FRTB: An Overview of Key Components

Fundamental Review of the Trading Book: An Overview of Key Components

By Sergiy Ladokhin & Aijaz Siddique / March 2021

Executive Summary

Fundamental Review of the Trading Book (FRTB) is a regulatory framework to calculate capital requirements for banks’ market risk. The FRTB is a long-awaited response of the regulators to the financial crisis of 2008, and it overhauls multiple elements of the current capital calculations. The FRTB consists of various elements; the banks will need to report to the national regulators based on some of these elements starting as early as September 2021.

Let us summarise the key elements of the FRTB regulation. First of all, the framework defines a strict boundary between the trading and banking books. All the securities that fall under the trading book will be subject to FRTB calculations. The regulation defines requirements on banks’ internal rules for assigning products to either the banking or the trading book. The FRTB allows several approaches for the calculation of market risk capital requirements:

Standardised Approach (SA) is a set of rules to calculate capital requirements based on regulatory prescribed parameters. The SA calculations are detailed in nature and are based on the Sensitivity-Based approach, namely sensitivities of the positions to a pre-defined risk factor.
Internal Models Approach (IMA) contains a number of changes compared to the current requirements. A switch to a new risk measure (Expected Shortfall) with multiple liquidity horizons and a P&L attribution test are the main new elements.
Simplified Standardised Approach (SSA) is a light version of SA and can be used by smaller local banks without a significant trading book.

FRTB SA will be used as the floor to calculate the capital requirements according to FRTB IMA. It means that even if a trading desk decides to use IMA for capital reporting purposes, the SA approach should always be calculated.

Successful FRTB implementation poses a number of challenges for financial institutions. The FRTB puts a significant emphasis on the data requirements and data granularity. Additionally, banks may consider a cost-benefit and impact analysis to determine whether to implement the IMA approach or stay with the SA.

The FRTB regulation will affect all Dutch banks. However, the level of the impact will differ significantly per financial institution, depending on the business model and the chosen type of the FRTB approach. Some banks will need to build governance around the split between the banking and trading book only, while others will implement the new models on top of the new data vendors’ feeds and other significant IT changes. Nevertheless, all banks will be involved with some level of reviews of the market risk activities and associated risk calculations.

This paper is the first in a series of white papers that Probability & Partners will publish about the FRTB regulation. This article provides a general overview of the FRTB, while the Standardised and Internal Models Approaches’ details will be covered in the consequent white papers.

Introduction

The financial crisis of 2007-2008 exposed deficiencies in the Basel capital framework. These were partially addressed in the BASEL 2.5 framework of 2009 by the Basel Committee on Banking Supervision (BCBS). However, it was clear that a more thorough review of the trading book regime was required to address the framework’s shortcomings. This comprehensive assessment, also called Fundamental Review of the Trading Book (FRTB), has led to sweeping reforms of the market risk framework to make banks more resilient to crisis.

In this paper, we examine the critical changes to the revised market risk framework. The next section contains a general overview of the FRTB framework, followed by a detailed discussion of the FRTB Standardised Approach (SA) and the FRTB Simplified Standardised Approach (SSA). This paper gives a high-level overview of the new regulatory standards without going into all the methods’ calculation details. It is primarily focused on a general overview of the FRTB framework, with a greater focus on the SA approach. Final section describes the new framework’s challenges for banking institutions.

FRTB Regulation – General Overview

After several consultative papers, the Basel Committee on Banking Supervision (BCBS) published its Standards for Minimum Capital Requirements for Market Risk [1], also known as the Fundamental Review of the Trading Book (FRTB), in 2016. These were further revised by January 2019, with February 2019 being the latest version [2].

The initial deadline for the implementation of the FRTB was January 1, 2022. Due to the impact of the COVID-19 pandemic, the deadline has been revised to January 1, 2023. However, certain jurisdictions (i.e., Capital Requirement Regulation 2 in Europe) will have to start reporting market risk under the FRTB Standardised Approach (SA) in September 2021.

