The DRM model as the future IFRS portfolio hedge accounting model is intended to resolve the long-standing tension between interest rate risk management at portfolio level and its presentation in the IFRS financial statements and overcome the weaknesses of the still relevant IAS 39 Portfolio Fair Value Hedge. Some risk managers in particular are demanding for the risk management view to be fully incorporated into the IFRS financial statements. The question therefore arises as to the extent to which a complete risk management view is possible or sensible in IFRS financial statements and how the risk management view can be adapted as far as possible in the DRM model.
Thought experiment complete adoption of the risk management view
The risk management view is characterized by the full consideration of all risk management activities, the symmetrical assessment of current risk on the basis of expected future developments, e.g. market interest rates, customer behavior and new business volume, as well as the similar consideration of all risk-induced valuation results on a present value (EVE) or periodic basis (NII). Their complete adaptation would have the following consequences in IFRS financial statements, among others:
> Measurement of all hedged items at hedge fair value through profit or loss
> Cumulative consideration of expected interest rate developments and expected customer behavior as well as future transactions relevant to earnings
> Recognition of the refinancing funding effect of equity in profit or loss
> Consideration of internal transactions for risk allocation
> Summary of all interest-induced earnings components in one line item
This would violate some fundamental IFRS principles of the IFRS framework and the relevant IFRS standards for financial instruments, including the accrual principle and the reporting date principle, the measurement of equity as a residual value of assets and liabilities, the neutrality of dividend payments in profit or loss, the prohibition of offsetting, the use of the different subsequent measurement categories AC, FVTPL and FVOCI for debt instruments on the basis of cash flows and business model (“mixed model”), the prohibition of accounting for exclusion of internal transactions and the separate recognition of interest in the income statement.
Nevertheless, there is a broad consensus that the general IFRS rules do not lead to meaningful accounting of the underlying financial instruments in the case of a bank’s interest rate risk management activities (“accounting mismatch”, lack of transparency). The purpose of specific IFRS hedge accounting rules is therefore to define exceptions to the above-mentioned principles that enable meaningful accounting to be achieved without, however, distorting the fundamental nature of IFRS financial statements or allowing arbitrary accounting.
Remaining differences when applying the DRM model
The DRM model as communicated by IASB until 11/30/2024 to fulfill this complex task can be understood as an ALM process adapted for accounting purposes. Compared to the current IAS 39 Portfolio Fair Value Hedge, the IASB would like to expand the scope of permitted financial instruments, take into account the dynamics of risk management and increase transparency about the effectiveness of risk management activities and their impact on the income statement in IFRS financial statements. Figure 1 provides a schematic overview of the 6 steps of a DRM cycle, which are run through continuously in the DRM model, deliberately avoiding accounting-specific terms.
Figure 1: Schematic representation of the DRM cycle, source: own representation based on IASB Staff paper 4B for IASB meeting September 2021, page 5
The DRM model represents a clear approach to the risk management view. The first 5 of the 6 steps of a DRM cycle are based on the methods and metrics of interest rate risk management. However, the DRM model does not completely adopt the risk management view. Rather, various differences will remain. According to our investigations and simulations to date, the most significant remaining differences are
> Different populations between the ALM and DRM portfolios
Permissible in the DRM model include demand deposits and both puttable and FVOCI classified financial assets. However, equity instruments issued, including AT1 bonds classified as equity under IFRS and financial instruments classified as FVTPL, are still not eligible. An explicit option for offsetting inadmissible transactions, such as proxy hedging as part of the Portfolio Fair Value Hedge in accordance with IAS 39, is not yet provided for in the DRM model.
> Elimination of accounting mismatch only for risk-mitigating activities (asymmetric risk assessment)
The prospective and retrospective DRM tests (steps 2 and 4 in the DRM cycle) and the calculation of the DRM adjustment on the basis of lower-of tests (step 6 in the DRM cycle) mean that the accounting mismatch is only eliminated to the extent of the risk mitigation actually achieved. There is no unconditional recognition of risk-induced changes in the value of hedged items, in contrast, for example, to the recognition of the hedge adjustment in the Portfolio Fair Value Hedge in accordance with IAS 39.
> Different assessment intervals (daily management vs. DRM cycle)
The DRM model is designed as a continuous cycle consisting of 6 steps. The portfolios of hedged items and actual and hypothetical derivatives are continued over the cycles without de- and redesignations. However, new underlying and hedging transactions are only included in the DRM model at the beginning of each cycle. Postings from the DRM model are only made at the end of each cycle. If the length of the DRM cycle, which is not defined by the standard setter, is longer than 1 day, this results in deviations from the usual daily ALM process of a bank.
> Limited mapping of risk management activities for multi-level risk management strategies and the use of various risk metrics
Transactions already included in another hedge relationship, such as asset swap packages in a micro fair value hedge, or transactions denominated in a different currency, such as fixed-interest USD loans that are turned into a variable EUR position with a cross currency swap, may not be included in a DRM model. However, the resulting net positions may be included in the DRM model as “aggregated exposures” under certain conditions.
Banks’ interest rate risk management is usually based on several risk metrics such as pv01, key rate duration, value at risk or nominal exposure, which differ in their nature, e.g. present value vs. nominal, maturity structure, sensitivity, etc., and therefore also in their effect when using the DRM model, sometimes considerably. In contrast, the DRM model is based on a single risk metric, so that the effects of risk management activities on the other risk metrics are not reflected in the DRM model.
Starting points for reducing the differences and the resulting negative implications
The principles-based standard setting that is prevalent in IFRS as a whole and is already recognizable for the DRM model makes it possible to reduce the differences between the risk management and accounting view by means of a methodological design aimed at this within the framework of the individual implementation of the DRM model itself.
