Key takeaways & industry challenges following the ECB TRIM project – a focus on CCR (Part 1)

Fri 07 May 2021

As articulated by the ECB in its recent TRIM reporting, the 236 findings cover different key aspects of supervised entities Internal Model Method (IMM) models & frameworks.  Remediation actions are underway in all institutions, however, we observe across the industry and through our clients persistent & common challenges that need to be tackled in a swift way.

Closing the highest severity findings is expected to be accompanied by a removal of the alpha multiplier “add-ons” and RWA buffers. At a time when capital is still scarce and an expensive resource, working on remediation can be seen as a strong and valuable investment that will sustain long term profitability for global markets activities. 

Among the various topics noted by the ECB, some governance, validation, collateral management investments could help achieve higher ROE. 

Scope and Trade Coverage: TRIM showed that institutions treated not fully modelled IMM trades very differently, some having a systematic carve out of the IMM trades under the standardised approach. In contrast, others retained the trades within the IMM model with more or less sophisticated proxy for the trades exposure profiles. Enforcing a level playing field in that front generally translated into significant capital hikes for institutions allowing not well-modelled trades within the IMM. Work is underway to reintegrate them back with better modelling. This is yet another opportunity to align IMM and front-office xVA pricing better. 

MPOR (margining period of risk) and Trade-Related Cash Flows (TRCF): The first challenge is implementation. For a certain period within each MPOR, the institution needs to assume that it continues to pay trade flows while receiving none. Many pricing libraries used for IMM struggle to capture trade cash-flows that need to be adequately netted for each CSA, even for trades that are fully modelled within IMM. We see investment across all institutions to strengthen these pricing libraries, sometimes nurturing further alignment with the front office xVA pricing libraries. The second challenge is governance and risk management to articulate better the bank Default Management Process (DMP), which is critical to reducing the time during which asymmetric cash-flows payment is assumed that drives most of the RWA inflation via the so-called cash-flow spikes.

Collateral Modelling and monitoring: The TRIM guide insists on reconciliation between the collateral assumed within the model and the “real” collateral effectively held/posted by institutions. Collateral management systems and risk systems are generally not aligned, and such reconciliation can prove very challenging. Unjustified discrepancies will have to be integrated within the IMM. It is an opportunity to bring further alignments to these systems, increase data quality and offer a single and consistent view of the positions at risk. 

Validation: TRIM has highlighted once more the importance of the validation unit and its role in providing effective challenges to the IMM, notably on its backtesting piece. Some institutions have taken this opportunity to redesign their backtesting framework, filling the potential coverage gaps (single risk factor, trade, portfolio), but more importantly (re)designing it in a way that its outcome can be useful to model developers, model validators and model users alike. Clear actions and escalations need to be articulated if and when the backtesting results deteriorate. A sound backtesting framework can be used to identify new “risk not in models”, indicate poor calibration/modelling of risk factors.

Governance: TRIM has spotted many documentation issues (related to the modelling itself and process & infrastructure documentations). Some institutions still do not sufficiently distinguish roles and responsibilities across model developers, model users and model owners. Identification of individuals within the institutions take clear & distinct responsibilities for their mandate, alongside committees with a clear reporting line & escalation process is seen as best practice.

Risk factor modelling and calibration: The simulation of risk factor processes for each relevant asset class plays an important role in a well-defined IMM framework. As observed by the ECB, various SIs had shown weaknesses in the model assumptions and the length of the stress period and the corresponding stress calibration. Different institutions are establishing or improving their risk-not-in-model IMM framework to quantify deficiencies captured by their backtesting.

Time steps and scenarios: The definition of the scenario grid is one of the cornerstones in establishing a precise and accurate IMM. The TRIM report hints that there is no “one-size-fits-all” sort of solution when it comes to defining a robust scenario grid. As this is specific to each FIs portfolio, one of the main challenges in constructing a resilient scenario grid is finding the right compromise between an FIs performance and resources. The most advanced firms are looking into the dynamic grid, where the time points are adapted to each portfolio to capture the main cash flows. TRIM offers some tolerance between the production grid and a daily grid and Monte Carlo error (which decreases when scenarios increase). Firms are optimizing their resource/accuracy trade-off accordingly.

TRIM highlighted the need to continue investing in IMM to maintain the waiver. Despite the capital increase generally seen on the back of this investigation, particularly due to the integration of TRCF in the exposures, overall, the IMM is still much more efficient than the standardized approach from a CCR RWA perspective. Keeping a robust framework in IMM governance, management, development, backtesting, and improvement processes is key for any Global markets division. Mitigation action, adjusting to regulation, and deficiency remediation have proven to be sources of increased risk control and profitability.