KNect365 is part of the Knowledge and Networking Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 3099067.


FRTB: the case for dynamic capital assessments

The Fundamental Review of the Trading Book (FRTB) will introduce unprecedented volatility to banks’ market risk capital quotas. This diminishes the value of traditional point-in-time quantitative impact studies (QIS) as a gauge of firms’ capital demands – as these can only provide snapshot calculations based on assumptions, which will vary in appropriateness as market conditions and model accuracy fluctuates.

Assessing desk- and firm-level requirements through these static lenses would undermine capital planning initiatives because they cannot account for the ebbing and flowing of capital consumption over time, nor permit the range of scenario testing necessary to inform the most efficient FRTB internal models approach (IMA) configuration and implementation strategy. One FRTB lead at a European bank says that “dynamic scenario analysis” would be the bank’s preferred method of estimating its capital needs under the framework.

Interactive capital study frameworks

Thus, there is demand for a smarter analytical tool – namely an interactive capital study (ICS) framework capable of charting FRTB constraints on a dynamic basis. Through an ICS, banks can gain a comprehensive view of their resource requirements that considers the effects of assumption changes, varied data inputs, different desk configurations and internal model failures. It can also adapt to various iterations of the regulation, allowing firms to see how their capital demands would be affected by different calibrations of the regime. This is especially useful as no-one can guess how FRTB will be implemented across jurisdictions.

ICS frameworks equip firms with the means to restrain their costs and minimise the risks associated with their FRTB change programmes – saving time, money and effort.

“Many firms have a target risk architecture for FRTB, but to achieve this they are using a static QIS – which locks them into making decisions now that they can’t validate,” says Andrew Aziz, global head of financial risk analytics at IHS Markit. “We think it is suboptimal to do a static impact study based on imperfect data, then have to wait until implementation to know if it was the correct decision. It’s much better to do dynamic assessments, considering several scenarios without becoming locked into a decision – particularly as certain aspects of FRTB may change,” he adds.

A dynamic analysis is indispensable, considering the fragility of permissions surrounding the use of internal models. While firms have the option of applying an IMA to their trading desks – which should produce smaller market risk capital increases than those associated with the new standardised approach (SA) – the conditions under which this approach can be deployed are limited.

Model robustness is challenged on an ongoing basis through a new profit-and-loss attribution (PLA) test and value-at-risk backtesting. The failure of too many tests within a rolling 12-month period will force a desk off of the IMA and onto the more capital-intensive SA.

Many firms have a target risk architecture for FRTB, but to achieve this they are using a static QIS – which locks them into making decisions now that they can’t validate.

Even desks that clear these hurdles may attract punitive capital add-ons in connection with non-modellable risk factors (NMRFs). Indeed, a 2016 industry study demonstrated that these could account for 30% of total market risk capital for IMA banks.

The ability to assess how the balance of IMA and SA desks will shift over multiple time horizons – as certain portfolios move in and out of model eligibility – will allow firms to better judge where to focus their modelling efforts, and save resources squandered on desks unlikely to retain model eligibility over time.

Similarly, an ICS framework would offer an insight into model risk factors that are vulnerable to an NMRF downgrade – assisting decision-making related to risk factor proxying, which can markedly reduce the capital add-on burden. It could also take firms even further by providing a window into how certain modellable risk factors can degrade over time, allowing early warning of when add-ons will increase, and helping to visualise seasonality effects. This intelligence could in turn help firms fine-tune their FRTB calibration – by, for example, identifying critical risk factors for which model approval should be a priority, or earmarking desks where the NMRF burden is such that reversion to the SAwould be more capital-efficient.

“NMRFs are a big wild card, and you have to plan for different scenarios. Some businesses may survive and some may not – some may be borderline. You need flexibility to keep checking the viability of your plans,” says a head of analytics at a large North American bank.

Complementing existing infrastructure

Banks are hungry for these capabilities. But with implementation projects already in train, and calculation engines in the front and middle office already undergoing convergence in response to the demands of the PLA test, appetite for uprooting existing infrastructure is lacking.

IHS Markit is therefore promoting a solution that can leverage in-house calculation engines or vendor packages to produce an ICS.

“Banks have little appetite for expensive or risky projects. What we provide with an ICS leverages the sensitivities already used for front-office risk management, and can provide end-to-end workflow capabilities. These include everything from real-price observations – such as committed quotes, modellability assessments and NMRF proxying – to capital impact across IMA and SA. While we do have a risk engine as part of our offering, we are really focused on providing modular components that can be easily embedded within banks’ existing ecosystems,” says Aziz.

The industry will therefore converge around common methods of assessing desk modellability. This increases the added value of leveraging software-as-a-service type solutions to perform these assessments, as they can liberate resources in-house to work on those risk management tasks that actually sharpen a bank’s competitive edge.

“What we’ve done with our service is build an application programming interface layer that allows firms to be completely proprietary with how they develop risk factor proxies and generate scenarios, but at the same time is largely turnkey. That combination of features doesn’t exist in bank infrastructure today – which is why IMA is becoming too expensive for many banks,” says Paul Jones, global head of FRTB solutions at IHS Markit.

This article originally appeared on on 12th September 2017 as an IHS Markit sponsored article.

Read the full collection of FRTB thought leadership articles from FTRB here >>

Get articles like this by email