Banks to ramp up credit risk if Basel scraps internal models

Without IRB, banks could resort to barbell portfolios of high- and low-risk assets

  • The Basel Committee wants to move large corporate exposures onto the standardised approach for calculating credit risk capital.
  • This is the culmination of concerns about the comparability of banks' credit risk-weight models.
  • Return on capital under the standardised approach significantly favours high yield over investment-grade credit.
  • Liquidity requirements also push banks to hold low-risk instruments, resulting in barbell portfolios with few median risk exposures.

Proposals to drastically limit banks' ability to use internal risk-based models to calculate credit risk capital will encourage lending to less creditworthy counterparties, trade associations have warned. This would have the perverse result of increasing the amount of credit risk on their balance sheets.

In its March 24 consultation, the Basel Committee on Banking Supervision proposed to strip away from banks the option to use the internal ratings based (IRB) approaches to determining their capital requirements for credit exposures facing other banks and large corporates with assets of more than €50 billion.

The advanced IRB relies on internal risk models developed by banks to calculate capital requirements based on their credit risk exposures. These models are validated by regulators but are rarely comparable from bank to bank, and can calculate wildly disparate capital requirements against relatively similar loan exposures.

To reduce variability and increase comparability, the Basel Committee has proposed that banks drop the advanced IRB and simply follow the less risk-sensitive standardised approach set out by the committee. If the rule is finalised, lobbying groups caution that with no capital incentive to hold safer but lower-yielding assets, banks will likely ramp up credit exposures with higher risk counterparties to achieve stronger returns on assets.

“Under the liquidity coverage ratio, banks have to hold a proportion of high-quality liquid assets like treasuries and cash. These low-yield assets are a drag on ROE [return on equity] under a flat capital measure like the leverage ratio. To optimise ROE, banks are encouraged to hold higher risk assets where they can make a high yield over that same flat number,” says Brad Carr, deputy director in regulatory affairs at the Institute of International Finance (IIF) in Washington DC.

“Banks become naturally encouraged to retain a barbell balance sheet where they hold assets at either end of the risk spectrum, but they have less capacity to invest in assets in the middle, such as prime mortgages and investment-grade credit. Using an internal models approach, capital requirements increase as yield goes up and there is a reasonably sensible trade off of risk-return across the full range of the spectrum,” Carr adds.

The perverse risk incentives created by the use of risk-insensitive models can be seen in the example of a $10 million corporate loan accounted for under three different risk models (see table). Under the proposed standardised approach, for example, an investment-grade loan receives a 75% risk-weight, just a one-quarter reduction from the full 100% weighting handed to non-investment grade loans.


The difference in the return on capital (ROC) between the loans is startlingly larger, however. Despite only being recognised under the standardised approach as 25% more risky, a B+ rated loan generates an ROC of 12.2%, almost three times the ROC a bank would achieve on an A+ loan. The incentive to extend credit to higher risk counterparties under the standardised approach is clear.

The situation is reversed under the advanced IRB approach, where the risk weight on the A+ loan at commencement is just 40.4%, generating a healthier ROC of 13.9%. Conversely, the B+ loan is calculated to have a risk weight of 203.9%, reducing the ROC to 7.2% and making the investment-grade loan the better use of capital.

“Major step backwards”

The proposal to scrap the advanced IRB and shift to the standardised approach for certain exposures has provoked alarm throughout the banking community. In an unusually strongly worded letter, the Securities Industry and Financial Markets Association (Sifma) wrote to the Basel Committee on April 14 that dropping internal models and replacing them with the standardised approach would “do severe damage” to the prospects for growth in the world economy.

“The paper does not mention the importance of maintaining a risk-sensitive capital framework. Our worry is that the consultation represents a retrograde conceptual shift in the design of the international framework. This is reinforced when considered alongside other proposals to revise the standardised approaches for the various risk categories and to introduce output capital floors. Relying on regulators’ assessments of risks, which will be rough approximations at best, to determine capital would be a major step backwards,” Sifma wrote.