FRTB incorporates several crucial enhancements to the market risk framework. First and foremost, the new rules propose an objective boundary between the trading book and the banking book, supplemented with a list of instruments that must be included in the trading book or excluded from it. Restrictions have been imposed to limit the movement of the instruments between the books. These include:

Receiving supervisory approval for any deviations from the prescribed list of trading book instruments.
Disallowing capital benefits resulting from switching instruments between the books.

Any re-classification of instruments would be only possible in extraordinary circumstances, which implies that reasons such as changes in market illiquidity or trading intent would not suffice anymore. Banks have been advised to enhance the trading book’s internal policies and procedures to ensure proper compliance with these rules.

BCBS has proposed a revised Standardised Approach (SA) to compute the trading book’s risk capital requirements. FRTB SA effectively captures risks in the trading book by incorporating weighted risk sensitivities for different asset classes and aggregating them with regulatory prescribed correlations. It is also more closely aligned with banks’ risk management systems due to its reliance on banks’ pricing models to calculate risk sensitivities. These adjustments enhance the risk-sensitivity of the FRTB SA so that it can serve as the floor to the FRTB Internal Models Approach.

The corresponding revised Internal Models Approach (IMA) for the trading book results in a more coherent and comprehensive methodology. It adequately accounts for tails risks and market illiquidity by replacing Value-at-Risk (VaR) with Expected Shortfall (ES) and introducing varying liquidity horizons for instruments. The Expected shortfall approach determines the riskiness of a position by considering both the size and likelihood of losses above a certain level.

However, the approval to use ES at the trading desk level requires strict compliance with qualitative (infrastructure) and quantitative (back-testing) criteria put forward by the BCBS.

Profit and Loss Attribution Test (PLAT) determines whether banks can use internal models to estimate market risk and compute capital requirements. It is calculated as the difference between risk-theoretical P&L (daily P&L based on the models that banks use to calculate ES) and actual P&L (marked-to-market) value of instruments.
Back-testing assessment (conducted at both bank-wide and individual trading desk levels) determines whether the observed outcomes conform to VaR measure calibrated to 97.5% and 99% level of confidence. Although the ES is proposed as a risk measure, the back-testing will continue to rely on the well-developed methodology for VaR back-testing.

The Expected Shortfall approach determines the riskiness of a position by considering both the size and likelihood of losses above a certain level.

Banks will have to invest in additional IT resources to support these requirements.

Another significant change that adds to this burden concerns the eligibility test for risk factors before they are included in the Expected Shortfall approach. Risk factors must have adequate real observations over a specific period. Otherwise, they are capitalised by adding a stressed charge to the ES approach for modellable risk factors.

FRTB Standardised approach

The amendments to the market risk framework revise the standardised approach to make it a credible
fallback and the floor to the FRTB IMA at the trading desk level. According to the Basel III reform package, there will be a 5-year transitional period for implementing the floor. The floor level will gradually increase from 50% in 2023 to 72.5% in 2028.

Specifically, capital requirements under FRTB IMA will be floored at 50% of the FRTB Standardised Approach in 2023 and 72.5% in 2028. This output floor is not binding during the first two years of the transitional period (2022 and 2023). Therefore, no additional capital is required, and the full benefit of moving from FRTB SA to FRTB IMA will apply.

The adjusted FRTB SA is composed of three components that quantify market risk in the trading book:

Sensitivities-based Method (SBM).
Default Risk Charge (DRC).
Residual Risk Add On (RRAO).

The SBM is measured and aggregated for each of the prescribed risk classes, which include General Interest Rate Risk (GIRR), Credit Spread Risk Non-Securitization, Credit Spread Risk Securitization1, Credit Spread Risk Securitization Non-Correlation Trading Portfolio, Equity Risk (EQ), Commodity Risk (CM) and Foreign Exchange Risk (FX). To this sum, we add the capital charges for DRC and RRAO.

