The Big Freeze

The big freeze
ABCP: Anatomy of a liquidity crunch.

By Marlene K. Puffer, managing director, Twist Financial Corp.

Fixed income markets have been in the headlines since mid-summer 2007, with the spotlight on mortgage-backed securities, structured products, corporate credit, and liquidity. The impact on Canadian institutional investors who steered clear of U.S. Credit Default Obligations (CDOs) or Structured Investment Vehicles (SIVs) has been mainly indirect through the repricing of risk and mark to market valuation impacts on other credit investments like corporate bonds. However, there are many investors who have taken a direct hit and remain mired in negotiations to sort through the mess in Canadian Asset-Backed Commercial Paper (ABCP).

ABCP provides short-term funding to conduits (typically trusts) that hold portfolios of longer-term (typically five to 10 years) financial assets such as credit card receivables, prime and subprime mortgages, and other types of loans, as well as CDOs, which themselves contain financial assets and derivatives, often on a highly levered basis. In this article, I explore the dynamics behind the ABCP market and the role of the U.S. subprime market in creating the liquidity problem currently facing Canadian investors.

The trigger for the global liquidity mess was rising defaults in the U.S. subprime mortgage market. This was coupled with a lack of transparency in CDOs and ABCP conduits. This lethal mix exploded in fear and panic as investors worried that there could be more subprime mortgages buried within structured investment products that could lead to more defaults than they had previously anticipated. Concerns arose that the structural enhancement within the conduits and CDOs would provide less protection against default than previously believed, and there was insufficient information to assuage these fears where warranted. Liquidity evaporated rapidly as investors feared the worst. In Canada, the tinder was particularly dry as the liquidity backstop protection provided by banks to many ABCP conduits differed from the global standard and was not triggered for conduits that were not sponsored by the big Canadian banks. ABCP investors in these third-party conduits were left holding portfolios of long-term assets when what they thought they bought was a high-quality short-term liquid investment.

Credit Conundrum – overview
It is easy to get lost in the details and complexities of the market for structured financial products such as CDOs, SIVs, and conduits that issue ABCP. The key feature is that structured products pool together many different assets (such as credit card loans, mortgages, CDOs, etc.), and raise funds to purchase them by issuing various types of securities that appeal to different investor risk appetites. For CDOs, the term of the funding and the term of the assets generally match fairly closely. SIVs tend to use a mix of short- and longer-term funding and ABCP conduits rely entirely on short-term paper so are subject to the most liquidity risk (i.e. risk that the current ABCP investors will not be able to sell to new investors at a reasonable price). The quality of the ABCP depends on the default rate of the underlying portfolio of assets, and on other specific product features that include credit enhancement, and liquidity enhancement. The market meltdown was exacerbated by a lack of transparency and information about the details of the underlying assets.

The simplest form of credit enhancement is overcollateralization, which means there are more assets in the pool than are required to pay off the investors in full, with the cushion determined by the default assumptions. Other forms include reserve accounts to cover excess losses. Fears were sufficiently intense this summer as investors perceived that the credit enhancement would offer insufficient protection, and that even highly rated securities could suffer losses. This fear contributed to refusal to roll over their investments in ABCP or other short-term funding for structured vehicles.

Structures are also designed with liquidity enhancement to provide protection when buyers do not materialize. Liquidity enhancement takes various forms, including extendible notes, and liquidity facilities such as liquidity loan facilities, where a loan can be drawn to finance the assets, or liquidity asset purchase agreements, where the assets are purchased out of the conduit. This type of liquidity is available on a same-day basis. Liquidity enhancement does not cover defaulted assets in most cases—this is usually covered through credit enhancement. Clearly the rating of the liquidity provider is an important determinant of the rating of the ABCP.

Global style liquidity can be triggered in two general circumstances: if the specific conduit is unable to obtain funding within a certain time period and at a specified spread level, or a general market disruption that affects all ABCP. For most Canadian ABCP, the triggers were weaker than in other markets and only included general market disruption clauses. Only a handful of conduits had global style liquidity. Strong liquidity enhancement can mean that there is less need for extensive credit enhancement.

The added complication is that many structured products have poor transparency. The information flow is particularly limited and complex for structures that contain CDOs that may themselves contain other CDOs. Investors do not have access to detailed information about the quality of the individual loans or their delinquency rates, and in the case of CDOs, the relevant details may be layers deep in documentation. This lack of transparency began because banks did not want their competitors to see details about the loans they were securitizing, and other conduit sponsors wanted to prevent competitors from easily copying their structures. Rating agencies and investors did not insist on change. This approach continued due to strong demand for these assets by investors who were willing to accept tight spreads and limited disclosure in a market hungry for yield.

The U.S. Situation
About 25% of new mortgage originations in 2006 in the U.S. were subprime (compared to less than 5% in Canada). As housing prices chugged higher, lending standards deteriorated and risky features proliferated in the U.S.. Now that housing prices are falling, the U.S. subprime problem is very bad. Many of these loans are in or nearing default, and the number of loans that are up for renewal at sharply higher rates is still climbing. There have been downgrades of many mortgage-related securities, and shoes continue to drop. The rate of delinquency increased rapidly in 2006, and got even worse in 2007, with roughly 6% of loans going delinquent or worse within three months, compared to less than 2% in 2005 and 3% in 2006 (Figure 1). The rapid pace of defaults is exacerbated by falling real estate valuations as borrowers walk away from loans obtained with little or no equity in the home.

