Proposed S&P counterparty criteria place too much emphasis on concentration and diversification at the cost of flexibility, and may do little to help risk diverisifcation, according to some market participants. Whether it will go forward as proposed is still unclear, but should it do so, it could have more effect of the rating agency than on the covered bond market. Denise Bedell reports.
To view the digital version of this report, please click here.
To use or not to use, that is the question on the minds of many covered bond market participants pondering the activities of Standard & Poor’s, as they await final details of its new covered bond counterparty criteria.
The ratings agency is currently digesting market feedback on the changes, the most significant of which are credit enhancements linked to the exposure of any single swap counterparty and also the number of counterparties included in swap arrangements.
Although most people engaged in the covered bond market acknowledge the importance of increasing the link between ratings and counterparty exposure, some feel the proposed criteria go too far, making the process inflexible and failing to acknowledge the diverse nature of covered bond programmes in different locales.
Whether the criteria will be adopted as is or will take into account market views on the issues remains to be seen. The result, however, may have more of an impact on the ratings agency’s position within the covered bond market than on the market itself.
In December, S&P published its new counterparty criteria, which at the time it intended to extend both to securitisation and to covered bonds. However, after something of an uproar from the market and following much lobbying from covered bond market groups – such as the European Covered Bond Council – in January the rating agency announced that covered bonds were outside the scope of the new criteria. It resolved to issue new criteria specific to that market.
In March Standard & Poor’s published a new set of covered bond counterparty criteria and launched a consultation period and request for comment that went until May 4. Within the covered bond market in Europe, the ECBC coordinated responses and sent in comment.
The key issue of debate is the dual-requirement for low counterparty concentration and a high number of counterparties in derivatives linked to a transaction. As S&P noted in its proposal: “In assessing counterparty risk, the proposed covered bond criteria for derivative obligations consider three factors: The number of derivative counterparties in the programme, the weighted-average rating on the counterparties and the ICR on the covered bond issuer.” In order to get a Triple A rating, a transaction would need a highly rated covered bond issuer and numerous highly-rated counterparties, said S&P.
Not alone
Linking ratings more closely to issuer rating is to be expected, according to Heikko Langer, senior covered bond analyst at BNP Paribas. Further connecting covered bond programmes with the ratings of their issuers is a trend that has been developing for a while. “Not just with Standard & Poor’s, but also, for example, with Fitch - who put more emphasis on counterparty criteria,” he said. Fitch launched its own updated counterparty criteria in March this year.
All the major rating agencies are linking covered bond ratings closer to issuer ratings, which makes the rating closer to what is actually happening in the market, according to Langer: “In the last three years issuer risk and issuer quality were the keys to pricing. We saw a huge diversification of spreads on covered bond programs with the same rating that came from different issuers. So this is a bit of a catch up of rating methodology to market environment,” he said.
“By increasing the link we most likely will end up with a more diversified rating landscape – fewer Triple A ratings – which would be closer to market pricing spreads. Also, rating agencies are focusing more on liquidity risk in cover pools,” he added.
The big problem with S&P’s proposals, however, lie in the counterparty concentration and diversification requirements for the highest ratings. The S&P criteria outline a differentiation of covered bond programmes based on the number of counterparties and the involvement of each particular counterparty. Bucket one would include programmes with 11 or more unrelated swap counterparties and no more than a 30% single counterparty exposure. Bucket two would include programmes with either one to 10 counterparties, or a single counterparty exposure greater than 30%.
According to S&P, the 30% limit reflects the typical counterparty concentrations in rated covered bond programs. S&P explained: “Having a larger number of existing derivative obligations in place may give greater flexibility to the covered bond issuer to manage its exposures and replace counterparties more quickly.”
Concerns arise, however, over credit enhancement based on the number of counterparties. Although in theory it sounds like a good thing, within the world of covered bonds the 11-counterparty cut-off set by S&P for Bucket one could be more harmful than helpful. “They are encouraging you to diversify counterparties, but it is not very practical,” explained John Wong, partner at Allen & Overy.
In fact, of those market participants approached for this article, most pointed out that it would be difficult to even find 11 counterparties that would provide swaps for such a transaction. Assuming one did find that many counterparties, it would mean that many, if not most, transactions would be using the same set of counterparties – thus negating the risk diversification that the rule attempts to engender.
In addition, it makes no allowance based on the size of the covered bond programme. For large bank programmes with multi-billion-dollar transactions, it is practical to have a number of swap counterparties. But for a smaller institution launching just a few hundred million or billion dollar transaction, using more than a couple of counterparties does not make economic sense.
Neither do the criteria acknowledge the self hedging practice used under various covered bonds programmes, where the issuer or parent bank of the issuer tend to be the hedging counterparty to the covered bond programme. This is often seen in Scandinavian countries, and there are various operational and cost benefits to that.
Diverse models
The S&P counterparty criteria does not recognise that there are various covered bond models in use, according to Andrew Vickery, partner of capital markets in London at law firm Linklaters. “For example, a cover pool held by the issuer and isolation achieved through statutory ringfencing versus a cover pool isolated in an SPV – as seen in the UK and the Netherlands, respectively – are quite different,” he said. “This is yet again another one-size-fits-all approach.”
Such programmes follow a range of different structures based not only on the specific frameworks in whichever jurisdiction they fall under, but also based on the specific needs of the issuer. “Just to say there is a universal prescription based on these criteria doesn’t make any sense,” said Lawton Camp, partner at law firm Allen & Overy.
Programmes in Scandinavian countries, such as Norway and Sweden, for example, are quite differently structured to those in the UK or in Spain. One of the big draws of the covered bond structure is that it provides that flexibility. By adding such specific criteria for counterparty credit enhancement, issuers may have to choose between the rating and the flexibility.
As such, this model is not compatible with the diversity seen in covered bond programs, according to Julia Hoggett,head of EMEA FIG flow at Bank of America Merrill Lynch: “The challenge for S&P has been that the changes to their methodology have lead a number of issuers to feel that it is not appropriate to use them any more. S&P’s footprint in the covered bond market is changing as a result.”
Upon release of the criteria in January, a number of S&P-rated covered bond issuers requested their ratings be withdrawn, including German institutions WestLB and NordLB. Clearly if the criteria go ahead as is, others may be asking themselves whether a rating from Standard & Poor’s is worth the cost, time, and complexity.