The integration of the DRM model into all of the hedge accounting variants permitted under IFRS and the fair value option on the basis of an adjusted hedge accounting strategy also offers starting points for reducing the differences. The use of leeway in risk management can be of lesser importance but useful in individual cases, provided that a reduction can be achieved without causing negative economic effects.
Like the entire functional and technical implementation of the DRM model, all of these starting points are subject to the essential side condition of operationalizability, taking cost-benefit considerations into account. It should also be noted that it is not only the accounting department that receives the results of the DRM model, but also other areas of the bank such as supervisory law, risk controlling and overall bank management.
For example, differing populations between the ALM and DRM portfolios produce differences at various points in the DRM process. In particular, the organic net risk position in the ALM portfolio may differ from that in the DRM portfolio (CNOP) and thus also the managed overall net risk position from the risk mitigation effect (RMI) derived and taken into account in the DRM process. Furthermore, the RMI can be influenced by the fact that the risk limits used in the prospective risk limit test (step 2) relate to the net risk position managed in the ALM portfolio and not to the overall net risk position resulting from the DRM portfolio. The extent of the differences is heavily dependent on the risk metric used, the maturity band structure and also on the type of financial instruments uneligible in the specific portfolio in the DRM process. One possible solution is to map underlying transactions with an offsetting risk position as part of a micro hedge when determining the hedge accounting strategy and therefore no longer include them in the DRM process or only include them with their aggregated net risk position. Depending on the volatility of the net risk position of the underlying transactions not permitted in the DRM process, this may be more or less operable and requires an early portfolio analysis in any case. Although avoiding transactions not permitted in the DRM process as part of risk management is preferable from an accounting perspective, it is often not an option due to undesirable control effects in risk management. It is therefore worth considering granting explicit compensation options not previously provided for in the DRM model.
As already mentioned, the individually defined length of the DRM cycle also has a significant influence on the extent of the differences between the ALM and DRM processes. From a technical point of view, a DRM cycle of one day leads to a largely synchronized ALM and DRM process. A longer DRM cycle, on the other hand, leads to a delayed consideration of new underlying and hedging transactions, which leads to distortions, particularly in the case of retrospective adjustments to the RMI (step 4). A daily DRM cycle is operationally challenging as it requires, among other things, a daily update and valuation of the actual control current Designated Derivatives and the hypothetical Benchmark Derivatives as well as a daily booking process.
Implementation aspects
The potential of these and other points for avoiding differences also depends largely on the chosen technical implementation solution.
Operationalizing a daily DRM cycle can be easier with an ALM-related IT solution, for example, than with an accounting-related or DRM-specific individual solution. In order to realize this potential, an appropriately structured implementation procedure is necessary, which is exemplary outlined in Figure 2.
Figure 2: Outline of the implementation procedure
In the first, functional stage, the key issues to be determined in the DRM model (e.g. risk metrics, frequency of DRM cycles) are to be identified with the relevant functions (e.g. determination of CNOP, construction and portfolio management of hypothetical benchmark derivatives). Possible technical implementation alternatives are to be derived and evaluated for each function. The specifications made here form the basis for the responsibilities and processes to be defined (see (1) in Figure 2).
In the second process-related step, the selected functional characteristics must be compared with the existing structures in risk management (e.g. process for deriving control derivatives) and accounting (e.g. end-of-day processing) in order to work out organizational and process-related options that form the basis for the selection of applications (see (2) in Figure 2). This raises questions regarding the responsibility for DRM-specific calculation processes, such as the derivation of the RMI or the execution of lower-of-tests to derive the DRM adjustments, as well as the provision of the necessary master, transaction and market data. The particular challenge of the DRM model lies in its functional location at the interface between risk management and accounting, which is also of considerable importance for supervisory law and overall bank management.
In the final step, the base of the possible applications are then collected and evaluated. In particular, the dimensions of location, degree of standardization and source of supply based on the company’s own IT strategy as well as information on DRM applications available on the market in the future must be assessed. Only on the basis of the findings from all three levels (functional level, process and organizational level and the application level) can a valid estimate of the costs and benefits for the institution be made (see (3) in Figure 2). The cost-benefit assessment can also be an iterative process (see (4) in Figure 2), in which cost-benefit considerations for implementation and operation are collected and evaluated after the general feasibility check.
Figure 3: Application variants
The discussion on the implementation procedure often starts at an application level, driven by the stakeholders of the applications and processes. This approach is unlikely to lead to an optimized DRM implementation for the institution, but the application variants can support a meaningful structuring of the various implementation options.
Conclusion
The DRM model means significant conceptual changes for hedge accounting. It can lead to a significant reduction in the differences between the risk management and accounting perspectives. Reducing the differences as far as possible is a complex and institution-specific challenge. The DRM model offers numerous technical design options and implementation approaches that can be worked out through targeted analysis and a structured implementation process. The time required for this should not be underestimated, nor should the need for meaningful quantitative simulation capabilities.
However, even with optimal implementation, the DRM model will not completely overcome the differences. It is therefore important, according to Friedrich Christoph Oetinger, to make full use of the design scope of the DRM model, to communicate unavoidable differences transparently to the various stakeholders and to be able to distinguish between the two on the basis of an intelligent analysis.
The views expressed in this article are solely those of the respective authors of the article. They do not, or do not necessarily, reflect those of EY ifb SE or any other member firm of the global network of EY firms, or its subcontractors, members, shareholders, directors, partners or employees (“EY”). EY therefore assumes no responsibility or liability for the content of the views expressed.
