Though cash and U.S. Treasury securities are the usual collateral for derivatives transactions, in recent years non-standard collateral (NSC) such as corporate bonds and derivatives of various descriptions has been increasingly used for this purpose. There is risk related to this type of collateral, i.e., devaluation, which is managed primarily reduced by haircuts, ratings triggers, netting, and mark-to-market valuation, and limiting the types of acceptable NSC (for example some counterparties who accept NSC will not accept interest-only or principal-only securities, or those linked to the residual interest in an asset pool). However, these risk minimization strategies do not address the matter of 'issuer' risk related to certain types of NSC. Issuer risk could be described as risk related to securities representing property (i.e., cash flows) that might be subject to competing claims.
Issuer Risk
For example, say a counterparty posts as collateral securities from a tranche of a collateralized debt obligation of a General Conglomerate (GC) special purpose vehicle as collateral. The CDO was created when GC set up GC Auto Receivables LP III (GC LP III), to which it 'sold' car loans. If GC gets into financial difficulty GC's creditors should not have recourse to GC LP III's assets--the car loan receivables--because GC transferred them to GC LP III as a 'true sale'. Of course, GC LP III continues to have its own risk issues, such as erosion of value of its assets, although this may be mitigated through credit enhancements such as financial guaranty insurance.
If assets/cash flows underlying NSC are challenged as really belonging to the entity which originally held them (in our example belonging to GC as opposed to GC LP III), the counterparty that took the NSC as collateral now has issuer risk in addition to the other risks previously discussed. This issuer risk envisions, at worst, bankruptcy of the originator with the accompanying forensic examination of its financing transactions (i.e., securitization deals creating NSC). This challenge might be made in a Chapter 11 scenario where the creditors of GC seek to avoid its securitizations as fraudulent or preferential transfers, or might seek to substantively consolidate GC and GC LP III on the theory that GC LP III is a sham corporation operated as a single-entity with GC to hinder, delay, or defraud creditors. Substantive consolidation can occur when corporations disregard corporate forms, such as the failure to observe corporate formalities, commingle assets, or assume obligations for affiliates in less than arms' length transactions.
Arguably, financial guarantees from monolines like AMBAC and MBIA would preclude NSC payment defaults, but only if the financial guarantees of payment were:
* Absolutely default risk free; and
* To the exclusive benefit of the holders of the NSC, irrespective of the validity of their claims to the property collateralizing the NSC.
On this latter point, if the guaranty is a contract between each NSC holder and the guarantor, the guaranty payments would be shielded from an attack by the issuer or its creditors in a Chapter 11 context. Of course, it is best if the guaranty is paid for by funds unconnected to the issuer.
Risk Is Still Risk
Given that companies convert 'risky' cash flows (i.e., receivables) into hard cash, thereby strengthening their balance sheets, the Chapter 11 scenario described above would seem a theoretical risk to a NSC holder. However, bankruptcy remoteness is still a concern because risk, theoretical or otherwise, is still risk. So, how is issuer risk mitigated in light of demands by counterparties to post NSC as collateral? In addition to the standard risk reduction techniques like haircuts, parties accepting NSC can further reduce their risk by limiting the amount of NSC they will accept for a particular transaction, or from a single issuer. The latter strategy is important because dollar limits on NSC can be circumvented unless such a restriction is used.
