By Luis Araneda, Member and Director of Structured Capital Partners, LLC
Structuring credit card ABS (CCABS) for a decade while at Banc of America Securities, I've seen issuers born and I've seen them die. I thought I had seen it all but new surprises continue. I would like to provide another view of CCABS structural weakness and the broader implications on the ABS market.
Surprises include downgrades, botched servicing transfers, servicing abandonment, and regulatory actions including curtailing issuer activities and overriding deal documentation, in varying degrees touching many issuers. These include NextCard, Metris, Capital One, Providian and now First Consumers. Although not CCABS, Heilig-Meyers used a revolving CCABS structure and served as the first hint of what was to come. CCABS deals are structured to withstand rising charge-offs and originator/servicer insolvencies. These events aren't surprises. Surprises are structure breakdowns. The key factors leading to breakdowns include the revolving feature of CCABS combined with aggressive industry practices. The revolving process exposes other vulnerabilities that are just now receiving a lot of attention: servicer risk, servicing fee adequacy, trusteeship, issuer liquidity and accounting methods.
Revolving means the trust mechanism of adding to existing receivables future purchases on both existing accounts and on new accounts automatically added to the trust. In certain respects, card pools are like operating assets. Analogously, securitization of non-revolvers such as autos, student loans and most mortgage pools is like peeling a layer off an onion. Securitization of revolving assets is like an entire onion because of a broader scope of activities. How does a revolving trust expose other vulnerabilities? Let's begin with servicer vulnerability.
Unlike non-revolvers, CCABS are exposed to terms and credit quality of future receivables. Investors trust that future receivables are underwritten and serviced consistently. However, issuer and ABS investor interests don't necessarily align. Issuers may increase credit limits, dig deeper in credit in originating, or offer lower-priced cards to gain market share. Some issuers have attempted to outgrow aggressive underwriting and a deteriorating economy. This was easy to attempt, using revolving structures. Deal documents allow servicers flexibility in generating and servicing future receivables - how they will be solicited, underwritten and priced. Issuers oppose restrictions and investors' only protection is rating agency approval or notification for material changes. Servicer vulnerability is intimately intertwined with future receivables risk. In cross-border future flow deals the gap between deal rating and servicer and sovereign rating is intentionally limited because of risk of interference in the transaction for the latter's benefit. How is this risk different in CCABS where repayment may be dependent on future receivables?
Servicing fee adequacy: The standard 2% CCABS servicing fee was originally established by examining costs for large issuers with healthy revolving portfolios. This became standard for CCABS without considering higher liquidating portfolios costs (rating agencies assume liquidation in stress scenarios) and subprime expenses. When riskier portfolios become troubled, servicers might note that 2% is inadequate and in combination with a limited amount of their own capital invested, realize it's advantageous to hand portfolios over to trustees. Another effect of low servicing fees is lower credit enhancement levels because the servicing fee is deducted from modeled cash flows in sizing enhancement, leaving less credit and interest rate risk protection.
Trusteeship: Rating agencies recently published excellent trusteeship articles: see Moody's Re-examines Trustee's Role and Fitch's Seller/Servicer Risk Trumps Trustee's Role. Moody's expect more of trustees while Fitch instead implores closer examination of seller/servicers. I agree with both and believe the solution is for trustees to charge more when needed and to subservice tasks others can perform better.
Liquidity is another vulnerability unique to CCABS. Because CCABS revolve, additional financing is needed to fund both portfolio balance increases and amortization, unlike amortizing asset classes where an asset-liability match exists from closing to ultimate repayment. As banks, corporate bond and equity investors have turned bearish, potential funding for new purchases and amortization has become more scarce - exacerbating servicer vulnerability.
By accounting vulnerability I mean that revolving pool performance numbers can be easily misinterpreted. It's difficult to assess credit risk in revolving pools because of new purchases masking performance. New purchases inflate the loss-rate calculation (period charge-offs divided by period average or point-in-time receivables) denominator, reducing the calculated loss rate. CCABS static loss analysis to date isn't truly "static". In the case of amortizing assets the "cumulative default rate" is truly static and reflects how much of a loan is ultimately charged off. In cards this information doesn't exist. CCABS version of static loss includes new purchases, thus providing less liquidation value information. Here's an example of misunderstood accounting. One rating agency's criteria states: (a) "The purchase rates modeled for bank cards range from 2% to 5%" and (b) "In a AAA scenario, the ultimate charge-off level is increased by three to five times the issuer's steady-state charge-off level". Assume a portfolio with equal month-end pool balances, 6% annualized loss rate, 10% monthly payment rate (MPR). Solving for a monthly new purchase rate (NPR) sufficient to maintain stable month-end pool balances results in NPR of 10.5%. If the loss-rate increases to 15%-25% and MPR is halved per criteria, can we assume new purchases continue at 2%-5%? And who would throw good money after bad? Experience shows that as losses within a portfolios increase, NPR moves to significantly lower levels.
Aggressive industry practices include subprime trusts that quietly funnel into public prime trusts, liberal overlimits, low minimum payments resulting in negative amortization, shuffling accounts between workout programs to avoid charge-off, payment matches, and aggressive accrual and recovery accounting practices. Of course, other sectors and industries are also aggressive, but in the case of revolving structures these practices can be harder to detect. NextCard reportedly inappropriately booked bank capital through intercompany transactions and didn't record as charge-offs obligor first payment defaults. Why the surprise over intense regulatory scrutiny?
Broader ABS market implications? Maybe little because the main issue is specific to cards - their revolving nature. Revolving trusts make it easier for issuers to downplay risk through a credit "barbell" strategy (one end subprime and the other end low-value superprime teasers), more difficult to assess servicing fee, more difficult for trustees to understand undertakings, and easier for issuers to shift risk to investors post-closing. The revolving nature pushes CCABS somewhere between truly isolated ABS and corporate bonds in credit risk.
Credit card structures do work, especially in allocating credit and interest rate risk among multiple security tranches. The subordination mechanism has been challenged in a previous blowup of a revolving subprime installment contract trust but subordination prevailed, despite a servicing meltdown, with subordinate investors bearing more losses than senior investors. In addition, unsecured bondholders were wiped out while AAA ABS bondholders experienced principal repayment. My main point is that because of CCABS' revolving nature, investors perhaps should rely on additional credit risk gauges besides historical performance, demand more due diligence, keep a keen eye on servicer quality and carefully examine the risk/reward tradeoff.
Copyright 2003 Thomson Media Inc. All Rights Reserved.