What troubles us is that in this mix of where banks can still use the foundation and advanced IRB, their capital requirement can’t fall below 60% to 90% of the standardised calculation for that credit exposure
Regulatory specialist at a trade association

The Basel committee is proposing to allow banks to retain the use of the foundation IRB approach when calculating credit risk capital exposures facing corporates with assets of less than €50 billion, and revenue of more than €200 million. Banks could continue to use the advanced IRB approach for corporates with revenues below €200 million. But even in these circumstances, the committee is considering implementing an aggregate output floor calibrated in the range of 60% to 90% of the standardised approach to set a minimum baseline below which calculated capital cannot fall.

“What troubles us is that in this mix of where banks can still use the foundation and advanced IRB, their capital requirement can’t fall below 60% to 90% of the standardised calculation for that credit exposure. For example, the best risk weight you’ll ever get for a prime mortgage under the standardised approach is 25%, but if your IRB calculation is floored at 90% of that number, a 22.5% risk weight is the best you’ll be able to achieve. That loss of risk sensitivity is very troubling,” says one regulatory specialist at another trade association.

Bank trade bodies have also attacked the very basis of the consultation paper. The Basel Committee states the two primary goals of the revised framework are to improve comparability between credit risk models and address excessive variability in the capital requirements that the models produce.

Critics say these regulatory goals are self-serving. By limiting the use of the IRB approach, the committee is making the supervisory role of regulators easier while ignoring the fact that fostering risk weight comparability through the use of the standardised approach diminishes banks' ability to accurately model risk.

Deliberate opacity?

Former regulators have nonetheless defended the new proposals. They say banks have deliberately used the esoteric nature of their capital models to wrong-foot supervisors, and make it more difficult for them to understand the assumptions and inputs underpinning their capital requirement calculations.

“When I was at the Fed, there was only so much the banks were willing to share with regulators in terms of modelling. We were reliant on the disclosures that the banks made, and on the credit risk side we are talking about complex stochastic modelling exercises involving very subjective inputs. As a regulator, you have limited resources, so you may only have a few weeks to explore, understand and potentially try to validate those immense and complex black boxes. That was a significant challenge then, and I suspect it remains a significant challenge today,” says Andrew Huszar, senior fellow at Rutgers University Business School and a former examining officer at the Federal Reserve Bank of New York.

“During my time with the Fed, I think we were overmatched when trying to understand the models. Banks have an incentive in maintaining the opacity of their models because it works to their advantage. That may have finally led regulators to determine that they could not get a sufficient degree of confidence with what the banks are doing. This proposal seems to be a response to those difficulties. I think this is a very good proposal, and I wish the committee had come up with it sooner,” Huszar adds.

We understand there are different views on this issue within the Basel Committee. Some are very sceptical of modelling, whilst others are supportive of it, albeit with some caveats and constraints
Brad Carr, Institute of International Finance

Sources with knowledge of the Basel Committee’s deliberations report that the body was far from unanimous in its decision to put the proposal restricting the use of internal models out for public consultation. The committee reportedly spent eight hours discussing this single proposal during its December 2015 meeting, and a further four hours debating the measure when it met in March.

“We understand there are different views on this issue within the Basel Committee. Some are very sceptical of modelling, whilst others are supportive of it, albeit with some caveats and constraints. This consultation looks like a form of compromise in which modelling is preserved – with some extra constraints – for the retail and small to medium enterprise space, but with the sceptical view prevailing for the wholesale portfolios,” says the IIF’s Carr.

The consultation period runs until June 24.

Are we now in the Basel IV era?

Keen observers of the Basel Committee on Banking Supervision began to notice a curious but important shift in the work of the body in late 2015: the frequency with which the committee was issuing proposals for public consultation began speeding up.

Since November 2015, the committee has issued a flurry of 12 new consultation documents – six of which have been published in the past seven weeks. To put that in context, it only issued five consultation papers in the first 10 months of 2015 up to November, and only published 10 proposals in the whole of 2014.

What accounts for the sudden rush of new consultation papers in the past two months?

“This is Basel IV. I know [Bank of England governor] Mark Carney and the Basel Committee hate it when you call it that, but this is absolutely Basel IV. This process is proceeding in a completely different fashion to Basel III. It is a bunch of isolated, piecemeal changes with no underlying or compelling rationale as there was in Basel II and in Basel III,” says one US regulatory lawyer and former bank supervisor.