The contagion effect of subprime mortgages arose due to widespread securitization that placed prime and subprime mortgages in the hands of investors who had no role in the origination of the loans, and thought they were buying controlled risk exposure to diversified portfolios of financial or synthetic assets through ABCP, CDOs, and SIVs etc. mortgages and mortgage-backed securities are held directly in conduits, and indirectly through CDOs that may hold additional mortgages. For some conduits, investors may only know that some assets are residential mortgages, with no assurance whether they are prime or subprime. In some cases, investors may be aware that a fraction of the portfolio is subprime mortgages, but have no information about the quality of the loans. The fear factor kicked in and liquidity deteriorated, as investors feared the worst in the absence of full disclosure. Conduit sponsors who underestimated the potential for a liquidity crunch and were slow to reassure investors (or may not have had reassuring information to offer) exacerbated this.

Canadian ABCP Market Dynamics
The U.S. subprime problem impacted ABCP and other structured investments globally and has pushed corporate spreads wider, but the impact on liquidity in the Canadian ABCP market was most severe. The Canadian ABCP market was very large and vulnerable to a shock for several specific structural reasons. With the advent of government budget surpluses around 2000, the Government of Canada reduced issuance of bonds and T-bills. With less government supply and limited growth in issuance of corporate commercial paper and Banker’s Acceptances (BAs), ABCP conduits filled the void. They began with financial assets such as mortgages, credit card receivables and consumer loans, and eventually came to rely mainly on synthetic assets including CDOs. This meant virtually unlimited supply since conduits could be quickly created with a great deal of flexibility.

The U.S. and European ABCP markets are larger in absolute terms, but only a fraction of their CP markets: 50% in the U.S., and 30% in Europe. By June 2007, the $280 billion Canadian money market consisted of $117 billion ABCP, compared to $21 billion of Government of Canada T-bills and $56 billion of Banker’s Acceptances (BAs), and $60 billion in other Commercial Paper (CP). Trading volume in ABCP far exceeds volume in BAs or T-bills, roughly $40 billion per month compared to roughly $30 billion for BAs and $20 billion for T-bills.

For years, fixed income investors have faced low yields and tight credit spreads and have been on the hunt for extra yield. Many money market investors were keen to invest in high-quality assets that offered a spread despite ratings from a single rating agency and poor disclosure. Most global ABCP is sponsored by a major bank, which also provides a liquidity backstop. This type of bank-sponsored ABCP dominates the Canadian market. However, Canada also had a sizable market for independent conduits that drove most of the growth in Canadian ABCP since 2004. As of the end of 2006, independents were roughly 48% of the $117 billion of ABCP outstanding, up from roughly 15% of the market in 2004 (source: DBRS).

Liquidity Canadian Style
In 2000 OSFI, the Canadian banks regulator, published regulation B-5, which implied that banks could provide “General Market Disruption” liquidity backstops to ABCP programs without having to set aside capital. However, global style liquidity required capital reserves. Canadian ABCP with only the General Market Disruption liquidity backstop was rated AAA by DBRS and accepted by investors. For bank-sponsored conduits, there may have been reason to believe that despite the language, banks would support their own conduits under broader conditions. When independent conduits were created, they continued to be rated by DBRS and many investors continued to accept them with little distinction in spread relative to bank-sponsored structures. The prevalence of CDOs in the conduits raised alarm bells, and in 2006 DBRS changed their ratings criteria, requiring stronger liquidity provisions for CDO-based structures and effectively putting ice on that market.

Bonds that are sold in Canada generally have ratings from at least two agencies. ABCP has been the exception with only one rating available. The other rating agencies clearly stated their opinions that they could not rate Canadian ABCP conduits since the liquidity provisions were inadequate by global standards. Prior to August 2007, investors did not vote in sufficient numbers with their dollars and insist on more than one rating. Deals were successfully sold with tight spreads, so there was no incentive for sponsors to change their single-rating approach.

Times have changed dramatically as DBRS has finally changed their standard to match the other rating agencies requiring global style liquidity for all ABCP. Banks are, and will continue to be under Basel II, required to provide capital reserves for this style of liquidity backstop since it means they bear more risk than with General Market Disruption style liquidity. This makes ABCP more expensive to issue. The evolution continues. The ABCP market may revive itself eventually, likely without third party-sponsored structures.

The Implications
Liquidity and credit risk have been repriced. Credit spreads are wider in money markets and longerterm bond markets, despite the absence of any major corporate defaults. Fear is still a factor, and information about Canadian ABCP and global CDOs is slow to flow. Opportunistic investors are willing to provide liquidity at a price, but they need information in order to determine that price. The Montreal Accord is aiming for a restructuring of Canadian ABCP to longer-term notes, but there are many hurdles and wildcards that could derail that process along the way. The amount and detail of legal documentation associated with the many different types of structures, and sorting through leverage and swap agreements embedded in the more complex deals are daunting tasks.

Major U.S. banks are creating a super fund to try to clean up the SIV market without taking the assets on balance sheet, but at this point the plan is to only buy high-quality assets into the fund and it is not clear how this will help. Some banks have taken assets back on balance sheet, and have funded in the term markets. In Canada, in the third quarter of 2007, 75% of new bond issuance was by banks. Along with fear and writedowns due to subprime and CDOs, this bank funding need has contributed to corporate bond spread widening, despite an absence of major corporate credit events.

Regulators and others who monitor or influence the financial markets are all looking for answers and responses. Disclosure is likely to be the main focus of their responses to facilitate smooth functioning markets without undue regulatory interference. The liquidity crisis has highlighted that investor due diligence is critical and that reliance on ratings alone for opinions about complex financial assets is insufficient for investment decision-making. Clearly, in order to conduct adequate due diligence, information is required. Investors have finally spoken with their dollars and markets are responding.

Acknoledgement
This article is based in part on presentations made at the Treasury Management Association of Canada (TMAC) conference in Vancouver on September 18, 2007 (with Colin Kilgour) and at the Insight Fixed Income conference in Toronto on October 29, 2007 and reflects only my opinions.