Industry
Banking
Banking Services Transformation: Unlocking Exponential Value
Highlights
- The Fundamental Review of the Trading Book (FRTB) reshapes market risk capital by driving greater consistency, transparency, and risk sensitivity across trading activities. While the Standardised Approach (SA) offers regulatory certainty, the Internal Models Approach (IMA) presents an opportunity for banks with sophisticated trading books to improve capital efficiency and risk insight.
- Realising these benefits requires strong data foundations, disciplined governance, and closer alignment between trading, risk, and technology teams. Regulatory expectations around modellability, attribution, and ongoing controls significantly raise the bar for operating maturity.
- Banks must define a phased, well‑governed adoption strategy for FRTB compliance, choosing the approach that best suits their portfolio after a comprehensive analysis of costs, benefits, and trade-offs.
On this page
Introduction
Following the 2008 financial crisis, the Basel Committee on Banking Supervision (BCBS) introduced the Fundamental Review of the Trading Book (FRTB) to strengthen market risk capital rules.
FRTB defines minimum capital requirements, improving resilience under stress while enhancing consistency, transparency, and comparability of market risk calculations across banks. Major global jurisdictions such as the EU, the UK, and the US have adjusted their FRTB implementation timelines to align with geography specific developments and maintain a level playing field. As of early 2026, the implementation landscape has shifted toward a 2027–2028 timeline. Given these developments, financial institutions are moving toward FRTB implementation, putting in place robust controls to soften its impact in the initial years.
FRTB prescribes two approaches for calculating market risk capital requirements—the Standardised Approach (SA) and the Internal Models Approach (IMA). Many financial institutions have adopted SA for FRTB compliance due to lower complexity and costs as well as fewer operational challenges. On the other hand, IMA adoption, though beneficial, has been slow due to implementation complexity, regulatory rigour, and higher operational risk. That said, IMA is gaining momentum—the US recently proposed a calibrated implementation of IMA for FRTB compliance.
IMA adoption in adherence to FRTB rules demands careful desk-level planning, meticulous modelling, provision for flexible switchovers, and appropriate data velocity. Essentially, it is a tightrope walk requiring banks to strike a balance between the certainty of SA and the volatility, but potentially lower capital provision of IMA. Despite these constraints, banks such as BNP Paribas, Deutsche Bank, and Intesa Sanpaolo have sought approval from the European Central Bank (ECB) to use IMA. A key reason for this step could be a desire for more precise measurement of market risk and the potential for lower capital requirements—a leading Japanese bank adopted IMA to significantly reduce market risk capital requirements.
The business case for IMA
Many financial institutions have decided to delay moving to the IMA.
However, players with significant trading books and non-linear portfolios can realise substantial benefits such as better risk sensitivity and lower capital requirements for specific trading desks by adopting IMA. These benefits will be capped by a 72.5% output floor, phased over a few years, with the initial years promising attractive benefits. Even after the full output floor sets in, IMA will continue to benefit players with significant trading exposures. IMA adoption in compliance with FRTB rules, however, needs architectural alignment to tackle challenges throughout the implementation cycle.
The Risk Factor Eligibility Test (RFET) is a core component of the IMA approach to FRTB compliance. It determines if a risk factor is ’modellable’. However, data availability is a challenge for some instruments such as OTC derivatives that lack observable prices. To qualify as modellable, a risk factor must demonstrate a minimum of 24 real prices within a 12‑month period such that no 90‑day window has fewer than four prices. Alternatively, there should be a minimum of 100 ‘real’ price observations over the previous 12 months. If a risk factor is deemed non-modellable, its capital requirement is calculated on a standalone, stressed basis, significantly increasing capital requirements. So, developing ‘modellable’ risk factors is a crucial prerequisite of IMA.
Assigning trading desks for IMA should strike a balance between capital efficiency and compliance. Desks must pass the profit and loss attribution test (PLAT) and satisfy RFET requirements. Utilising granular risk factor buckets produces good results in RFET due to better alignment between risk factors and similar price observations. However, it increases the complexity and operational burden during PLAT as every risk factor must be regularly tested for eligibility and back-tested across all the selected buckets. Consequently, optimisation of risk factor buckets is crucial to pass both RFET and PLAT, as regulators will challenge the results of both tests.
IMA necessitates a shift from value at risk (VaR) to expected shortfall (ES), which introduces difficulties in simulating tail behaviour. Here, non-linear products tend to amplify ES instability and correlation assumptions, thus increasing aggregation complexity. Moreover, ES must be scaled from a 10-day liquidity horizon to longer ones spanning 20, 40, 60, and 120 days. Liquidity horizon determination is subjective; they are not static and change during periods of market stress. The process is compute-intensive, and its management is difficult.
Under IMA, the default risk charge (DRC) replaces the incremental risk charge (IRC). Limited default data, high compute costs, and lack of correlation information pose difficulties in DRC calculation. To obtain an accurate understanding of the impact, millions of combinations of scenarios, desks and risk factors should be considered, which is cost- and compute-intensive.
