First Reds Then Black

In January 2016, the Basel Committee on Banking Supervision published the final rules resulting from its Fundamental Review of the Trading Book (FRTB). The rules are due to come into effect at the end of 2019 and are encapsulated in BCBS 352, Minimum Capital Requirements for Market Risk.

The economic impact on all impacted banks is large. According to an impact study published by three industry groups on April 18th, 2016 the new rules will increase market risk capital requirements by between 1.5 and 2.4 times. However the economic burden is not the only problem and quite possibly the least of the problems in each bank. The effect on the way banks operate will be equally significant:

• The changes needed to provide a robust environment to ensure compliance and minimise the economic impacts will affect all functions, not just risk management.
• Both the standardised approach (SA) and the internal model approach (IMA) have changed from earlier iterations. Existing systems may not be suitable without significant investment/upheaval.
• IMA approvals will now be granted at the individual trading desk level rather than firm-wide.
• Any desks given IMA approval will have to undertake new and stricter performance tests.
• Failing tests need to be reported immediately and the bank will have to revert to the SA.
• Failure for a model means the bank will have to re-apply from the start

The ultimate aim of FRTB is to tighten the loopholes of the current Basel 2.5 regime and ensure that banks are able to withstand any potential market crash. FRTB imposes the most stringent capital requirements of any regulation to date and involves significant required challenges for the Risk, Treasury, Capital Management departments and the trading activity of banks. (See Figure 1)

The Red Balls

• The output of FTRB isn’t a risk measure, but a capital charge:

The purpose of FRTB is to calculate the capital charges on your business, replacing the existing Basel approach. FRTB (as you will see below) has a greater level of structure to the required calculations, a more prescribed approach on how to model your business, and a greater impact on how your entire trading business is run.

• The boundary between the Trading book and Banking book needs to be tightly defined

FRTB requires that business be allocated to either the trading or banking book, and not jump the boundary in an arbitrary way. This means re-visiting any previous approach to modelling this structure, embedding it in an electronic representation, and ongoing monitoring.

• The way market risk is measured is changing

The new requirements will impose significant changes to measuring Market Risk (which will become inputs to the Capital Charge). Moving from VaR to Expected Shortfall will increase the amount of computational effort and the fact that Liquidity Horizons differ from BAU systems to FRTB required models, in itself will need additional care to avoid duplication, wherever possible, to minimise operational costs. Moving both the Current Market Risk BAU systems and implementing the required Sensitivity Based measures will necessitate significant additional investment

• All models need regulatory approval

For business which requires an internal model, you must obtain, receive and maintain regulatory approval for the model. Any internal calculation model will need an approval from your regulator at a desk level. Approval maintenance will be based on three criteria – P&L attribution, back testing and a model independent assessment. Determining which desks rely on an internal model and which on the standard approach has timing constraints which prevent switching desks rapidly from one to the other.

• The success of the output of your FRTB platform needs desk level monitoring

Each desk will need to be monitored to observe price inputs for the last quoted price, last traded date, and adequacy of inputs, and should be alerted in real-time to avoid punitive capital charges

• The availability of data has a direct impact on your FRTB output

The underlying market data must be updated daily to avoid the above exceptions – any portfolio which fails under an internal model due to poor data will fall back to the standard approach and a potentially higher capital charge.

• Expanded Computing Power Requirements

The amount of computation needed to create the output is dramatically greater than for market risk including Expected Shortfall as the main exposure measure whilst maintaining VaR for back-testing, stress testing and default risk (under IMA). These risk treatments will mean there will be significant increases in computational processing power, aggregation and reporting which current systems will not easily be able to accommodate

The Black Ball: Fundamental problems that banks will face.

FRTB, which some banks have started to implement already, will be mandatory from January 2019, but as the name suggests will need a fundamental approach to deciding how best to structure each bank’s trading activities, how to maintain business as usual risk and regulatory governance processes and how to implement the required additional computational components and the ongoing maintenance of compliance under the new regime.

The requirements of FRTB mean that existing Basel market risk rules will be substantially altered. Your existing infrastructure will need updating or augmenting to deliver an effective and efficient solution, particularly in the following areas:

• Controls between the boundaries of the banking book and trading book.

The bank will need to deploy and support additional controls, documentation and regulatory audit capabilities, possibly utilising a scenario analysis framework to identify the optimum treatment for all trade types noting any prescribed treatments

• Including Expected Shortfall (ES) as the main exposure measure whilst maintaining VaR for back-testing, stress testing and default risk (under IMA).

Any solution will have to be able to handle existing VaR related calculations as well as the prescribed Expected Shortfall method. Handling both will mean there will be significant increases in computational processing power, aggregation and reporting which current systems will not easily be able to accommodate. By augmenting existing systems through deployment of an FRTB specific module, the bank will be quickly able to undertake all necessary computational requirements without changing BAU processes.

