In a bid to boost the risk-sensitivity of banks’ counterparty risk calculations, the Basel Committee is gunning to revamp the existing exposure-model, triggering a familiar volley of criticisms from industry participants, who claim the proposed model is a one-size-fits-all approach with unintended negative consequences.
At present, large banks have two simple methods of determining counterparty risks – the current exposure method (CEM) and the standardized method (SM) – and one supervisory-approved model: the internal model method (IMM).
The non-internal model method (NIMM) has been proposed to replace the latter, which seeks to better incorporate stress-testing, collateralization and economic offsetting.
In June, the Basel Committee on Banking Supervision released a consultative paper with the deadline for industry feedback elapsing end-September.
According to Jean-Marc Schwob, credit risk expert at SunGard, the shift from IMM to NIMM represents a crude and conservative attempt that inflates counterparty exposures, which will trigger a “big increase in regulatory exposure levels for single transactions or non-diversified, unsecured portfolios”.
He adds: “On the other hand, perfectly offsetting transactions – and hence well-diversified portfolios, but only within the narrow concept of hedging sets – will have a more advantageous treatment, allowing for a total offset of add-ons rather than the 60% offset allowed under the CEM’s net-to-gross ratio formula.”
Essentially, Schwob charges, like RWA-calculations and the Basel framework, more generally, the NIMM is overly prescriptive and fails to take into account individual asset volatilities and correlations.
According to SunGard, which provides processing solutions to broker/dealer houses, the Basel approach to place trades in hedging sets – subsets of transactions within asset classes: interest rates, foreign exchange, credit, equities and commodities – is arbitrary and crude.
“For example, in the foreign exchange set, banks are only allowed to offset within a currency pair bucket – not between different currency pairs: for interest rates, only trades in the same currency can be offset – there is no recognition of correlation between interest rates in different currencies,” it states.
“The NIMM doesn’t take individual asset volatilities into account, either. Volatility is the key ingredient in potential future exposure calculations – the more volatile the underlying instruments are, the higher the potential future exposure.”
It continues: “Basel’s NIMM prescribes fixed weighting factors for each asset class, which are merely based on average volatilities. For example, all equities will have the same weighting factor of 32%.
“This may give rise to moral hazard, as flat supervisory factors effectively provide a regulatory incentive to trade highly volatile assets – by understating their supervisory factor – and a regulatory penalty on trading in low volatility assets, by overstating their factor.”
Underscoring regulatory distrust towards derivatives as an asset class, the framework also mandates short-term unsecured transactions to be subject to the same factors as one-year transactions, while margined transactions will be subject to much lower capital costs.
Meanwhile, way-in-the-money options compared with forward deals will be substantially privileged, given the fact option deals are set to be weighted at a flat 50% irrespective of whether they are in or out of the money, concludes SunGard, which prefers Monte Carlo simulations to calculate, in a realistic fashion, options pricing.
The other risk, of course, is that some banks might benefit from supervisory forbearance, using SM rather than the NIMM, a de facto form of regulatory arbitrage, though at present few banks use the SM approach.
Still, given the dilution of liquidity coverage ratio and policymakers’ increasing recognition of the pro-cyclical impact of bank regulation, industry participants are hopeful of greater flexibility from the Basel committee and a lengthy implementation timetable.
SunGard’s full critique can be found here.