine if a PSE is in competition with the public sector based on their own criteria, which must be documented in an internal policy. |
| Chapter 3 – Equity Investment in Funds | |
| The draft Guideline proposes a conservative approach to determining exposure at default (EAD) under the Standardized Approach to Counterparty Credit Risk (SA-CCR) under the mandate based approach. We believe the calculation is overly punitive and request that the multiplier of 1.4 be removed. | OSFI does not believe that the multiplier needs to be removed, as the mandate based approach requires a level of conservatism to be included. In addition, the treatment under the mandate based approach should be consistent with other parts of the CAR Guideline. Finally, the look-through approach, which does not contain the embedded conservatism of the mandate based approach, is available to banks making equity investment in funds. |
| Chapter 3 – IFRS 16 | |
| The adoption of IFRS 16 does not change banks' current obligations or risk, and therefore should not affect regulatory capital requirements. OSFI should clarify that capital should be held only for the net lease asset amount. |
IFRS 16 provides greater transparency to institutions' statement of financial position. Prior to IFRS 16, lease exposures were underrepresented (on the balance sheet) and, in OSFI's opinion, so were the capital requirements. OSFI supports the IFRS 16 approach whereby, from a lessee's perspective, no distinction is made between assets that are leased ("operating leases") and assets that are financed to be owned ("finance leases"). The CAR Guideline reflects this approach by applying the same capital treatment to all leases subject to IFRS 16. Capital requirements consider the carrying amount of property (i.e., the risk exposure), whether leased or owned. They also capture other types of risk associated with leases (e.g., early cancellation penalties, fluctuating market lease rates, costs associated with renegotiating or finding a new lease, as well as uncertainty in the realizable value of right of use assets in times of stress). |
| Chapter 3 – Treatment of short-term bank exposures | |
|
OSFI should consider allowing the use of a 20% risk weight for the calculation of the capital floor for short term exposures to banks, which is consistent with the Basel I approach. The change from a Basel I based floor to a Standardized Approach based floor, effective in Q2-2018, has significantly impacted trade finance exposures. Under Basel I, the risk weight was generally set at 20% given a carve-out for claims with a residual maturity of less than 365 days. Under the applicable standardized approach, most trade finance exposures receive a 100% risk weight. The new risk weight is out of line with the risk of trade finance, which is widely recognized as a low risk form of financing. |
OSFI's approach to the calculation of the capital floor is to use the existing applicable standardized approach for credit risk under the CAR Guideline. Short term bank exposures should not be treated differently solely for the capital floor, and, as such, if changes are made to the treatment of bank exposures, these would apply to both the calculation of the capital floor and to bank exposures under the standardized approach. Further analysis and broader consultation would be needed to consider changing the treatment of bank exposures. |
| Chapter 3 – Restriction on the use of unsolicited ratings | |
| OSFI should reconsider the current restriction on the use of unsolicited ratings in determining the risk weights in the CAR Guideline. This restriction contributes to misperceptions on the nature of unsolicited credit ratings. | OSFI conducted a review of the treatment of unsolicited ratings in 2013, particularly for sovereigns. This review determined that unsolicited ratings could be used for sovereign exposures when a solicited rating is not available. OSFI believes that this position is appropriate. The review also concluded that we should not extend the same treatment to corporates without additional study. |
| Chapter 4 - Implementation of the SA-CCR | |
| We are concerned with OSFI's decision to move ahead with the implementation of the SA-CCR when other jurisdictions have yet to do so. This could place Canadian banks at a significant disadvantage compared to global peers. |
The SA-CCR is a significantly better measure of counterparty credit risk exposure than the Current Exposure Method. The new method provides better incentives to enter into netting agreement and exchange margin with their counterparties. In addition, many global peers make use of the internal modeling method (IMM), which allows banks to model their derivative exposure. Canadian banks with large derivative portfolios are eligible to apply to OSFI to use the IMM. This approach has been available to banks for more than five years and places Canadian banks on a level playing field with their global peers. |
| Chapter 7 – Grandfathering and transitional arrangementsFootnote 2 | |
| The lack of true grandfathering provisions will unnecessarily put Canadian banks at a distinct disadvantage relative to competitors in jurisdictions that are either not adopting the Revised Securitization Framework or are incorporating grandfathering provisions for existing structures. Risk weights will increase upon adoption of the revised framework and banks will not have the ability to address these changes until the next renewal date, or not at all in the case of amortizing transactions, without giving the industry time to re-price transactions. | OSFI recognizes international competition at the large banks as well as funding costs at smaller banks may be disrupted if grandfathering is not provided. In the guideline, OSFI has provided grandfathering of the current capital treatment for one year through a negative adjustment to risk-weighted assets that effectively eliminates the initial increase in risk weights. |
| Chapter 7 – Quantitative significant risk transfer (SRT) test | |
| We remain concerned that the SRT quantitative requirement is overly conservative and will, more often than not, disqualify transactions that have genuine risk transference. The proposed calibration is notably more constraining than similar tests used by other international regulators and will make Canadian banks less competitive than their international peers, thereby discouraging transactions that transfer risk out of the Canadian financial system. | Recognizing the concerns, OSFI will provide an exemption from the SRT test if all of an institution's exposures are risk-weighted 1,250%. OSFI will also modify the calculation of the test such that banks can perform the test without applying the risk weight floor. In addition, OSFI will raise the SRT test threshold from 30% to 40%. |
