With the 2022 Fundamental Review of the Trading Book (FRTB) deadline looming, banks are fast coming to grips with the amount of work still to be done to achieve a successful implementation
In August, attendees of Risk Australia were surveyed on their preparedness for FRTB. More than half of respondents said they were looking to implement an internal models approach (IMA), with the most pressing challenge being to improve their integration across credit, planning, treasury, capital calculations and regulatory reporting.
Primary concerns around implementation include organisation-wide integration of calculation methods -namely across credit, planning, treasury, capital calculations and regulatory reporting. Ioannis Akkizidis, lead technology product manager, and Sam Mukhopadhyay, director, risk and finance, Asia‑Pacific, at Wolters Kluwer explore the FRTB implementation challenges firms face, and how banks in Australia and New Zealand can prepare for the transition.
How can banks best analyse potential cost implications of the standardised approach (SA) versus the IMA?
Ioannis Akkizidis: Credit financial institutions with trading activities must perform an impact analysis on implementing standardised models and – for more advance trading portfolios – internal models. The measurement of such impacts should consider the costs of capital adequacy and implementation.
The minimum effort and implementation cost to trading desks is thus to perform the analysis of the SA by employing the sensitivities-based method, the default risk charge and the residual risk add-on.
More sophisticated trading desks will implement internal models by including "tail risk" events, default risk and stressed capital add-on. For this, banks must also apply both backtesting and profit-and-loss attribution tests, which determine eligibility and measure the robustness of the internal model. Although the IMA may demand greater implementation efforts, a robust model will result in stable and effective trading outcomes, as measured by the comparison between the model-generated risk measures and the actual profit or loss.
This can result in higher income and minimised losses, and thus optimal cost of capital. However, any failure in performing SA and IMA analysis may have implications on the cost of regulatory capital.
Sam Mukhopadhyay: Typically, the hurdles for a bank to justify the IMA have been high, owing to both complexity and qualifying criteria as defined by the Bank for International Settlements. Issues have ranged from the transparency of the modelling approach, the defensibility of the regulators themselves, consistency in calibrating stress scenarios, and proof points around how these materially impact capital at a given point in time or in the future.
As a result, only more sophisticated banks have attempted the IMA route, and the industry as a whole is still evaluating the options.
In terms of cost and effort, the SA also demands deep exploration of risk factor sensitivities and impact analysis on model choices, by virtue of the SA methodology prescribed. So, while the demands are not as onerous as those for IMA, the benefits in terms of capital management are substantial. Therefore, for many banks, taking up the SA is the more prudent choice.
Given the implementation timeline of 2022, what infrastructure or processes need to be put in place now?
Ioannis Akkizidis: For implementing FRTB, banks must have in place infrastructure on data and analytical tools. Data must be processed and aligned to the FRTB requirements, and also be available at the unified model consistent with FRTB analysis parameters.
As such, data must be well defined, with shared views across both banking and trading books. This will ensure minimum cost and effort in data management. It will also provide a common understanding of the input information, the output result and reporting taking place across different trading desks, banking books and supervisory authorities.
Infrastructure using a single calculation engine will manifest consistent analysis across various desks, and can provide dynamic and robust risk management without the danger of inconsistent risk analysis outcomes.
Sam Mukhopadhyay: Key elements of data management infrastructure should include consistency across factors, positions and exposures; the right level of granularity; the ability to establish data lineage; and auditability. Together, these factors make the models and methods transparent and defensible. Therefore, banks that have not made the right investments will need to upgrade their data infrastructure.
In addition, given the underlying FRTB requirement of credit and market integration requirements, the data framework should address the need for a common pool of risk factors that can be jointly modelled and stressed across risk classes – this is a critical factor. So, beyond data, the solution choice must also address an integrated credit and market framework.
What unexpected challenges might financial institutions face – especially regarding data processing and pricing models?
Ioannis Akkizidis: One of the main challenges banks face is inconsistent data and calculation processes. Pricing models rely on the types of financial instruments and external market data, with dependencies and correlations of models and underlying data playing a significant role.
Sam Mukhopadhyay: Some banks have made progress with initiatives for the Basel Committee on Banking Supervision’s standard number 239, known as BCBS 239 – giving them an advantage regarding data management. Those that have not paid specific attention to this area are not only exposed to broader model risk, they are also likely to encounter serious challenges in model validation, stress-testing and the reconciliation of results from multiple model outputs.
Will variations in regional implementation result in regulatory arbitrage? Which jurisdictions are likely to set a precedent for others?
Ioannis Akkizidis: In principle, the actual capital requirements conforming to local and regional regulatory authority jurisdictions must always be higher than the capital that would have fulfilled the relevant Basel framework. The differences in jurisdictions defined across local regulatory authorities should result in a minimum delta of capital impact.
If these principles fail, a regulatory arbitrage may arise across different market segments and geographic regions where the bank’s trading desk might operate.
Sam Mukhopadhyay: It varies by jurisdiction. A recent survey of capital adequacy returns from banks headquartered in Australia and New Zealand indicates that the prudential standards set by the local regulators are more rigid, attracting a higher quantum of risk capital at deal/exposure level. For example, in many instances, an Australian bank has been outpriced in the market while bidding for a global asset.
Currently, there is an effort by regulators to harmonise these standards, to prevent inequity for banks in Australia and New Zealand.
This article was recently featured in Risk.net, to view the article please follow this link.