Refinements to the Dynamic Risk Management (DRM) model—Risk Limits (Agenda Paper 4A)

During the September 2021 meeting the Board discussed potential refinements to the DRM model in order to address the challenges identified during outreach and to align the DRM model more closely to entities’ risk management practices by incorporating the concept of risk limits into the target profile. In this paper, the staff is proposed further refinements to address the comments and questions received from the Board during the meeting in September and is also explaining the benefits of these refinements.  

Proposed refinements to the DRM model

• Revising the definition of the target profile—the target profile will be defined as “the range (risk limits) within which the current net open risk position can vary while still being consistent with the entity’s risk management strategy”. The target profile will have to be linked to the entity’s risk management strategy and documented at initial designation of the hedge. Any changes to the target profile would result in the discontinuation of the hedge. The staff believe that the revised definition is more closely aligned with entities’ risk management strategies and also better aligned with the general hedge accounting model in IFRS 9 which will help with understandability and operability of the DRM model.
• Introducing the risk mitigation intention—the risk mitigation intention will be described as “the extent to which an entity intends to mitigate the current net open risk position through the use of derivatives”. This will be a single-outcome element determined on the entity’s preferred risk metrics. The risk mitigation intention cannot create new risks, it has to transform the current net open risk position to a residual risk position that is within the target profile. It is a fixed amount of risk to be mitigated through derivatives and is set for a specific period of time. The maximum amount of risk mitigation intention is capped to the current open risk position and is not affected by the entity’s target profile determined at the inception of the hedge. On the other hand, the target profile determines the minimum amount that an entity needs to designate as the risk mitigation intention must be consistent with the entity’s risk management strategy. Once the risk mitigation intention is determined, it is documented via construction of benchmark derivatives. Changes to the risk mitigation intention can occur without affecting the continuation of the DRM model but any changes can be made only prospectively (i.e. entities are not allowed to amend the risk mitigation intention retrospectively). Any unexpected changes for the period under review will be captured through retrospective assessment and may affect the measurement of misalignment in the financial statements. The risk mitigation intention will have a similar role to that of the risk management objective in the general hedge accounting model of IFRS 9.
• Revising the construction of benchmark derivatives—benchmark derivatives must be based on the risk mitigation intention because the risk mitigation intention is a ‘fixed amount of risk’ while the target profile represents acceptable open risk positions that do not specify the extent to which an entity decides to mitigate the risk. These benchmark derivatives will be used to measure the performance (similarly to hypothetical derivatives in the general hedge accounting model of IFRS 9).
• Introducing prospective and two retrospective assessments—these assessments would ensure that the DRM model is used to mitigate the interest rate risk, that the target profile is achieved and that any misalignment that is driven by the effect of unexpected changes is captured in the financial statements. However, as the recognition of the misalignment will be determined based on the ‘lower of’ test, it may not always lead to the recognition of a gain or loss in the P&L (i.e. the entity only recognises gains and losses in the P&L on the over-hedged position similarly to the general hedge accounting model in IFRS 9).

Benefits of the refinements to the DRM model

The staff indicated that the introduction of the risk mitigation intention provides a clear link between the application of the DRM model and an entity’s risk management strategy. With this approach, entities can explain better the DRM results in their financial statements and the users of financial statements can have a better visibility of the alignment between the actual results achieved and the target profile (or the entity’s risk mitigation intention). Furthermore, this will reduce the need for ‘proxy hedging’. The introduction of the risk mitigation intention will result in an operational simplification, and reduced cost and/or errors because the retrospective assessment will be based on a ‘period-to-period’ comparison and hence entities will not be required to track the changes in the expected cash flows across several periods.  

Question for the Board

The staff asked whether the Board agrees with the proposed refinements to the DRM model.