| Chapter 7 – Use of the Internal Assessment Approach (IAA) for the overall capital floor | |
| Not allowing banks to use the IAA when calculating RWA for the output floor raises costs by requiring banks to pay for an external rating on their liquidity facilities without a commensurate benefit. | While OSFI appreciates that there is a cost to obtain external ratings, there is model risk in applying the IAA, which the floor is in place to mitigate. Therefore, OSFI has not modified the guideline on this point. |
| Chapter 7 – Definition of Tranche Maturity | |
| The definition of tranche maturity is overly conservative because it materially overstates the time period during which banks are exposed to unexpected losses and should be modified to a tranche's expected weighted-average life. | Allowing an expected weighted average life calculation would entail a reliance on banks' internal models and assumptions on defaults, pre-payments and other cash flows. This reduces comparability and leaves no assurance of conservatism. Therefore, OSFI has not changed the definition of tranche maturity. |
| Chapter 7 – Top down approach concentration limit | |
| The 4% concentration limit in applying the top-down approach is unnecessary because transactions normally have other mitigating features. | OSFI has imposed a concentration limit of 4% to prevent a lumpy pool from distorting the averages that banks calculate when using this approach. While OSFI continues to believe the limit is warranted, the guideline clarifies that the approach may be applied to sub-pools. This allows pools to be split between large exposures measured using a bottom-up approach and other exposures measured on a top-down basis. |
| Chapter 7 – Treatment of Excess Spread | |
| Excess spread acts as a first level of credit protection to absorb losses, prior to losses being absorbed by the more traditional credit protection tranches. | While profits on other loans are the first buffer to absorb losses, these profits can disappear quickly in a crisis. For this reason, these profits cannot be recognized in capital until they are earned. OSFI has not made any changes to the treatment of excess spread. |
| Chapter 7 – Simple, transparent and comparable (STC) securitizations – Applicability of the short-term STC criteria | |
| Sponsors of asset backed commercial paper (ABCP) have the ability to move freely most transactions between their own balance sheets and ABCP conduits. Identical transactions may be funded either on-balance sheet or through an ABCP conduit. Bank balance-sheet funded exposures should be subject to the transaction level short-term STC criteria (Appendix 7-2). | OSFI agrees with the view that the funding source for the transaction does not affect the credit risk of that transaction, while also maintaining a clear distinction between which set of STC criteria applies to a transaction. Reflecting this view, OSFI has added language to the chapter that allows banks to apply the short-term STC criteria to balance-sheet funded exposures under certain conditions. |
| Chapter 7 – Simple, transparent and comparable (STC) securitizations – Increased operational burden | |
| Several criteria require collection and/or disclosure of information that is not currently available or is not collected according to the specified metrics. This could make certain criteria difficult to meet and, in many cases, would entail a significant change and an increase in work compared to current practice. |
OSFI recognizes the increased level of risk measurement, monitoring and disclosure being required by the STC criteria and the potential difficulty and/or operational burden for certain securitization transactions to qualify for preferential capital treatment. As such, OSFI has modified language where possible to reduce the potential operational burden of meeting certain STC requirements while retaining the substance and intent of the particular STC criterion. For example, the guideline now specifies that certain checks may be performed on a representative sample of underlying obligors, that disclosures may be limited to only those items applicable to the securitization transaction, that certain disclosures need only be provided if requested by an investor and in a timely manner as legally permissible, or that, in some cases, tests similar to those specified may be performed provided they adequately measure the relevant risks. |
| Chapter 7 – STC securitizations – Practical application of STC criteria | |
| We recommend that the failure to meet any one individual criterion should not disqualify the transaction from STC treatment. In addition, verifications and disclosures of risks required in various STC criteria should be subject to a materiality threshold. Further, the criteria should take a practical rather than prescriptive approach and should reflect how securitization transactions are conducted in Canada. |
The STC criteria are designed to collectively safeguard the ability of investors in Canadian securitization transactions to properly and conveniently assess the characteristics, ongoing risks and performance of STC securitization transactions. To this end, institutions must meet all criteria in order for an exposure to be deemed STC compliant and receive preferential capital treatment. That said, several requirements set out within individual criteria have been made subject to materiality and practical judgement when warranted, while preserving the spirit and intent of the criteria. Further, we have made targeted adjustments to better align STC criteria with characteristics of the Canadian securitization market and the Canadian legal and economic environment. |
| Chapter 9 – Practical application of the hierarchy of approaches and STC criteria to trading book exposures | |
| The short term nature of trading book exposures introduces operational challenges to applying the model-based approach (SEC-IRBA) as well as challenges to determining whether exposures are STC compliant due to the manual intervention required. | OSFI has added language to the chapter which allows institutions to follow internal policies to mitigate the highlighted operational challenges. |
Footnotes
- Footnote 1
-
November 1, 2018 for institutions with an October 31st year end and January 1, 2019 for institutions with a December 31st year end.
- Footnote 2
-
Due to the large number of comments received regarding the significant revisions to the CAR securitization framework, comments and OSFI responses in this area have been summarized according to general themes.