PrefBlog

I’ve just read a good paper by Engel & McCoy, Turning a Blind Eye: Wall Street Finance of Predatory Lending, which, while certainly having an axe to grind (they want more regulation), does give a good overview of the problems. Their central point is:

As this excerpt from one prospectus illustrates,securitization turns a blind eye to the underwriting of subprime loans:

With the exception of approximately 20.82% of the mortgage loans in the statistical mortgage pool that were underwritten in accordance with the underwriting criteria of The Winter Group, underwriting criteria are generally not available with respect to the mortgage loans. In many instances the mortgage loans in the statistical mortgage pool were acquired by Terwin Advisors LLC from sources, including mortgage brokers and other non-originators, that could not provide detailed information regarding the underwriting guidelines of the originators.

As this suggests, Wall Street firms securitize subprime home loans without determining if loan pools contain predatory loans. In the worst situations, secondary market actors have actively facilitated abusive lending.

The big problem (to me) is loan re-negotiation, which has been the subject of some political chatter in recent weeks:

Securitization complicates and often blocks work-outs with borrowers who are harmed by predatory loans. This is because the underlying securitization contracts tie the trustee’s and servicer’s hands if they attempt to negotiate a repayment plan in lieu of foreclosure. The value of the securities and the amount of their returns are based on cash flows that are determined, in part, by the loan terms. To protect these cash flows, securitization contracts typically prohibit changes to the terms of the underlying loans. In addition, securitization contracts often prohibit servicers from waiving prepayment penalties and other loan provisions.

There is a very good table in the Engel & McCoy paper (on page 2056 of the Fordham Law Review), showing S&P Upgrades and Downgrades of Public Subprime RMBS, 2003-2006. It’s not a proper transition matrix, but it’s a start. 

Anyway, what brought on this surge of interest in the mechanics of sub-prime was the recent announce by Fitch:

Fitch has affirmed three classes and downgraded one class of notes issued by Northwall Funding CDO I, Ltd., (Northwall). The following rating actions are effective immediately:

–$165,326,758 class A-1 notes affirmed at ‘AAA’;
–$46,500,000 class A-2 notes affirmed at ‘AAA’;
–$40,500,000 class B notes affirmed at ‘AA’;
–$18,000,000 class C notes downgraded to ‘BB’ from ‘BBB’ and remain on Rating Watch Negative (RWN).

Northwall is a collateralized debt obligation (CDO) that closed May 17, 2005 and is managed by Terwin Money Management, LLC (Terwin). Northwall has a substitution period that grants Terwin limited trading ability until September 2007. The portfolio is composed of approximately 89% subprime residential mortgage-backed securities (RMBS), 8% Prime RMBS, and 3% CDOs.

The downgrade of the class C notes reflects the deterioration in credit quality of the portfolio.
Approximately 11% of the portfolio has been downgraded since last review and as of the most recent trustee report the WARF has increased to 5.07 (‘BBB/BBB-‘) from 4.35 (‘BBB/BBB-‘) at last review. In Fitch’s view approximately 13.5% of the portfolio is below investment grade quality, including approximately 5.9% ‘CCC’ or lower quality. There is one defaulted asset comprising $994,341 of the portfolio. In addition, approximately 7% of bonds in the portfolio are on Rating Watch Negative (RWN).

As far as I can make out from a google-cached report by Credit Suisse, the original issue came in the tranches indicated above, with an additional “equity tranche” representing 5% of the issue.

There was another announcement by Moody’s:

Moody’s Investors Service today announced downgrades on 120 securities originated in the second half of 2005 and backed by subprime, first-lien mortgage loans. The actions follow a review of the securities rated in the second half of 2005 and affect securities with an original face value of over $1.5 billion, representing 0.7% of the dollar volume and 4.1% of the securities rated by Moody’s in the second-half of 2005 that were backed by subprime, first-lien loans.

 

The actions reflect the higher than anticipated delinquency rates of first-lien subprime mortgage loans securitized in the second half of 2005. These loans were originated in an environment of aggressive underwriting, although not to the same degree as the subprime loans originated in 2006. Aggressive underwriting combined with the prolonged slowdown in the housing market has caused significant loan performance deterioration and is the primary factor in these rating actions. Moody’s has noted a persistent negative trend in severe delinquencies for first-lien subprime mortgage loans securitized in late 2005 and 2006.

