Lawbitrage

India Embraces CDSs Just as EU Gives them the Cold Shoulder

While the European Union is in the midst of banning naked credit default swaps (CDSs) referencing the debt of its member states, the Reserve Bank of India (RBI) has approved the introduction of CDSs referencing corporate bonds.

The EU justified banning naked sovereign CDSs due to the instruments "fuelling market volatility" and "potentially aggravating Greece's troubles" in the form of increasing its borrowing costs. On the other hand, the RBI justified its own embrace of CDSs in the following way:

Since CDS have benefits like enhancing investment and borrowing opportunities and reducing transaction costs while allowing risk-transfers, such products would increase investors' interest in corporate bonds and would be beneficial to the development of the corporate bond market in India[.]

Although the moves by the two regions are not mutually exclusive, it seems that India is the one headed in the right direction.

While there is little doubt that CDSs generally improve credit risk price discovery, the evidence for corporate CDSs improving the underlying cash market is not overwhelming. But if Asia's experience from recently introducing corporate CDSs is any indication, India is in good company. According to a paper by Ilhyock Shim and Haibin Zhu, in Asian markets "CDS trading has lowered the cost of issuing bonds and enhanced the liquidity in the bond market."

By contrast, there is good reason to be skeptical of the EU's ban on naked sovereign CDSs. In a paper published by the Banque de France, Darell Duffie notes his research finding no significant relationship between CDS speculation and countries' borrowing costs. More recent research by several Italian economists came to a similar conclusion:

Having regard to the European government bonds market, there is no clear evidence that speculation through CDS has affected the prices of the underlying bonds, nor that it is possible to manipulate the price of CDS in order to generate de-stabilising informative signals on the credit risk of sovereign issuers.

Hedge Funds Are Safely Helping Banks Free Up Capital

Basel III's new capital requirements for loans made to businesses are in some cases higher than banks' internal risk assessment, so banks have been looking for ways to offload loans, or at least the credit risk associated with the loans. Hedge funds are assisting banks in this process, and through structures that seem relatively stable.

A story in FINalternatives last month reported about this phenomenon in the case of a World Bank subsidiary-sponsored hedge fund established to incentivize banks to make loans to developing market companies. Yesterday, a Financial News story by Harriet Agnew shed further light on how hedge funds are helping banks to free up capital, noting that the funds are primarily using synthetic securitization structures to take the first- or second-loss risk from the banks.

Agnew describes the capital-relief transactions as involving banks that

securitise some of the assets on their balance sheets using credit derivatives and other techniques, offloading a portion of the risk to non-bank [hedge fund] investors....

Typically, risk-sharing transactions work by packaging parcels of assets, such as high-grade corporate loans or counterparty risk, and transferring a portion of their risk to a non-bank investor, who agrees to absorb the first or second tranche of losses should the loans go sour or counterparties default.

For the banks, this reconciles the high cost of holding these assets with the need to continue lending money to important clients....For the investors, such transactions based on corporate risk are attractive because they are taking on a very precise credit risk, which is cushioned from falls in the bank’s share price and trading losses in other parts of its business.

Importantly, banks are requiring the funds to lock in their investors' capital for five years, reports Agnew. This feature largely removes any concern that these hedge funds have much in common with the commercial paper conduit transactions that caused banks to take substantial losses when, during the financial crisis, the conduits' short term investors chose not to roll over their debt and reinvest in the conduit.

Another important contrast with conduit transactions is that credit oriented hedge funds tend not to be highly leveraged. A recent estimate found their average gross leverage to be 2.4 whereas, by contrast, conduits were 12-15 times leveraged, assuming they were even capitalized with any equity at all.

The lack of short-term funding and high leverage means that the hedge funds will actually be able to pay the banks if their packaged corporate loans default--the risk is being completely transferred to the funds and their investors. Banks may be wrongly deciding that less capital needs to be held against their corporate loans. However, with these capital-relief transactions structured as they are, it is a decision whose wisdom will fall on the hedge funds and their investors, not the banks. In this respect, hedge fund investors may be helping banks make loans to corporations by subsidizing a portion of their risk.

Credit Risk Transfer Governance Theory for Practitioners

A valuable theory should be able to guide practice. I have written an academic article developing a theory of credit risk transfer (CRT) governance that at a general level should help investors, dealers, attorneys, and other practitioners when approaching securtizations, credit default swaps (CDSs), and other CRT transactions. The theory provides a broader context for best practice guides.

CRT governance is the extent to which a CRT transaction protects investors (or counterparties) from losses in the underlying credit risk being transferred. CRT governance mechanisms include collateralization and netting for CDSs, and credit enhancements and active collateral management in securitizations.

The general principles of CRT governance include the following. First, good (or high levels of) governance can protect investors (or counterparties) from losses even if the underlying assets whose credit risk is transferred experience significant losses. Bad CRT governance, by contrast, creates transaction structures that leave parties with highly sensitive exposures to losses in underlying credit instruments. In addition, either good or bad governance can be efficient. Efficiency requires the price of a CRT instrument to accurately reflect the protection provided by governance mechanisms in relation to the credit risk being transferred. However, governance mechanisms are limited in their ability to protect investors or counterparties if the underlying credit risk is too high. The direct implication is that practitioners in CRT markets should increase or decrease the governance of a particular transaction relative to the risk of the underlying credits--and price accordingly.