Board discussion

One Board member raised the question whether it will be possible to apply other hedge accounting models if an entity applies the DRM model. The question was raised because of the wording used in the agenda paper which suggests that when an entity applies the DRM model, it may no longer designate derivatives against the gross positions. It was indicated that the group may have a segregation of the portfolio, similarly to the business model for a particular group of assets, and may wish to apply the DRM model only to a part of the portfolio and to designate the gross position to the other part. The staff answered that the same item cannot be designated twice for same risk, i.e. both in the DRM model and then in micro-hedge accounting. But the hedged item may be hedged for more than one risk, hence the DRM model may be used for the interest rate risk and micro-hedging for the FX risk. The staff confirmed that the item cannot be designated on the gross basis if it is already in the DRM model.  

The question was raised as to whether the concept of ‘unexpected changes’ is clear. The staff confirmed that it is clear because the current net open risk position which is set up by an entity is based on the expected changes and repricing. Any changes that are outside of that are ‘unexpected changes’. This can be caused, for example, by more prepayments than originally expected, or a bigger growth in the book. Then a point was raised in respect of retrospective effectiveness tests and that the documentation has to be very clear as to which is the inception day for each time bucket given the test is based on a cumulative basis. The staff reiterated that the Board will need to distinguish between effectiveness assessment which is performed via a bucket-by-bucket and period-on-period comparison and measurement that is based on a cumulative basis as is currently done. However, the entities still need to know what was in the bucket at the beginning of the period and at the end of the period to compare those two and determine the retrospective assessment. It was confirmed that each bucket will be assessed.

Several Board members had different wording recommendations. Generally, they were in respect of clarity or definitions and consistency of the wording used. For example:

• The term ‘current net open risk position’ is used in the definition of the risk mitigation intention and therefore it is important to determine the date of the current net open risk position, i.e. whether that is the beginning of the reporting period or the designation date. Then, when the misalignment is measured, the term is used again but the ‘current net open risk position’ would have already changed, so the wording has to be clear what exactly is compared
• The term ‘trading derivatives’ is used when explaining risk changes in risk mitigation intention. These derivatives were not executed for trading purposes only for hedging purposes hence the wording is misleading
• In the same paper two different words are used, namely ‘adjust’ and ‘amend’ when describing the same thing. The staff was asked to always use the same word because even if these words may be interchangeable in English, they may be later translated as two words with different meanings into other languages
• The term ‘period’ could be a ‘reporting period’ or a ‘risk management period’ and determines how much disclosures will be later required
• It was also challenged that as currently described, the ‘lower of’ test is trying to limit the amount recognised in profit or loss while in the general hedge accounting model the ‘lower of’ test is limiting the amount recognised in OCI, i.e. it is currently trying to limit the wrong feature

One Board member commented that the consistent ideas between the DRM model and the general hedge accounting model in IFRS 9 are good as they help in understanding the model. The staff was encouraged to use the same language for the same ideas. Another comment was raised in respect of ‘reducing interest rate risk’ and that often by reducing one type of the risk, another risk is created. For example, if the cash flow variability is taken away, fair value variability is introduced so it is not a totally neutral risk position. Therefore, very specific language should be used, and the Board should be very precise which risk is reduced. For example, say ‘reducing re-pricing risk due to interest rate changes’ and use the term everywhere. Similarly, net open risk position may be different from the position that is hedged at the group level, hence it should be very specific to the population hedged.

The staff was challenged whether the misalignment will be visible. The staff answered that there are different types of misalignment and because the ‘lower of’ test will be used, not all will be recorded in profit or loss. It was suggested that the staff should further explore the disclosures in those cases when the ineffectiveness is not recognised in profit or loss due to mechanics of the ‘lower of’ test. It was said that such information about the misalignment is still very important and useful to the users of financial statements. The staff agreed with that.

Another Board member reminded the Board that it had to be very clear that the DRM model is for external hedging, using external derivatives. Then a question was raised in respect of retrospective testing, i.e. whether the changes that happen during the period will be captured as the testing suggests comparison between the beginning of the period and the end of the period. For example, during the period there can be unexpected prepayments and increases on the portfolio. It follows that overall, at the end, the position may not change. It was asked how such a misalignment that happened during the period will then be captured. Furthermore, it was asked to clarify that the period does not mean ‘reporting period’. This was important with regard to the frequency of testing and adjusting the risk management intention. The staff was also asked to be clear that any new loans could be only added prospectively.

