FRTB Audit Support & Documentation

During FRTB audits, supervisory authorities assess several core areas: the correct assignment of positions to the trading or banking book (boundary), the quality and documentation of internal models (IMA) or the application of the standardised approach (SA), data quality and availability for risk factor calculations, governance structures including the three lines of defence, and backtesting results and P&L attribution tests. Typical focus areas include the completeness of model documentation, the traceability of model decisions, and the operational implementation of regulatory requirements in day-to-day business.

The preparation follows a structured process: first, we conduct an audit readiness assessment, reviewing all FRTB-relevant documentation, processes and governance structures against supervisory expectations. Based on this, we create a gap analysis with specific recommendations. During the implementation phase, we close identified gaps in documentation and processes. Before the actual examination, we conduct mock audits that simulate examination situations and prepare your teams for typical examiner questions. During the audit itself, we provide both technical and organisational support.

For a successful FRTB examination, banks require at minimum the following core documents: complete model validation documentation with backtesting results and P&L attribution, the trading book/banking book boundary policy, risk management policies with responsibilities and escalation processes, data quality policies and process descriptions for risk factor collection, documentation of the internal model approval process, and evidence of regular model monitoring and re-calibration. ADVISORI prepares and reviews this documentation in accordance with current CRR III and EBA standards.

Under the IMA (Internal Model Approach), the examination focuses on model validation: examiners assess the quality of Expected Shortfall calculations, the treatment of non-modellable risk factors (NMRF), backtesting results and P&L attribution. Under the standardised approach (SA), supervisors primarily examine the correct application of sensitivity calculations (SBM), assignment to risk classes, default risk charge calculation and residual risk add-on determination. In both cases, complete documentation of the calculation methodology and data basis is critical.

From our experience in FRTB audit support, the following findings occur most frequently: incomplete or outdated model documentation, insufficient evidence of data quality and risk factor availability, gaps in governance documentation particularly regarding responsibilities between risk management and front office, missing or incomplete backtesting evaluations, lack of traceability in trading book/banking book assignment, and deficiencies in documenting model changes and re-calibrations. An early gap analysis by experienced audit support consultants can significantly reduce these risks.

The preparation duration depends on the maturity of the existing FRTB implementation. For banks with fundamentally existing documentation and processes, we estimate

8 to

12 weeks for comprehensive audit preparation. Where there are significant documentation gaps or governance deficiencies, preparation can take

4 to

6 months. We recommend beginning the readiness assessment at least

6 months before a scheduled audit. This allows sufficient time to address identified weaknesses and conduct mock audits.

Mock audits simulate a real examination situation and offer three key advantages: first, they identify remaining documentation gaps and process weaknesses under realistic conditions. Second, they prepare your specialist teams for typical examiner questions and interview situations, enabling structured and technically sound responses. Third, they test the organisational audit infrastructure, meaning the ability to provide requested documents quickly and coordinate across departments. Experience shows that a professionally conducted mock audit reduces audit duration by up to

40 percent and significantly decreases the number of critical findings.