“There is an unspoken rationale in all this which is to undo an enormous amount of risk sensitivity that was inserted into the regime during Basel II and move towards a one-size-fits-all approach that doesn't recognise real risk distinctions between different firms, markets and jurisdictions. At least some of the people on the Basel Committee are willing to admit in quieter moments that the point here is to raise banks' effective capital requirements. But that is not something that the committee can come out and say publically as a matter of consensus,” he adds.

Governor Carney insisted during a December 1 press conference that Basel IV is not already underway but that the rash of new proposals issued by the committee in late 2015 merely represented improvements to Basel III. “There is no Basel IV. There is no big wave of additional capital. There are things that need to be done to clean up the system,” Carney said.

In the months since, however, consultation papers from the Basel committee have been proposing changes that amount to draconian increases in capital requirements. These include scrapping the advanced approaches for modeling operational risk capital and introducing strict limitations on the use of internal ratings based approaches for calculating credit risk capital requirements.

The result of both of these changes will be to force operational risk and credit risk managers to instead follow the less risk-sensitive standardised approaches in most instances – with the effect of markedly increasing the amount of capital banks have to hold.

“The industry is struggling to keep up with all these consultations, which are fundamentally changing the nature of the regulatory capital framework. I would guess we are seeing a proposal each month right now between step-in risk, operational risk modelling and restrictions on credit risk models. It’s a steady flow of proposals that are coming thick and fast. As an industry, we don’t have a holistic view of the direction the committee is running in,” says Hugh Carney vice-president of capital policy at the American Bankers Association in Washington, DC.

“The Basel Committee has taken a piecemeal approach, and each of these pieces appears to be developed by individual Basel task forces that are doing tweaks here and there. We are concerned that the committee is not collectively evaluating these changes and what the big picture impact is,” he adds.

Basel Committee public consultations: Jan 2014 – April 2016

Apr 21, 2016: Interest rate risk in the banking book final standards

Apr 14, 2016: Prudential treatment of problem assets – definitions of non-performing exposures and forbearance

Apr 6, 2016: Revisions to the Basel III leverage ratio framework

Mar 24, 2016: Reducing variation in credit risk-weighted assets – constraints on the use of internal model approaches

Mar 11, 2016: Pillar 3 disclosure requirements – consolidated and enhanced framework

Mar 4, 2016: Standardised measurement approach for operational risk

Jan 14, 2016: Minimum capital requirements for market risk final standards

Dec 17, 2015: Identification and measurement of step-in risk

Dec 10, 2015: Revisions to the Standardised approach for credit risk second consultative document

Nov 10, 2015: Capital treatment for "simple, transparent and comparable" securitisations

Nov 9, 2015: TLAC Holdings

Nov 5, 2015: Haircut floors for non-centrally cleared securities financing transactions

Jul 16, 2015: General guide to account opening

Jul 1, 2015: Review of the credit valuation adjustment (CVA) risk framework

Jun 8, 2015: Interest rate risk in the banking book consultative document

Feb 5, 2015: Developments in credit risk management across sectors: current practices and recommendations

Feb 2, 2015: Guidance on accounting for expected credit losses

Dec 22, 2014: Revisions to the standardised approach for credit risk consultative document

Dec 22, 2014: Capital floors: the design of a framework based on standardised approaches

Dec 19, 2014: Fundamental review of the trading book: outstanding issues

Dec 11, 2014: Criteria for identifying simple, transparent and comparable securitisations

Dec 9, 2014: Net stable funding ratio disclosure standards

Oct 10, 2014: Corporate governance principles for banks

Oct 6, 2014: Operational risk – Revisions to the simpler approaches

Jun 24, 2014: Review of the Pillar 3 disclosure requirements

Jan 23, 2014: Revised good practice principles for supervisory colleges

Jan 12, 2014: Basel III: the net stable funding ratio

Only users who have a paid subscription or are part of a corporate subscription are able to print or copy content.

To access these options, along with all other subscription benefits, please contact or view our subscription options here:

You are currently unable to copy this content. Please contact to find out more.

You need to sign in to use this feature. If you don’t have a account, please register for a trial.

Sign in
You are currently on corporate access.

To use this feature you will need an individual account. If you have one already please sign in.

Sign in.

Alternatively you can request an individual account here