Then a question was raised as to why the staff suggest to have two retrospective tests (i.e. testing the target profile and risk management intention) if the consequences are the same. The staff commented that the risk management intention is to be within bands and due to unexpected changes, the entity may fall outside the bands. That is what one of the tests is trying to capture.

Another Board member commented on the benefits of the DRM model that were stated in the paper. In his view, the DRM model would allow to better explain the movement in net interest income but the disclosures will be critical to get the insights into the risk management activities of the entity.

It was also indicated that the focus should not only be on the backwards tests but also on the prospective basis and how that was communicated. It was indicated that one of the limiting factors may be that entities are sometimes overhedging in particular time buckets, but not overall, and this is likely to provoke discussion.

One Board member asked whether one of the benefits mentioned, namely the reduction in proxy hedging, is due to entities not being able to use proxy hedging or due to the DRM model being more attractive. The staff confirmed that it is the latter.

One of the Board members was unsure whether the statement that “anytime that the entity sees the increase in the risk in the bucket, has a consequence” was correct. There may be situations in which the entity cannot do anything, but most often the entities act in order to bring the exposure below the line. So, the question is when this reassessment should take the place i.e. (i) when the entity decided to act and change the derivatives position, bring the risk back below the line, or (ii) at the end of period.

Then it was explained that one area where it cannot be done is the last bucket. It was explained that the ‘bucket’ is a very neat assumption, but the buckets move, so the question is what is kept static, i.e.

• Buckets - the cash flows are moved from one bucket to the other bucket)
• All the buckets move on the time horizon and then a new bucket is added at the end, but this last bucket cannot be controlled

These are the issues that has be looked at in the future. The staff agreed that much work needs to be done with regard to the model.

Board decision

All Board members tentatively agreed with the proposed refinements to the DRM model.

Designation of a portion of prepayable assets in the DRM model (Agenda Paper 4B)

In this paper, the staff addressed the second of the challenges identified during the outreach, namely the designation of a portion of prepayable assets in the DRM model. The participants of the outreach recommended that the Board allow the designation of the bottom layer of a portfolio of prepayable assets in the DRM model instead of the percentage approach.

The staff indicated that the proposed refinements to the DRM model (as described in Agenda Paper 4A) would also resolve the need to designate a bottom layer for risk management purposes. This is because the risk mitigation intention would allow an entity to decide the extent of the current net open risk position to mitigate by using derivatives. Specifically, when deciding how much of the open risk position to hedge, the entity would have to take into consideration that the prepayment levels need to be consistent with its risk management strategy. As such, an entity applying the DRM model would not need to use a bottom layer approach.

In the light of the above, the staff is not recommending any further refinements with regard to this issue.

Question for the Board

The staff asked whether the Board agrees with the staff view that no further refinements to the DRM model are needed in respect of the designation of a portion of prepayable assets.

Board discussion (Agenda Paper 4B)

Board members questioned whether by using the bottom layer approach, entities could achieve a particular accounting outcome. The staff reiterated that the DRM model is driven by the risk management and not by the accounting outcome. The entities look at the books holistically and not individually. Furthermore, the staff explained that if an entity has to capture the economic effect of what is happening through the bottom layer, the DRM model does exactly that, i.e. it captures the expected maturities of the assets.

One of the Board members agreed with the principle but as a caution, he wanted to check with French banks.

One Board member asked whether the FASB paper on multiple layers had any impact on the DRM model. The staff said that they have something similar that the Board had in the portfolio hedging requirements. It is a slightly different model.

Decision

All Board members tentatively agreed that no further refinements to the DRM model are needed in respect of the designation of a portion of prepayable assets.