Basel Committee to consult on scrapping op risk modelling

Regulators are set to propose a single simple method for calculating operational risk capital

On the scrapheap: the AMA may soon be obsolete

Global banking regulators plan to publish a consultation in December on disposing of operational risk modelling, in a surprise response to frequent criticisms of the decade-old advanced measurement approach (AMA).

The Basel Committee on Banking Supervision will also seek views on a new standardised approach, designed to replace all existing and proposed methods of calculating regulatory capital for operational hazards, according to Bill Coen, the committee's secretary general.

"There are always two sides of a debate. But when it comes to the advanced measurement approaches for operational risk, the views largely converge in the same direction – that the AMA has not worked as intended," Coen says in an exclusive interview with Risk.net. "When we consult by the end of the year on a revised standardised approach, I expect we will also propose removing the advanced modelled approach from the regulatory framework."

The future of the AMA has been the subject of intense discussion among regulators and industry figures since the Basel Committee suggested in November 2014 that the method might need to be simplified. The modelling approach has come under fire for being too complex and inconsistently applied and therefore producing wildly divergent capital levels.

Regulators have floated a number of possible amendments to the AMA, such as a lower confidence interval and better use of scenario analysis, and have hinted that the method may even be scrapped – but the latter option has been viewed by operational risk professionals and people close to the Basel Committee as the less likely outcome.

The new method the international supervisors plan to put forward in December will be called the standardised measurement approach (SMA). It is meant to replace the techniques currently in use – the AMA and the basic indicator and standardised approaches – as well as the revised standardised approach (RSA) proposed by the Basel Committee in October 2014, Coen says.

He adds that the SMA is an improvement on current non-modelling methods and has been designed to suit all banks, irrespective of their size and risk profile. "It's a simple approach, but also much more risk-sensitive approach than what we've proposed and what's currently in place." He declined to provide further details, saying it is "still a work in progress".

The SMA will borrow from the AMA by incorporating a requirement for banks to collect data on operational risk losses, but internal modelling will not be factored in when calculating regulatory capital. While the usefulness of the AMA as a way of determining capital levels has been questioned, the industry and supervisors agree that the loss data aggregation and modelling of risks that the method involves improve risk management. 

The RSA – the Basel Committee's previous attempt at reforming simpler approaches – was aimed at providing a more accurate reflection of a bank's risk level than the basic indicator and standardised approaches, by using a new proxy called the 'business indicator' in place of gross income. Gross income has been widely criticised as a risk-insensitive measure, notably by the UK's Prudential Regulation Authority, which wrote in July: "During the recent economic downturn, incomes dropped but operational risk exposures, in many cases, remained the same or increased." However, the revised approach drew overwhelmingly negative responses from the industry.

It is unclear whether the SMA will rely on gross income or the RSA's business indicator – which adds up three measures of business activity linked to income – or on another proxy entirely.

The Basel Committee intends to publish the new consultation on operational risk capital after its next meeting, scheduled for December 1–2, and before the end of the year, Coen says.

Click here to read Risk.net's in-depth analysis of the consequences of scrapping AMA.