As Risk went to press, the Basel Committee on Banking Supervision sprang a surprise. Rolled up in a long-awaited consultation on the calculation of capital for bank loan books was the announcement that the committee had decided to end the use of internal models for a different capital requirement – the credit valuation adjustment (CVA) charge that covers variation in the market price of counterparty risk over the life of a derivatives trade.
The surprise was less in the decision itself – it comes hot on the heels of proposals to end the use of models for operational risk capital, and rules that will more strictly police market risk modelling – and more in the timing. A consultation on CVA capital had been launched last July, and the industry was given instructions only last month for a new quantitative impact study (QIS), in which the internal models approach (IMA) was included.
So, it was a little abrupt. It also appears to be final – not a proposal, not a placeholder. IMA-CVA is officially dead.
The immediate reaction of dealers contacted by Risk was surprise and disappointment. They pointed out that the regulatory justification for the move – greater use of clearing and bilateral margining will cut counterparty exposure – does not apply to a good-sized chunk of the client base. The vast majority of corporates and sovereigns are not subject to either clearing or margin regimes.
So, the capital impact for dealers, and the pricing impact for exposure-heavy customers, now depends on the results of the ongoing QIS and the calibration of the two remaining approaches – basic and standardised – that will replace the IMA. Under the July consultation, regulators opened the door to a host of changes the industry had been seeking since the CVA charge was first mooted: capital relief will be granted for a wider range of CVA hedges, and exposure will consider not just credit spreads but other sensitivities as well, such as foreign exchange movements or interest rates.
The bottom line is that some sensitivity will be lost with the death of IMA-CVA – a principle dealers value more highly than regulators – but that variability in numbers should also be cut, which should matter on both sides of the fence. Ultimately, it may be the case that unhedged, non-cleared trades see a significant jump in capital, while hedged, non-cleared trades do not.
Banks are not backwards in coming forwards, of course, and in response to last year's QIS, one trader was already speaking in the darkest possible terms about the prospect of losing the IMA: "If we have to switch to [the standardised approach], we would not be doing the derivatives business any more."