A February 2009 article in Risk by Ashish Dev and Bo Qian reflects CRT governance theory in the context of residential mortgage-backed securities (RMBS):

When PD [probability of default] is relatively low, it is possible to make high-quality (AAA-rated) CDO tranches with close to zero EL [expected losses] out of RMBS mezzanine tranches, although the CDO tranches often have ULs [unexpected losses] that cannot be ignored. On the other hand, for subprime mortgages or in situations where PD rises to 5%, even a credit enhancement of 50% is not enough to make the senior tranche of the CDO safe in terms of EL. Moreover, the corresponding UL is substantial. . . . [I]t is almost impossible to create an AAA tranche in a CDO structure backed by mezzanine or junior RMBS tranches with subprime mortgages as underlying collateral.

Unfortunately, the basic principles of CRT governance were violated by practitioners with the $300 billion worth of CDOs created from mezzanine RMBS that were issued from 1999 to 2007. Fortunately, post-crisis mortgage securitizations now seem to be more in line with the basic principles of CRT governance.

 

Why Most Credit Default Swaps Will Never Be Completely Regulated

Yesterday, the New York Fed published a report about the credit default swap (CDS) market. An important implication of the report is that most CDSs will likely never be completely regulated as envisioned by the Dodd-Frank Act. Market participants simply do not trade most CDSs with sufficient frequency to make them eligible for mandatory central clearing.

The Dodd-Frank Act requires all types of CDSs to be cleared by a central counterparty and traded on an electronic trading platform. However, the Securities and Exchange Commission (SEC) is empowered to decide that certain types of CDS don't need to be centrally cleared and traded. Section 763(a) of the Act requires the SEC to base its decision upon the contract and financial characteristics of the CDS in question, including how standardized and liquid (often traded) is the CDS.

According to the NY Fed report, about 19% of CDSs that reference a single issuer or obligation, and 52% of index CDSs, are currently eligible for clearing. This finding bears directly on what the SEC is likely to regulate because CDSs are currently non-clearable precisely due to the factors the SEC must consider in deciding whether to mandate clearing. According to the NY Fed:

Eligibility for clearing can be seen as an indirect indicator of the level of standardization and liquidity of an instrument, since central counterparties choose to support a product based on the ability to price, trade and effectively model the risks of the product.  

The NY Fed's statement reflects what is widely understood. The European Central Bank, for example, recognized the same over two years ago when in a 2009 report it noted that for a derivative contract to "be eligible for clearing, a product must, as a minimum, be liquid, have price transparency and be standardised" (page 79).

Accordingly, if non-clearable CDSs are to be eligible for clearing, they should become more standardized or more liquid (or both). Importantly, however, non-clearable CDS are already highly standardized, due in part to the Big Bang and Small Bang protocols of 2009. Non-clearable CDS liquidity, by contrast, is relatively low. According to the NY Fed report, the overwhelming majority of non-clearable CDSs do not trade more than once a day. Clearable CDSs are more liquid, with 50% averaging two to four trades per day. This means that if currently non-clearable CDSs are going to be cleared, they will have to become more liquid--there is far less room to grow in the area of standardization.

However, CDSs are unlikely to become more liquid anytime soon. CDS dealers and end-users seem unwilling or uninterested in trading them more often. As noted in a Bloomberg Businessweek article by Matthew Leising, dealers in particular may be reluctant to trade CDSs more often because doing so would prevent them from capturing gains from transactions with clients. End-users, for their part, are unlikely to increase their CDS trading frequency because they tend to approach the instruments with a relatively long-term purpose in mind.

But just because most CDSs are not liquid enough for the SEC to require central clearing does not mean that most CDSs will be completely unregulated. Uncleared CDSs will still be subject to Dodd-Frank's reporting and capital requirements. Because the CDS market was generally well governed and not a source of systemic risk even prior to the passage of the Dodd-Frank Act, the fact that most CDSs will fall outside of the full scope of the Act's regulatory regime should not by itself be a cause for concern.

(Updated on 9/29/11 with additional observation about governance in final paragraph.)

Central Clearing of Credit Default Swaps May Be a Systemic Risk Double Whammy

Although the derivatives industry has been moving toward a central clearing model for over-the-counter derivatives for years, and the Dodd-Frank Act requires OTC derivatives to be centrally cleared unless they are subject to an exception, efficient and safe central clearing may be easier said than done, especially when it comes to clearing credit default swaps (CDSs).

A Risk.net story from the September 20th ISDA conference in London highlights some of the technical difficulties in coordinating multiple central counterparties:

Interoperability between central counterparties (CCPs) in the over-the-counter derivatives market is not a realistic prospect in the near term, according to a panel of clearing experts. . . while CCP interoperability now exists in the cash equities market, it will be much harder to achieve in the OTC space. One hurdle is the different margin standards and default fund requirements used by individual clearing houses, meaning it would be difficult to calculate the exposures between two entities. Even if a best guess were made, the exposures between CCPs would then need to be collateralised – which may not be straightforward.

Numerous localized (i.e., non-interoperable) CCPs may increase systemic risk, the participants noted.

But systemic risk from CCPs may increase even if they achieve interoperability simply because they are required to clear CDSs. The following was also reported today about separate remarks made by the COO of the Depository Trust Clearing Corporation:

Credit-default swaps are “scary, scary products,” Bodson said in a discussion of the transaction infrastructure that is emerging after the 2008 global credit crisis. The exchange of risks involved in these swaps may not fit well on a transaction model that has been used for futures contracts, which are simpler forms of derivative securities....“We may be creating a much higher level of risk,’’ he said, by requiring swaps to be turned into standardized products, traded on electronic exchanges and where risks are covered by a central counterparty, backing up the two sides of a trade.

If CCPs may increase risk due to not being interoperable anytime soon, and cleared CDSs may increase risk for independent reasons as well, the central clearing model for CDSs is something market participants and regulators may want to reconsider. Indeed, there is strong academic research on why CCPs may increase systemic risk, which I note here (on pages 31-32).