© 2024 Arizent. All rights reserved.

S&P's new LEVELS model hits high LTVs hardest

Home equity ABS issuers continued to bombard the market last week - nudging spreads and pushing volume expectations to an amount more than May's $71.8 billion as well as last year's issuance over the same time period. The impressive number, however, was more likely the result of issuers itching to complete deals before quarter-end than a strategic move to sweep under a July 1 deadline before Standard and Poor's applies steepened credit enhancement requirements for most deals.

"Our initial interpretation of S&P's very short and unhelpful description of LEVELS was that it should not have a significant impact on credit enhancement levels required of subprime securities settling in August," said Michael Youngblood, director of ABS research at Friedman Billings Ramsey. "When we had the opportunity to compare LEVELS 5.7 to the prior version, we came to the same conclusion."

S&P this week will begin rating deals using the updated version of its LEVELS model, which its analysts use to determine the risk of default - and subsequently the amount of credit enhancement issuers are required to post - for all non-agency RMBS deals. The shift is expected to boost S&P's credit enhancement requirements nearer to those of fellow rating agency Moody's Investors Service, particularly in regard to penalties for higher loan-to-value ratios. As Youngblood put it, "If you are already subject to needing a Moody's rating, S&P has not changed your life at all."

While any spread widening, reported by JPMorgan Securities as a result of the HEL issuance deluge, is generally expected to be short-lived, some implications of the model shift could be more long lasting. Issuers could tighten underwriting guidelines to avoid more expensive subordination requirements, and, according to Nomura Securities, the change could result in larger classes of subordinate bonds relative to triple-A rated bonds.

Piggy-backs:

no longer a free ride

Primarily, the much talked about distinguishing factor between the rating agency's LEVELS 5.6 model and its 5.7 model now in play is the harsher treatment toward higher LTV ratios and so-called piggy-back or silent second-lien loans. And this is for no small reason. The portion of adjustable-rate loans baring the silent seconds increased by 19% to 38% in the first quarter of this year from the fourth, while fixed-rate loans with seconds increased by 60% to reach 24%, according to Moody's. Both rating agencies and market players have raised an eyebrow at the seeming uptick in the prevalence of piggy-backs in recent years. Moody's and S&P say that this could be an indication that borrowers are buying more house than they can afford. LTV ratios at origination in the first and second half of last year averaged 81.1 and 80.5, respectively, according to Peter DiMartino, ABS Strategist at RBS Greenwich Capital.

Up until the first of the month, issuers seeking an S&P rating were not required to post excess credit enhancement if the proportion of piggy-back loans within a given pool totaled less than 20% and excess credit enhancement required for a pool because of piggy-back loans was capped at 40%. However, the rating agency has changed its methodology from the pool-level approach to a loan-by-loan analysis. Meaning, it can require as much credit enhancement as it sees fit, based on each individual loan.

LTV ratios of greater than 92% take the hardest hit with the new model, according to FBR - particularly at the triple-A level. The absolute increase in default likelihood according to the new model was on average 13.18% for 92% LTV pools, and 22.50% for 94% and greater LTVs.

How much extra credit enhancement is required per loan depends on a sliding scale of credit score and collective LTV. For example, the new model requires 1.5 times more credit enhancement than the older model for a piggyback loan with an original LTV below 80%, a combined LTV of 91% and a lower-than 660 FICO, according to Nomura.

Interestingly, FBR's Youngblood last week found that loss coverage requirements at the triple-A level were in fact on average lower when piggy-backs constituted more than 40% of a given pool with the new model - although the difference was admittedly marginal. At the other end of the credit spectrum, at triple-B minus, a 40% piggy-back concentration will now cost 1.9% more in subordination - to a 6.02 loss coverage from 5.91%.

Because of the loan-level analysis applied in the new model, it will also be more stringent toward loans thought to contain "layered risk," such as a combined high loan-to-value ratio and low FICO score. The rating agency also updated its loss predictions depending on the chance that a given U.S. region will experience a decline in home prices. FBR found lower loss coverage levels across all regions at the triple-A level, except for Phoenix and Las Vegas, which increased from the old model by 0.4% and 0.97%, respectively.

Trouble for residuals?

JPMorgan analysts noted last week that the increased focus on such loan-level attributes as silent seconds is bringing with it more spread tiering, as investors - particularly in the lower end of the capital structure - trend toward combing through collateral characteristics with a finer-toothed comb. According to Nomura, a resulting larger class of subordinate bonds in relation to triple-As could cause a ripple of failed transaction triggers - since they are typically linked to the percent of triple-A bonds present in a given deal. Failed trigger tests could lead to a wider spread threat of extension risk for subordinate bonds, preventing over-collateralization from filtering down, and, making residuals a tougher sell.

(c) 2006 Asset Securitization Report and SourceMedia, Inc. All Rights Reserved.

http://www.asreport.com http://www.sourcemedia.com

For reprint and licensing requests for this article, click here.
ABS CDOs
MORE FROM ASSET SECURITIZATION REPORT