Regulatory Compliance

Navigating the Challenges of FRTB

The Fundamental Review of the Trading Book (FRTB) framework promulgated by the Basel Committee on Banking Supervision establishes new standards for banks for calculating the minimum capital requirements for market risk. The Basel Committee expects national banking supervisors globally to adopt final regulations implementing the FRTB standards effective January 1, 2019 with reporting by global banks under the new framework to commence by the end of 2019. 1

In line with current regulations codified in Basel II and Basel III, the FRTB standards define two separate approaches for quantifying an institution’s exposure to market risk: (1) the standardized approach and (2) the internal models approach. Under FRTB, both approaches have been overhauled in an effort to add greater rigor in how banks and regulators identify and calculate market risks. In addition, compared with prior Basel Committee standards, the new standards:

  • Apply an expected shortfall measure (the successor to value-at-risk or VAR) to calculate capital adequacy under stress
  • Require more elaborate analysis of market liquidity risks
  • Modify the prior rules separating a bank’s trading book from its banking book

The Standardized Approach

According to the Basel Committee, the standardized approach is intended to serve as a “credible fallback” and a floor in relation to the internal models approach, while still providing an appropriate standard for banks “that do not require a sophisticated treatment for market risk.”2 In order to attain regulatory approval to use the internal models approach, a bank must pass a series of qualitative and quantitative tests. A key portion of those tests assess the quantity and quality of market data inputs used in the bank’s risk models. (All banks, including those approved for the internal models approach, will also be required to calculate risk according to the standardized approach.)

A bank’s risk capital calculations must incorporate default risk, interest rate risk, credit spread risk, equity risk, foreign exchange risk and commodities risk, among others. Securitized products (such as Residential Mortgage-backed Securities) are ineligible for the internal models approach; their capital charges can be calculated using the standardized approach only.

FRTB includes a prescriptive boundary between trading and banking books, with few exceptions to change between the two

The Internal Models Approach

The FRTB standards describe the criteria and requirements to which banks seeking approval to use internal models will be subject. The process for determining the eligibility of trading activities for the internal models-based approach is comprised of three steps:

Step One

In the initial step, the regulator assesses the bank’s internal risk capital model and related control processes against a series of qualitative and quantitative criteria.3


The second step requires breaking out a bank’s market risk exposures at the trading desk level. An institution seeking permission to use the internal models approach must submit a separate application on behalf of each of its trading desks for which the bank seeks to use internal models.4

Then, the bank’s internal models must be tested against 12 months of historical profit and loss data for each trading desk to count "exceptions" and "breaches" (deviations between model-calculated and actual daily P&L for the applicable desk). If the number of deviations exceeds specified levels, the specific desk is denied permission to use the internal models approach.


The third and final step consists of analyzing risk factors within a bank’s internal risk models for regulatory capital, to determine whether each individual risk factor qualifies as "modellable." The "modellability" question hinges on the amount of qualifying market data the bank can access to feed its model.

In the Basel Committee’s words: "For a risk factor to be classified as modellable by a bank, there must be continuously available ‘real’ prices for a sufficient set of representative transactions."5 The standards define a "real" price as either a transaction conducted by the same bank or at arm’s length by other parties if verifiable, or a committed quote.

An institution that has obtained regulatory approval to use the internal risk model approach shall calculate its minimum regulatory capital requirement by the aggregation of the following three components: global expected shortfall; a default risk charge; and a "stressed capital add-on" meant to capture risk associated with non-modellable risk factors in model-eligible desks. One facet of the global expected shortfall component is a stress test calculation with an observation horizon going back to 2007, hence, requiring access to historical market extending data back to at least January 2007.

ICE Data Services’ Approach for FRTB Data Offering

To initiate a dialogue with the industry on FRTB challenges and strategy, we sponsored an industry round table in London on January 31, 2017 - a start to a series of client engagements on the topic.

We’re developing a platform designed to assist global banking institutions in managing market data content associated with the FRTB standards. The offering will support a broad range of asset classes, including commodities, foreign exchange, equities, credit (credit default swaps/fixed income), and interest rates. Our approach aims to provide you with data that can help support your assessment of which risk factors are eligible to be included in the bank’s internal models according to the FRTB standards. The main components of our approach are:

Historical valuation and pricing data that generally extend back to January, 2007

"Real price" evidentiary support

New FRTB-specific security-level attributes

Data pooling service

FRTB User Interface

The repository of global market data underlying the our FRTB offering will draw upon our pricing and analytics businesses, trading platforms, and proprietary data contributed by participating institutions. Data generated by existing ICE businesses will be augmented with additional FRTB-specific reference data and risk inputs. The platform incorporates data quality control measures including data cleaning, identifying committed quotes, and "de-duping" (identifying and excluding duplicate transactions).

The offering will provide participating banks with considerable flexibility. User groups contributing data on transactions and committed quotes will have the ability to select the license associated with the contributed data set. In addition, each user will have the ability to specify categories of market data to include or exclude in tests of that user’s internal risk models. For example, if a particular bank’s policy excludes certain sources of market data from being counted in market risk assessments, our FRTB service will allow them to exclude such data from their calculation of inputs necessary to support a "real price" determination. Similarly, a participating bank will have the ability to override the system’s default setup for identifying duplicates by inputting the user’s definitions of duplicate transactions prior to running the service’s de-duping function.

  1. Basel Committee on Banking Supervision, "Minimum capital requirements for market risk," January 2016.

    Policy rationales were set out in three earlier consultation papers by the committee:

    "Fundamental review of the trading book," May 2012
    "Fundamental review of the trading book: A revised market risk framework," October 2013
    "Fundamental review of the trading book: Outstanding issues," December 2014

  2. "Minimum capital requirements for market risk," op cit., page 1
  3. The qualitative criteria include having an independent risk control unit, conducting regular back-testing and profit and loss attribution programmes, a distinct unit responsible for model validation, board and senior management involvement in the risk control process, conducting routine and rigorous stress testing, and regular audits of its risk control process by either an internal unit or an external auditor. See "Minimum capital requirements for market risk," op cit., pages 50-52.
  4. The P&L attribution and back-testing requirements are spelled out in "Minimum capital requirements for market risk," op cit., page 57.
  5. "Minimum capital requirements for market risk," op cit., page 57.