Sensitivities-based Method

The sensitivity-based approach captures linear (Delta and Vega) and non-linear (Curvature) risk sensitivities. Both Delta and Vega risks are price sensitivities of instruments to risk factors prescribed by regulation, for example, equity spot price and volatility in the equity risk class.

In comparison, Curvature risk captures the incremental non-linear risk not captured by the Delta charge. It can be described as a measure of convexity or second-order relationship between instrument prices and risk factors. The charge is calculated by doing a complete re-valuation of the instrument price by bumping the risk factor under a set of different stressed scenarios. By its nature, the Curvature risk factor is like Gamma sensitivity. A representative step-by-step process to calculate the risk sensitivities for any risk class can be described as:

Measure the relevant sensitivities (Delta, Vega, and Curvature) for all the trades in scope.
Assign the sensitivities to a risk class and an appropriate bucket within that risk class. Net the sensitivities to the same risk factor.
Calculate risk-weighted sensitivities by multiplying the sensitivity with the corresponding risk weight.
For each bucket, aggregate the sensitivities according to a specified correlation.
For each risk class, aggregate the sensitivities across all the buckets according to another set of correlations.

The SBM is more refined than previous methodologies. Its risk-sensitive nature arises from its granularity and the use of risk-weights. Sensitivities are calculated for instruments in scope before being assigned to risk classes and an appropriate bucket within that risk class. For example, for the equity risk class, there are 13 buckets in total. For instance, bucket 1 holds risk sensitivities to instruments that have exposure to large- cap emerging market stocks within the consumer goods and services sector. While for General Interest Rate Risk (GIRR), each currency has a separate bucket, all risk factors in risk-free yield curves for the same currency are grouped in the same bucket.

Within each bucket, sensitivities are weighed according to specific weights. Aggregation takes place within the bucket and then across the buckets according to pre-defined correlations. The weight and correlation parameters are prescribed by regulation considering the risk class and characteristics of risk factors. We provide a holistic overview of the sensitivities-based method that includes all risk classes in the figure below.

Figure 1: General overview of the calculations under the Sensitivities-based Method.

2. Default Risk Charge

The SBM is further enhanced by the default risk charge (DRC), which computes the liquidation risk not accounted for in the SBM. Like the SBM, the specific steps for the DRC charge involve calculating the jump-to-default risk for all trades in scope, netting it for the same obligors, and risk-weighting the netted risk for the credit quality of the obligor. Weighted risk sensitivities are then assigned to buckets (corporates, sovereigns, and local governments), allowing hedging benefits before being aggregated.

Jump-to-default can be described as the risk that a financial product, whose value directly depends on one or more entities’ credit quality, may experience sudden price changes due to an unexpected default of one of these entities.

In some cases, a look-through approach may have to be adopted. For example, for derivatives on indices, a look-through approach would involve calculating the derivative’s jump-to-default value for each constituent separately.

3. Residual Risk Add-on

A charge for residual risk add-on (RRAO) has been included in the FRTB Standardised Approach
to account for residual risks, particularly related to instruments with exotic exposures (non-linear payoffs). Examples of exotic exposures include gap risk, correlation risk, behavioral risk, and others, which are outside of the scope of the Delta, Vega, and Curvature charges. For example, a callable bond may be subject to behavioral risk because of the possibility that an investor calls it earlier due to unexpected or unforeseen reasons.

RRAO is calculated as the sum of the gross notional amount of instruments, subject to particular residual risk, multiplied by a regulatory prescribed risk weight for that residual risk.