 

The vast majority of these downgrades impacted securities originally rated Baa or lower. In total 54 securities originally rated Baa and 60 securities previously rated Ba were downgraded. Additionally, 6 tranches originally rated A were downgraded. No action was taken on securities rated Aaa or Aa.

 

In addition to the high rates of early delinquency predicating today’s actions, Moody’s notes that subprime mortgages originated in late 2005 and 2006 that are subject to interest rate reset present an additional cause for credit concern. Subprime borrowers from previous vintages of such collateral avoided “payment shock” and potential default by refinancing. However, with the recent pressure in home price appreciation and tightening of mortgage lending standards, such refinancing opportunities may be more limited. Moody’s has noted that transactions issued in the second half of 2005 have begun to exhibit slower prepayment speeds as they near the two-year interest reset than did prior vintages. Moody’s is actively surveying loan servicers to evaluate the impact of potential increases in loan modification due to these upcoming resets.

Why do I bring this up? Well … no real reason. I just wanted to point out that the highest rated tranche of the issue reviewed by Fitch was the best-protected $165-million of a $300-million issue … asset coverage of 1.8:1, in fact, to put it in terms familiar to those who invest in Split-Share preferreds.

Also, I’m really annoyed at all the weeping and wailing over sub-prime. There have been significant losses, but so far they have been borne by

i) those who bought the lower-rated or equity tranches, and it serves ’em right!

ii) those who have panicked and sold stuff into a panicked market because it has the word “sub-prime” in it somewhere. To say something is a sub-prime derivative is about as meaningful as saying something else is an equity. It comes in many flavours.

There’s more perspective over at Tom Graff’s Accrued Interest.

There has been some slight readjustment in the market lately at the intersection of the three titled subjects, and I’ve received some queries regarding how it all works.

So, for those who don’t wish to read Moody’s explanation of the market (hat tip: Financial Webring Forum) or the Federal Reserve’s Examiners’ Supervision Manual, here’s a stripped down version of how it all came together. Please note that I have no inside information whatsoever regarding the specifics of the holdings or clients of Coventree’s trusts; also note that I am recklessly making up the numbers with a view to showing how the system works, not with a view to calculating actual profits:

(i) A hedge-fund guy (HF) has $100 he needs to invest.

(ii) HF is offered some 6% 30-year sub-prime paper and decides that it’s a good investment.

(iii) HF buys $1,000 of this 6% paper and borrows $900 on margin at 7% to pay for it. At this point he has negative carry, which is a Bad Thing.

(iv) Coventree offers to lend him $900 at 5% for ten years against the assets, provided he over-collateralizes. HF borrows the $900 from the Coventree trust at 5% for a ten year term and uses these proceeds to repay his margin debt. The loan is secured by a pledge of the $1,000 worth of 30-year 6% sub-prime paper.

(v) Coventree then issues $900 of three-month paper to yield 4%. The buyer is … Investor Guy (IG), who needs a liquid investment but doesn’t want to buy T-Bills yielding 3%.

So at this point, everybody’s happy:

(a) The ultimate financer has $900 worth of three-month paper yielding 4%

(b) Coventree’s trust is borrowing $900 at 4% and lending $900 at 5%, for projected profit of $9 annually.

(c) HF is levering his $100 capital into a $1,000 investment which pays $60 interest annually and financing $900 at 5%, paying $45 annually. HF thus has a positive carry of $15 annually on an investment of $100 and has a chance at a capital gain … if the market goes his way, he can sell the sub-prime paper and collapse the loan.

Everybody’s happy. Coventree is confident they’ll be able to issue three month paper for the next ten years; HF is confident he’ll be able to take out 10-year loans for the next thirty years. The ratings agencies poke around inside Coventree and say, hey! There’s over-collaterallization here, and positive carry on the underlying investment. No problems. Until step six:

(vi) IG reads in the paper that sub-prime paper is worthless. All of it! Not only are all those deadbeats going to default on their mortgages, but the houses won’t be worth anything after foreclosure.

