On 31 March 2014 the Basel Committee on Banking Supervision (BCBS) published a final paper mandating a new standardised (non-internal models) calculation of counterparty exposure at default (EAD) that feeds into the overall Basel capital adequacy framework. This gives regard to the feedback received from respondents (including SunGard) to the BCBS consultative paper The Non-internal model method [NIMM] for capitalising counterparty credit risk exposures, as well as the results of a quantitative impact study.
The newly proposed 'Standardised Approach for Counterparty Credit Risk’ (SA-CCR) intends to address some of the long-standing criticisms of the current ‘simple’ methods - the Current Exposure Method (CEM) and Standardised Method (SM) - and will replace these old measures with what is meant to be a more risk sensitive measure.
Compared to the consultative paper, the revised standard provides some welcome improvements and clarifications.
- Transactions with a remaining maturity of less than one year will have their exposure scaled down via a square root of time factor. This will provide welcome relief, notably for Foreign Exchange transactions.
- Interest rate and credit derivative transactions will have their exposure weighted by a supervisory duration parameter reflecting the time period of the underlying risk factor.
- The calculation for certain complex or multi-factor instruments has been clarified.
- The delta adjustment for option transactions has been refined (from the initially proposed 0.5) to a type of Black Scholes formula, using supervisory volatility parameters.
- Some supervisory weighting factors have been revised. Notably, all credit derivative factors have been doubled, the FX factor has been reduced from 5% to 4%, and non-Electricity commodity factors have been increased from 15% to 18%.
- The exposure of a margined netting set is now capped at the exposure of the same netting set calculated on an unmargined basis. This addresses the previous criticism that a large Threshold amount (greater than the unmargined exposure of the underlying transactions) would have counted as credit exposure.
- Basis instruments are placed in separate hedging sets, based pairs of risk factors. Their supervisory factor is multiplied by one-half.
- Volatility instruments are placed in separate hedging sets. Their supervisory factor is multiplied by a factor of five.
It is satisfying that many of the points raised in SunGard’s formal response to the consultative paper were addressed by the Basel Committee.
Despite some welcome concessions (notably the lower weighting of short term transactions), it is yet unclear whether the new methodology will provide any significant capital relief compared to the CEM. Much will depend on the composition of counterparty portfolios.
For directional portfolios (such as those with corporate end-user counterparties), the new methodology will be quite punitive in terms of capital. For example, an unsecured 5-year at-the-money Interest Rate Swap will consume six times more capital under the SA-CCR compared to the CEM. A 9-month FX Forward will consume almost five times more capital (a 6-month deal would be 4 times more). On the other hand, portfolios with offsetting transactions will have a fairer treatment under the SA-CCR, but only within the narrow confines of ‘hedging sets’. In the two examples above, if the portfolios had completely offsetting opposite (long/short) deals, the CEM would still show positive exposure, whereas the SA-CCR would correctly compute a zero exposure.
We expect that margined portfolios, notably those with central counterparties, will benefit from a lower capital treatment under the SA-CCR. The weighting by the Margin Period of Risk will effectively deliver a reduction of approximately 50% in the amount of capital required for such transactions, compared to the CEM.
Despite improvements in the risk-sensitivity of the standardised approach, it remains a crude, conservative, one-size-fits-all approach which will continue to significantly overstate the true exposure on most counterparty portfolios. For example, supervisory add-on and correlation factors are flat percentages applicable to broad asset classes, with no recognition of individual asset volatilities and correlations: all FX transactions will be subject to a flat 4% weighting (with no regards to the actual volatility of a given currency pair); a flat 50% correlation is assumed for all single-name credit derivatives; no correlation is recognised between interest rates in different currencies.
It is also perplexing that the weighting by (the square root of) residual maturity only applies to transactions of less than 1 year. Whilst longer term interest rate and credit derivative transactions will be weighted by the duration of the underlying risk factor, long-term FX, equities and commodities transactions will have exactly the same weight as 1-year transactions.
We believe that best-practice measurement of counterparty exposures is to use a Monte Carlo simulation of potential future exposure, which provides a true reflection of all portfolio effects in a scenario-consistent manner, taking into account market-based volatilities and correlations. Such a method is available to banks under the regulatory ‘Internal Model Method’ (IMM), which requires specific approval from national supervisors. It is expected that an IMM approach to counterparty risk will still provide significant capital savings compared to the SA-CCR.
The implementation of the new measure will raise significant system challenges for banks worldwide, as it is based on many more parameters than the previous methods. Some banks may even (re-)consider whether they should apply for an IMM-waiver. SunGard is able to help its customers in this respect, as it has already implemented the new SA-CCR measure in its Adaptiv Credit Risk solution (which will be adjusted to reflect the latest changes), and can also provide an IMM-compliant Potential Future Exposure calculation via its Adaptiv Analytics engine.