• Strengthening the relationship between the standardised (SA) and internal models approach (IMA).

The SA is mandatory and will be the ‘fall back’ capital charge. IMA has to be tested and maintained daily with full auditability at desk rather than firm level. The FRTB solution will have to be able to handle the new time frame deadlines, accuracy and transparency needed to support the increased number and type of risk/capital charge aggregations, P&L attribution tests, stress tests, back-tests, daily reports, new procedures, new organisational structures and increased Board level involvement

• Obtaining desk by desk approvals for Internal Models.

Every bank will be doing same things, potentially; getting the approvals in the first place will require substantial effort, documentation and empirical evidence to support each application. By deploying an FRTB specific solution that runs alongside existing BAU Market Risk systems, the bank will be able to test models, support validation testing, provide auditable evidence and run the 250 day tests with real data sets. Once approved the models will need to be tested and maintained. This adds a huge amount of administration and desk level scrutiny. The FRTB solution should be able to provide the capacity to support each desk and specified users to test and maintain control over their models, PnL Attribution tests and back testing. Should a model fail then the bank will need to re-apply; this will need to either be avoided through stress testing or managed through regular reporting and model validations. Again an additional requirement not easily supported in existing BAU solutions.

The challenges FRTB poses can be further categorised as:

• Strategic – requiring high level insight in terms of Target Operating Models, optimisation of capital and business re-engineering
• Business – what risk and capital management treatments, processes, procedures, control, reporting and organisational structure should the bank be deploying to facilitate and maintain compliance
• Technology – how to augment existing ecosystems without impacting “business as usual” so that the firm can support the preparations for FRTB compliance, the migration from current to future steady state and then the ongoing maintenance under a robust, high performing, transparent and efficient technical environment. (See Figure 2)

Timing & Implementation

The deadline to comply with FRTB is January 1st 2019, leaving approximately two years to plan the development, testing, approval and deployment of your solution. The high level time plan looks like this. (See Table 1).

A typical implementation timeline for a bank would look like this, with assumptions about national rule making and the timing of model approvals being specific to each firm. (See Figure 3).

The latest date your firm can complete implementation is December 31st 2019 – but – every other firm will be making submissions for necessary approvals, and some public commentators have said that from mid-2018 onwards this has the potential to become a bottleneck for firms who don’t get going early.

Programme Overview

Your implementation programme has a typical framework, but for something with this level of impact, starting early and appointing a strong leadership team is essential. Steps in a typical FRTB programme could include:

1. Mobilisation of an FTRB team with an accountable executive to lead your implementation across multiple departments/disciplines (front-office, risk, treasury, finance, operations, IT).
2. Discovery and current state impact analysis of FRTB on your organisation
3. Allocating budget for the programme, from now until compliance is reached.
4. High level requirements for people, software and hardware. Whilst the calculations will require systems, the governance and allocation of roles and responsibilities is also expected to be non-trivial and will need to be carefully thought through.
5. Platform qualification and selection (which may revisit the suitability of existing in-house systems), examining what assets are available to build an FRTB solution upon
6. Prototyping using existing feeds and models, and gap analysis for future requirements. Making an early assessment of the impacts of FRTB and discovering the missing pieces for the full implementation.
7. Organisational changes, policy development, systems implementation. Preparing for the 250 test period.
8. Testing, calibration, benchmarking, initial results – a precursor to the 250 test period, which can’t start unless the systems and people are already signed off.
9. Back-testing and P&L attribution testing for 250 days – required by the regulation to achieve approval.
10. Model approvals, regulatory approvals, parallel running.
11. Production, monitoring, and permanent regular reviews.

Choosing an FRTB Platform

Razor Risk has made significant investments and advancements to create a platform highly focussed on enabling firms to comply with FRTB. Here we summarise the key platform requirements to enable you to evaluate your current platform versus the software available from Razor Risk.

We can work with you to produce a report showing a gap analysis between the requirements of FRTB and your existing platform, please get in touch if you need any additional information or would like to talk to us about a system assessment. (See Table 2)

Next Steps

1. If you haven’t already, create a regular communication process to track the state of FRTB and national rule-making.
2. Use the platform checklist above to benchmark your current systems suitability to support FRTB.
3. Appoint an accountable Executive, to prepare and communicate the strategy for your firm.
4. Appoint a leadership group to receive, review and own your FRTB strategy.
5. Talk to Razor Risk and find out how we can help, visit www.razor-risk.com for more information
6. Be aware of the model approval bottleneck – don’t be at the back of the queue for regulatory input.