(vii) IG gets a note telling him that Coventree trusts have sub-prime paper in them – something like that disclosed by Coventree in their latest MD&A:

Management continues to believe in the high quality of those underlying assets. Coventree-sponsored conduits have limited exposure to U.S. subprime mortgages – less than 4% of the total assets in Coventree-sponsored conduits are backed by assets related to U.S. subprime mortgages. Those assets continue to perform within the range of initial expectations and, as such, continue to be rated AAA, and are not expected to be materially affected by the recent increase in delinquencies and losses in that asset class.

(viii) IG just wants a really safe Money Market investment. He doesn’t like headlines. He therefore does not buy any more Coventree-sponsored paper.

(ix) Coventree-sponsored trusts can no longer operate since they’re unable to repay the money market notes as they come due.

(x) The Montreal Proposal is made; notes will be issued to reflect the underlying trust asset; thus, there will be a conversion of the overnight paper into a ten-year note, secured by Coventree’s ten-year loan to HF.

In short: Coventree was doing a classic bank thing: borrowing short and lending long, making money on the term spread and the quality spread.

They ran into the other classic bank thing: a run.

Update, 2007-08-24 : Tom Graff has explained a variation on the theme. 

Update, 2007-08-24 : The issuers of the short term paper did anticipate that a run might happen and set themselves up with one of two defenses:(i) Extendible Notes … if the paper can’t be rolled, they can just extend the maturity by a period of time, which gives them some breathing space.

(ii) Back up Liquidity – in the example above, they’ve got $9 net interest income. They can enter into a contract with a bank, whereby they pay the bank $1 annually for an emergency borrowing facility. The trouble starts when the word “emergency” is defined, as noted by DBRS:

During the weeks of August 13 and August 20, 2007, DBRS noted that while many ABCP issuers continued market issuance activities in the normal course, a number of ABCP issuers were unable to roll their ABCP maturities. In these instances, backup liquidity facilities were drawn upon and while liquidity was advanced in several cases, in others they were not. In cases where liquidity was not advanced, investors may now be exposed to mark-to-market risk in the underlying assets.

In further comments,

During the last two weeks, DBRS noted, a number of issuers of asset backed debt, also known as ABCP, were unable to roll over – or find buyers for – debt as it matured.

The situation was made worse when the issuers couldn’t get cash under liquidity provisions set up as a sort of safety net in case of market disruptions.

[DBRS group managing director Huston] Loke said this is one area that DBRS is looking at, since its rating system hasn’t previously looked at the reliability of the liquidity agreements put in place between the issuers and their financial institutions.

He noted that some financial institutions provided liquidity when requested by ABCP issuers and others did not, even though the contracts were worded similarly.

“I think that’s something we would need to look at in terms of our revisions to criteria,” Loke said.

Update, 2007-08-24: Coventree uses the term “credit arbitrage” to describe its process; Fabrice Taylor made fun of this term:

Coventree earns fees and spread income from trusts that buy assets like mortgages, leases and other receivables that earn interest income. They then issue shorter-term debt to investors on which they pay interest, hopefully earning a spread. The company calls this spread revenue “credit arbitrage,” which is a misnomer because, by definition, arbitrage is supposed to be a risk-free profit.

Fabrice Taylor’s track record was not disclosed. According to the Moody’s primer linked above:

Credit arbitrage programs are bank-sponsored programs that invest in securities rated Aa3 or higher. The programs are similar to a cash flow CDO, but funded with short-term liabilities instead of term debt. They generally have no credit enhancement because the securities are highly-rated and the program administrator must sell the securities or provide credit enhancement if the assets are downgraded. The liquidity facility is sized at the face amount of ABCP outstanding and purchases assets at book value, implicitly protecting investors from market value risk. These programs exist largely because the initial BIS regulatory capital regulations for commercial banks do not distinguish between highly-rated and lower-rated securities. By funding off-balance sheet, banks obtain regulatory capital relief. The program also serves to diversify the bank’s sources of financing.

And according to Coventree:

Credit arbitrage transactions closely resemble derivative transactions which are designed to transfer risk from one highly sophisticated financial institution to another. Coventree’s revenues for credit arbitrage transactions consist of the spread between the return on the underlying investment and the conduit’s cost of funds.