Issuer: SBC Glacier Finance Series 1997-2
Amount: $1.6 billion
Issue type: repackaging of credit-linked notes
Launched: September 9
Bookrunner: SBC Warburg Dillon Read
SBC Warburg Dillon Read brought a new degree of sophistication to the art of credit risk and bank capital management with the $1.6 billion securitization deal it announced in September.
SBC writes a series of credit-linked notes to hedge the credit risks of specific counterparties taken on in a range of its international activities, including mainly lending but also bonds and derivatives trading. These notes are sold to a special purpose vehicle (SPV), which in turn uses them to collateralize $1.6 billion of floating-rate bonds issued in two equal tranches of seven and nine years.
The deal follows the bank’s decision to create one division to manage its credit risk. This merges liquid and illiquid exposures from SBC’s loan book, securities trading and derivatives businesses. The deal helps SBC “strip off credit risk from the original transaction where SBC has an exposure to an obligor, then bundle and sell it in a unique way to investors”, says Linda Bammann, managing director at SBC Warburg Dillon Read in New York.
Credit-linked notes are reasonably straightforward credit derivatives. The buyer, in this case the SPV Glacier Finance, receives a spread over Libor for taking credit risk on a specific counterparty. In the event of the original borrower defaulting, SBC pays a principal amount to the SPV derived from the mark-to-market rate for a liquid reference security.
Say the underlying borrower is an emerging-market sovereign whose most actively traded bonds fall to 40% of face value following a default, the SPV receives a payment of 40% of the notional principal of the credit-linked note. Where there is no publicly traded reference security for the credit-linked notes, SBC will offer a fixed payment of 51% based on Moody’s analysis of average historical recovery rates for defaulted bonds.
The process does not quite take the loans off the bank’s balance sheet. SBC is not selling the loans. But it does obtain a release of capital because regulators and accounting bodies recognize the credit-linked notes as cash-collateralized exposures. By selling them to the SPV it has raised cash up-front, which flows through from bond investors. The objective of the transaction is to help the bank manage its capital more efficiently. Bond investors are attracted by the spread over Libor (16 basis points on the seven-year, 19bp on the nine-year Aa1/AA+ rated bonds) for a diversified portfolio of credit exposures. The rating is capped by SBC’s own rating. It would have been much harder for SBC to raise $1.6 billion by selling individual credit-linked notes to individual investors, and much less useful.
The subtlety of the deal is that the pool of underlying collateral – the credit-linked notes – may constantly be altered at SBC’s discretion. The notes are callable at par on every quarterly interest payment date and the SPV then reinvests the proceeds in new credit-linked notes. The structure reflects the fact that SBC’s credit exposure will be ever changing. Three months after selling a credit-linked note, the underlying borrower may have refinanced its bank loan from SBC, or the derivatives markets may have moved in such a way that SBC is no longer in the money on positions with that counterparty. The old credit-linked note will no longer match the bank’s actual exposure, so it will want to rehedge.
The cynical investor might ask what’s to stop SBC loading the collateral pool with all manner of toxic waste?
The rating agencies devised a model for the collateral pool and any new credit-linked notes going into this pool must fit within the guidelines the model draws for average credit rating, issuer, industry and geographic concentration. There is also an element of the protection used in more standard securitization, namely subordination. Subordinated tranches equivalent to 8.25% of the whole bond were placed separately from the main deal. SBC also retains a small sliver of each maturity portion of the securitized bond deal. It calls this the e-tranche, and deems it the bank’s capital investment in the deal. It’s a form of equity participation, and will absorb the first losses, followed by holders of subordinate bonds, before senior bondholders suffer a hit. Additional protection comes from a degree of over-collateralization.
But the the main source of comfort for investors in the bonds is Standard & Poor’s modelling. The rating agencies face hefty demands from many banks creating collateralized loan obligations, following the success of NatWest’s Rose deal in October 1996. SBC Warburg Dillon Read’s deal brought additional challenges. The initial collateral pool contains notes on some 120 different obligors. These may drop out and be replaced by others. The pool includes some credits that have never been rated by the rating agencies.
SBC wrote to all the obligors which might be included in the deal, informing them that they might be in it and providing a telephone number for borrowers to call to get additional information. The bank was keenly aware of the disquiet of some NatWest customers at the prospect of the British bank securitizing their loans off its balance sheet and the possible implication that it might be concerned about their credit quality. SBC did not want to disrupt its customer relationships. In the event, the response of SBC’s obligors was muted. “Borrowers shouldn’t really be concerned,” says Bammann, “because their presence in the collateral pool does not reflect SBC’s concern about their creditworthiness. This is all about capital management.”
Still SBC didn’t want the rating agencies calling on previously unrated borrowers in the initial collateral pool and trying to interrogate them. Instead of checking each individual borrower, the ratings agencies had to check SBC’s internal credit analysis and credit-scoring system. The rating-agency models require 88% of underlying credit exposure to be to borrowers that are rated investment grade, according to SBC’s own system. The agencies rated SBC’s rating system.
The only worry for investors is that there has been little experience of how pay-outs from credit derivatives will work in actual defaults and how the mark-to-market mechanism for reference securities will fare in potentially illiquid markets accompanying a default. There is also the risk that an immediate pay-out will be less than the eventual recovery rate.
But the deal was generously priced and increased from an initial target of $1.5 billion. It has been structured as a master trust indicating that SBC Warburg Dillon Read intends to sell more such deals in future.