A major issue with IMA adoption is passing the PLAT. Failing the evaluation requires an immediate switch to SA, changing capital provisioning for the next 12 months. Many desks fail PLAT not because they are wrong, but because data, booking, and pricing alignment is broken.
Even minor changes require extensive diligence and documentation for regulatory approvals, which is difficult to maintain. Moreover, changes create desk-level uncertainty as traders are suspicious of them given that many come with increased compliance burdens, strict desk-level approvals, and the threat of severe capital penalties in case of failure to meet back-testing and PLAT thresholds.
Making IMA work
Banks must make several changes to how they leverage the synergy between the trading desk and line 2 risk to efficiently handle FRTB.
A componentised and rationalised workplace, along with provisions for risk and front office integration and an early warning system (EWS), are the need of the hour to comply with elaborate regulatory requirements. Banks that plan to adopt IMA must focus on the following aspects.
Data management
Ensuring auditable data along with the right risk factor buckets is key to continued RFET success. External and proxy data usage should be documented and governed accurately. Third-party risk should be tracked and monitored. Banks must consider setting up a domain master data service to monitor risk factors.
Componentised IMA risk platform
Banks must establish a robust, auditable scenario management service to manage stressed ES and non-modellable risk factors. Liquidity horizons should follow a robust, policy-driven approach. Policy changes should be versioned and auditable. Modelling and curve construction engines should be centralised and componentised through risk clustering and configuration-based models where possible. Model maintenance should be automated to the extent possible, and models should be periodically validated and monitored.
Banks must build an engine to optimise capital based on Risk-Adjusted Return on Capital (RAROC). The platform must include an EWS to notify the probability of failures and changes to avoid wide fluctuations in capital requirements. The componentised system must include a feature flag mechanism to signal changes in the FRTB regulation or IMA requirements. Leveraging high-performance computing (HPC) technology can accommodate compute-intensive calculations.
PLAT readiness
Integrating risk and pricing engines at the data and component level to create a single source of truth can boost PLAT success. Setting up a shared pricing library will ensure uniform prices for front-office and risk teams. In addition, building a dedicated PLAT analysis layer, separate from the ES engine, and reinforced by application programming interfaces (APIs), root-cause diagnostics, and historical tracking, will enable daily P&L analysis.
Designing for control
While IMA is expected to unlock significant gains in capital planning, it has to be meticulously designed and monitored as it comes with many moving parts. Consequently, quality data, meticulous desk planning and governance, efficient analytics, high-speed computing, and the right degree of configurability and regulatory preparedness are key.
Banks must put in place a codified policy management process where every compliance document is metadata-driven. Adopting an API-driven and componentised system will simplify change implementation.
From theory to action
Successful FRTB IMA compliance demands a phased and well‑structured roadmap, specifically focused on IMA desk build‑out and lifecycle management.
While defining the roadmap, banks must consider the size and nature of the trading book, desks that can benefit significantly from IMA, RoI, and regulatory support.
Phase 1
- Implement SA across all desks, enabling banks to achieve initial compliance.
Phase 2
- Analyse whether the desk is IMA eligible and ready; evaluate the potential benefits of IMA implementation for specific desks.
- Build robust data foundations and risk factor mapping, ensuring data discipline and centralised reference data, including curves and pricing functions.
- Conduct RFET and identify the right risk factor buckets that will ensure a balance between RFET and PLAT.
- Establish ES modelling and stress period selection rules for centralised pricing and risk engines, tail behaviour simulation, choice of liquidity horizon, and stress period selection mechanism.
- Create robust profit and loss attribution (PLA) and back-testing frameworks to reduce failures and provide early warnings.
- Analyse scope and implement DRC.
- Conduct parallel runs with SA and ensure entry and exit governance.
Phase 3
- Strategise for regulatory approval and steady state productisation.
- Establish a mechanism for regulatory engagement and investigation.
How TCS can help
With widespread experience in risk management and deep expertise in market risk data and methodologies, TCS offers a suite of services to support the entire lifecycle of FRTB IMA implementation.
TCS helps financial institutions diagnose the current state and define a roadmap along with the target architecture for FRTB IMA compliance (see Figure 1).
Figure 1: Capabilities that enable TCS clients to adopt the internal models approach for FRTB compliance
Going forward, we expect intelligent AI agents to foray into the IMA ecosystem, helping to predict PLA failure, detect emerging non-modellable risk factors, recommend hedges to stabilise capital, and generate dynamic stress scenarios. This will offer experts a real-time view of emerging risks to IMA desks and help make timely provisions. We believe AI systems will flag potential IMA failures through an integrated dashboard, improving capital efficiency.
Furthermore, the risk function will move toward agentic AI-driven automation, helping drive compliance with the key FRTB objectives of rigour in data management and model governance. We believe the entry of AI agents will help avert failures and explain the reasons for systemic issues, which will ensure the continuity of IMA for allocated trading desks, prevent sudden spikes in capital requirements, proactive course correction before non-modellable risk factors emerge.
While both SA and IMA offer distinct advantages, there are compromises as well. Banks must carefully evaluate the benefits, costs, and the trade-offs, and ultimately choose the approach that best suits their portfolio.