The Basel Committee wants to remove internal models from the calculation of banks’ operational risk capital, but needs to find a suitable replacement.

What’s happening?

Stefan Ingves, the Swedish central bank governor and chairman of the Basel Committee on Banking Supervision (BCBS), confirmed long-standing speculation in November 2015 that the BCBS will soon propose removing the advanced measurement approach (AMA) to calculating operational risk capital. This would mean that all banks will calculate how much capital to hold against operational risks using a new standardised approach that the BCBS is currently drafting.

“Most would agree that the benefits of the AMA are not proportionate to the related costs and complexity,” said Mr Ingves.

Why is it happening?

There have been difficulties calculating risk-weighted assets (RWAs) based on the seven Basel operational risk categories – employment practices including workplace safety; business practices affecting clients or products; damage to physical assets; business disruption and systems failure; failures of execution, process and delivery; internal fraud and external fraud. Even the largest banks may only have a few hundred loss data points for operational risk, not necessarily covering all seven categories.

“RWAs are calculated at a 99.9% confidence level, but because datasets are insufficiently large, banks have seen disproportionate instability in their loss distribution approach model results in response to tail loss events,” says Eva Leighton, an independent consultant who was previously deputy global head of operational risk management at Morgan Stanley.

Recent misconduct fines at global banks for money laundering or market manipulation have sometimes exceeded total operational risk capital resources. Ms Leighton also notes that current Basel capital methodologies do not produce lower capital in response to a reduction in risk. 

“For credit or market risk, reducing risk by hedging or unwinding positions promptly reduces capital. For operational risk, managers can invest in process or control improvements to reduce operational risk, but it will take at least several years for the resulting decrease in losses to flow through their loss distribution approach models and result in lower capital,” says Ms Leighton.

What’s the risk?

The addition of operational risk to the Basel capital framework under Basel II directly contributed to the creation of the operational risk data exchange association (ORX) in 2002. ORX is designed to provide a secure and anonymised platform to collect historic external operational loss data, which is one of four inputs for AMA capital models.

Reg rage - denial

Marc Taymans, a managing partner at consultancy Risk Dynamics and visiting professor of operational risk management at France’s IESEG school of management, says he had anticipated a refinement of the AMA with more guidance to ensure greater comparability between banks. He is concerned that outright cancellation will not achieve desirable results from a regulatory viewpoint.

“Operational risk models are still in their infancy, and it is Basel and bank supervisors that have driven greater rigour on data, modelling and testing. There needs to be a very careful impact assessment of removing AMA, because without AMA you would not have ORX, and you would not have such a developed set of internal loss data,” says Mr Taymans.

Can we live without it?

Luke Carrivick, head of research and analytics at ORX, believes the momentum for continued improvements in operational risk loss data and analysis will continue. ORX now has 85 members, of which only about 50 use the AMA. Many of the other members had no plans to move to AMA even before the announcement by Mr Ingves.

“Although ORX was set up to collect loss data, our purpose has now moved well beyond that. Non-AMA banks use our data for their own economic capital models and for benchmarking, and we are running a whole range of projects on risk measurement and management that are not necessarily for capital purposes,” says Mr Carrivick.

Monsur Hussain, a senior director in the financial institutions team at ratings agency Fitch, is optimistic on the potential for shifting to a purely standardised approach to operational risk capital. He expects a new standardised measurement approach to borrow from the current AMA by incorporating a requirement for banks to continue to collect operational risk loss data.

To improve forecasting based on historic loss data, banks are keen for more descriptive information on the linkages between operational losses and specific exposures. This is where regulators can directly influence and improve risk management, and a number of countries such as South Korea and Italy already have detailed public databases. ORX members also requested in 2012 the launch of a scenarios programme, designed to map and share anonymised live portfolios to see how risks move across different business lines.

“For each scenario category, we try to find the risk drivers including the external environment. For instance, for mis-selling we could look at media coverage as a potential driver for customers to seek restitution,” says Caitlin Frost, the head of the ORX scenarios programme who was formerly in the operational risk capital team at Lloyds Banking Group in the UK.


All fields are mandatory

The Banker is a service from the Financial Times. The Financial Times Ltd takes your privacy seriously.

Choose how you want us to contact you.

Invites and Offers from The Banker

Receive exclusive personalised event invitations, carefully curated offers and promotions from The Banker

For more information about how we use your data, please refer to our privacy and cookie policies.

Terms and conditions

Request a demonstration to The Banker Database

Join our community

The Banker on Twitter