For example, assume a party to a swap transaction has a dollar limit on AAA CDOs as collateral set at USD10 million per CDO issuer. Accordingly, if USD20 million collateral is required its counterparty may post USD10 million (ignoring haircuts) of a CDO issued by X, and USD10 million of a CDO issued by Y. It would appear that this is proper as X and Y are different entities. However, what if X and Y are affiliated? Even if one assumes that each CDO has an AAA/Aaa rating (i.e., because its respective issuer is a bankruptcy-remote SPV whose creditworthiness is independent of that of the originator of its assets and enjoys the credit enhancement features discussed before), as stated previously, if an originator is in Chapter 11, any assets it has 'sold' to its SPVs might be claimed as by the originator's creditors. For this reason, when a counterparty offers NSC ostensibly from different issuers, but whose actual provenance is identical because the issuers are affiliates, the NSC should be considered as coming from the same issuer. This characterization puts teeth into dollar limits on NSC because it recognizes that a holder of NSC from affiliated SPVs magnifies its issuer risk relative to the originator. Recently I created a template containing such a provision:
"Notwithstanding anything to the contrary contained in this Agreement, on any date, for either party . . . . if any U.S. Corporation ("Corporation X") whose Private Securities comprise any portion of a party's Posted Collateral is an Affiliate of any other U.S. Corporation or entity ("Affiliated Corporation") whose Private Securities also comprise that party's Posted Collateral, the Private Securities of Corporation X and those of the Affiliated Corporation shall be considered as having been issued by a single U.S. Corporation for purposes of this Section."
Whether a dollar limit is placed on specific NSC for a transaction, a group of transactions, the aggregate exposure to a particular NSC issue (and that issuer's affiliates), or an asset pool (NSC backed by auto loans versus senior tranches of first-priority home mortgages), NSC issues should be of great interest to a derivatives player's most senior management. Specifically, senior management should recognize that by accepting NSC they absorb issuer risk as a fourth horseman after credit, transaction, (i.e., profit or loss on the trade itself), and market risk. Accordingly, in addition to familiarizing themselves with the basics of derivatives, such management should require credit personnel to:
* Be intimately familiar with any variety of NSC the firm accepts;
* Diversify NSC to limit issuer risk; and
Be familiar with industry measures and academic writings relating to collateral valuation (on the latter point, see, e.g., Didier and Saliasi, 'The Lender's Default Risk Mitigation Policy and its Implications on the Borrower's Optimal Trading Strategies', University of Lausanne, December 2002).
Closing Thoughts
It axiomatic that collateral, as a pledged asset, is valuable to all concerned (the lender, borrower, and prospective buyer) from the inception of a transaction through to default and liquidation. With 'standard' collateral, value is unquestioned so risk is truly mitigated. The opposite is true when the toxic tranche of a CDO or other type of NSC serves as security. An anecdotal illustration of this last point is the experience of one portfolio manager who watched USD5 million of a BBB collateralized bond obligation tranche dissolve into USD87,500. S (see 'Loose Accounting Rules Keep Defaults Off Books For Now', The Wall Street Journal, Jan. 15, 2003). Moreover, the Standard & Poor's definition of a BBB rating does not explain such a decline: "An obligation rated BBB exhibits adequate protection parameters. However, adverse economic conditions or changing circumstances are more likely to lead to a weakened capacity of the obligor to meet its financial commitment on the obligation." See Standard & Poor's Ratings Definitions at www.standardandpoors.com. Although this security was purchased as an investment, and may not have been subject to the daily vetting of mark-to-market valuation, a possible 98% write-down would make a NSC holder justifiably nervous. For a 50% haircut on USD10 million, no one could ask for USD285 million of collateral (i.e., * 57 x USD5 million in collateral where such collateral is later devalued by 98% to USD5 million) yet the magnitude of this manager's loss speaks volumes.
It is wise to remember that an AAA/Aaa rating has, at least in part, a marketing function, and that for SPV-issued securities bankruptcy remoteness is an element of that rating; so, those who tout bankruptcy-remoteness understand its value in the capital markets.
Recently the cfo of a major hedge fund was overheard grousing about the haircuts his prime brokers imposed and the "avarice" underlying their margin rates. Of course, the collateral he offers his prime brokers is, whenever possible, non-standard. Whatever may be said about prime brokers, the cold fact is non-standard collateral should cannot responsibly be treated as low-risk relative to standard collateral, and parties accepting it without safeguards like those discussed above, counteract the very purpose of collateralization.
This week's Learning Curve was written by Alice Griffin (linearoption@mindspring.com), a New York-based lawyer and consultant on derivatives transactions.
