Basel considered axing standardised approach to CVA calculation

Committee discussed axing standardised and basic approaches in recent months, sources say – but ruled out both

Red light: Basel decided against forcing dealers to use the heavily simplified basic approach

Photo: Ulrich Roth

Global policymakers have strongly considered axing the standardised approach to calculating credit valuation adjustments (CVA) in recent months, three people familiar with the matter tell Risk.net. Coming six months after banks were told they would no longer be permitted to use an internal models-based approach to calculate CVA, the move would have been a hammer blow, forcing dealers to use the heavily simplified basic approach, which some estimate to be 10 times more punitive.

Though the standardised approach is now expected to be included in the final CVA framework when it is published by the Basel Committee on Banking Supervision in December, it is understood there are still elements within the committee that would like to see it dropped, arguing the approach is still too complex for less sophisticated banks to compute.

In a further twist, the committee is also understood to have debated axing the basic approach. Some attribute this to committee members swinging in favour of greater co-ordination with the Fundamental review of the trading book (FRTB), of which the revised CVA framework is an offshoot. Under FRTB, all banks will have to implement the sensitivities-based approach to the revised market risk framework anyway.

A source at one national regulator says he believes both approaches are safe for now, but that the committee's thinking remains unclear: "The final decision will be made by the BCBS at the end of November, and the final CVA framework should be published in December. I do not think either of the two approaches is in danger now, but with the BCBS you never know."

The source adds that while no formal proposal was made to eliminate the standardised approach, there was a push by some regulators in this direction. A proposal was tabled to do away with the basic approach, however, but it was met with resistance.

The committee released the first consultative document on its review of the CVA risk rules in July 2015. The proposals were split into three broad methods: a highly punitive basic approach for firms with limited modelling capabilities (BA-CVA); the more sophisticated standardised method (SA-CVA); and the internal model-based approach (IMA-CVA).

I do think a fallback is required for banks unable to calculate SA-CVA. It is not straightforward, given the risk sensitivity requirements
CVA trader at a large European bank.

But in a move that shocked dealers, the committee decided to remove the own-models approach in March this year, citing concerns over whether CVA can be adequately captured by internal models.

In a speech delivered at a panel discussion in Washington, DC on October 7, Bill Coen, secretary-general of the Basel Committee, said the body was working on finalising the treatment of CVA risk and was "carefully weighing the benefits of a risk-sensitive treatment for this risk with the associated complexity and global applicability" – a line some interpreted as a veiled indication that SA-CVA was under scrutiny.

Coen also highlighted that CVA risk only accounts for 2% of dealers' minimum capital requirements on average, and is significant for only a relatively small number of banks.

If the standardised approach ends up being eliminated, banks will be forced to use the basic method. BA-CVA does not allow banks to recognise the benefit of any market risk hedges, and restricts hedging to single-name credit default swaps (CDS), single-name contingent CDS and index CDS.

One regional bank has estimated the capital requirement under the approach to be 10 times more punitive than the current standardised rules. Based on the latest impact study carried out in April, one large European dealer found the standardised approach required two to three times more capital than the internal models approach.

Destroying competition

When the Basel Committee removed the internal models approach from the CVA framework on March 24, banks raised concerns the move would destroy competition and discourage genuine hedging of risks. A further move to eliminate the more punitive standardised approach would be a huge mistake, many warn.

"Removing the IMA-CVA was clearly a step in the wrong direction, and further removing SA-CVA makes even less sense. I see a large potential for CVA risks being mismanaged due to these shadow-costs of a misaligned capital set-up," says Nicki Rasmussen, chief analyst in the counterparty credit and funding trading desk at Danske Bank. "These costs materialise through poor incentives to model and monitor the real risks, and create a false sense of security in the market risk management."

Removing the basic approach, on the other hand, might not impact larger banks that have more sophisticated risk management systems – but small banks that do not calculate risk sensitivities in the manner required by the proposals could have trouble complying.

"I do think a fallback is required for banks unable to calculate SA-CVA," says one CVA trader at a large European bank. "It is not straightforward, given the risk sensitivity requirements – smaller banks may struggle with the calculations, although there will probably be vendor solutions ready to step into the breach."

One CVA quant at a regional bank counters that smaller banks will likely be required to make SBA calculations under the FRTB anyway, however.

"Under the FRTB itself, you only have a choice between the internal models approach and the SBA, so all the small banks would be forced to implement the SBA anyway. So if you take that view, then it's perhaps not that much of a stretch to say you are going to push banks to do SBA and not provide an IMA-equivalent for CVA," he says.