FRTB Simplified Standardised Approach

It is generally acknowledged within the industry that the FRTB SA, due to its complexity, might lead to implementation challenges for many banks. Notably, banks would require sufficient infrastructure for computing the SBM under FRTB SA. For these reasons, the committee developed a proposal for an
alternative called the FRTB Simplified Standardised Approach (SSA). The primary adjustment is a simplified version of the SBM called the Reduced Sensitivities-Based Method (RSBM) while leaving the Default Risk and Residual Risk Add On charges as they are. The RSBM makes several adjustments to the SBM, such as removing capital requirements for Vega and Curvature risks. The RSBM would be applicable only for banks that meet specific qualitative and quantitative criteria. Banks intending to use the reduced method must fulfill the following requirements, based on a
quarterly assessment:

The bank must not be a Global-Systematically Important or Domestic-Systematically Important
Bank.
The bank must not be engaged in writing options (except for back-to-back options and covered
options whereby the bank owns the securities it may need to deliver under the terms of the option).
The bank must not use the internal models approach for any of its trading desks.
As measured by the bank’s relevant accounting standard, the bank’s total non-derivative trading
book assets and liabilities plus the sum of the gross fair value of its trading book derivative assets
and liabilities must not exceed a certain level.
The bank’s total market risk-weighted assets, when using the RSBM, divided by its total riskweighted assets, should be less than a certain threshold.
The bank must not hold any correlation trading positions.

Challenges

The new framework for market risk capital requirements is more complex and poses significant challenges to banks for several reasons. The main difficulties lie with the increased requirements for data used by banks. For example, risk-weights within the SBM depend on industry, sector, and regional information on the sensitivities. Some of the inputs may not be available yet for all sensitivities in the banks’ systems and may need to be sourced. Complexities can also arise due to the comprehensive nature of the framework. A look-through approach may have to be implemented to compute default risk, contributing to the data problem.

The potential data issue is particularly pronounced for the IMA approach, which enforces tight requirements for IMA approval at the trading desk level. To comply, banks will have to ensure that they source the necessary data for the P&L attribution tests and back-testing requirements. Furthermore, the frequency of observable outcomes for risk factors will have to be maintained before being used as an input for the Expected Shortfall approach. Possible solutions can consist of using new data vendors and implementing a single “golden record” data set.

As a result of FRTB implementation, the banks will need to update their IT systems and reporting capabilities. The new regulatory standard will trigger an update through the whole market risk calculation chain.

The SA approach introduces new sensitivity calculations (or Greeks) for the products in scope. A typical example is Curvature (Convexity); however, Delta and Vega’s calculation per supervisory bucket can also be viewed as a special case. Although the new sensitivity measures are relatively straightforward by themselves, they require changes to the Greeks calculations, which are usually part of the pricing models as opposed to the market risk calculations. Depending on the particular implementation, that might trigger a review of some elements of pricing models.

FRTB can have an impact on the business processes and organizational structure of banks. This is mainly driven by the new rules around the split between the trading book and banking book and the rules around trading desks. As a result of these changes, the governance (policies and procedures) must be updated to comply with the new rules. Since FRTB is a new regulation, there are several technical points on which the regulators do not have clarity yet (a typical example is the treatment of CIUs in SA). However, through consultations with the industry, such technical points will be resolved over time.

It will be prudent to do a cost-benefit analysis between IMA and SA. For many banks, implementing and maintaining IMA is associated with high costs. However, a potential reduction in capital is not always apparent. The cost-benefit analysis would involve partial implementation of the IMA (at least in a prototype form) to estimate the impact of the switch. Such a study would usually happen once SA is implemented or almost implemented within the bank.

We should note that IMA itself poses several new challenges which are worth mentioning. For example, for the P&L attribution test, alignment is necessary for pricing models (and input data) used by front-desk and risk models. The scarcity of data for certain risk factors and the application process to the regulator authorities for IMA approval are hurdles that banks need to overcome. However, in this article, we will not discuss IMA in detail. We plan to have a dedicated white paper on the FRTB IMA, the associated challenges, and possible solutions.

Despite some challenges, FRTB SA implementation also gives several opportunities for financial institutions. Banks could perform a review and update their data infrastructure; such an update could improve the quality of the risk calculation and greater automation of the data gathering process. FRTB can allow the banks to streamline and automate operations related to market risk, pricing, and validation. A holistic approach to the FRTB SA implementation could lay some quality groundwork which could help with a potential IMA implementation.

Conclusion

The Basel Committee on Banking Supervision (BCBS) published changes to the market risk capital requirements in early 2019, the Basel III framework’s last important element. These requirements are generally referred to as the Fundamental Review of the Trading Book (FRTB). The banks should adopt the new requirements by January 1, 2023, although the capital requirements calculated using the Standardised Approach should be reported to regulatory authorities as early as September 2021.

The FRTB is a set of complex regulatory requirements that, next to the rules about the split of the banking and trading books, cover three different capital calculation approaches. These are the Standardized Approach (SA), Internal Models Approach (IMA), and the Simplified Standardized Approach (SSA). The SA is the back-bone of the FRTB regulation because it will be used as the floor in capital requirements. We expect most banks to implement SA first, run quantitative studies on the IMA’s cost-benefit, and only then consider IMA implementation. Finally, smaller banks can choose the SSA approach, a simpler version of SA. Although much simpler compared to SA, to qualify for SSA, the bank must meet a set of regulatory criteria.

FRTB implementation poses several challenges for financial institutions. The increased requirements

towards input data, its quality and granularity are the hardest challenges for banks to meet. However, we can also see that banks can take the FRTB as a chance to review internal processes, procedures, models, and governance used in market risk management. By following the strict regulatory requirements, banks can increase the use of automation and IT technologies throughout the front-, mid-and back-office processes. These steps will result in stronger, well capitalised, more transparent banks and a less risky financial system as a whole.

Download White Paper

Contact

Sergiy Ladokhin

Senior Quantitative Consultant

E-Mail

Pim Poppe

Managing Partner

E-Mail

Contact

If our approach appeals to you and you want to know more about Probability & Partners, you can fill in the form or contact one of our people directly.

Executive Summary

This paper is the first in a series of white papers that Probability & Partners will publish about the FRTB regulation. This article provides a general overview of the FRTB, while the Standardised and Internal Models Approaches’ details will be covered in the consequent white papers.

Introduction

FRTB Regulation – General Overview

The Expected Shortfall approach determines the riskiness of a position by considering both the size and likelihood of losses above a certain level.

FRTB Standardised approach

Jump-to-default can be described as the risk that a financial product, whose value directly depends on one or more entities’ credit quality, may experience sudden price changes due to an unexpected default of one of these entities.

RRAO is calculated as the sum of the gross notional amount of instruments, subject to particular residual risk, multiplied by a regulatory prescribed risk weight for that residual risk.

FRTB Simplified Standardised Approach

Challenges

Conclusion

Contact

Sergiy Ladokhin

Pim Poppe

Visiting Address

Postal address

Information

Contact Sector Leads

Executive Summary

This paper is the first in a series of white papers that Probability & Partners will publish about the FRTB regulation. This article provides a general overview of the FRTB, while the Standardised and Internal Models Approaches’ details will be covered in the consequent white papers.

Introduction

FRTB Regulation – General Overview

The Expected Shortfall approach determines the riskiness of a position by considering both the size and likelihood of losses above a certain level.

FRTB Standardised approach

Jump-to-default can be described as the risk that a financial product, whose value directly depends on one or more entities’ credit quality, may experience sudden price changes due to an unexpected default of one of these entities.

RRAO is calculated as the sum of the gross notional amount of instruments, subject to particular residual risk, multiplied by a regulatory prescribed risk weight for that residual risk.

FRTB Simplified Standardised Approach

Challenges

Conclusion

Contact

Sergiy Ladokhin

Pim Poppe

Visiting Address

Postal address

Information

Contact Sector Leads

Download paper