ABS CDOs: False dawn for ABS CDOs

COPYING AND DISTRIBUTING ARE PROHIBITED WITHOUT PERMISSION OF THE PUBLISHER: SContreras@Euromoney.com

By:
Louise Bowman
Published on:

ABS CDOs are back buying sub-prime ABS but all remains far from well in the asset class.

A shadow of its former self?
US mezz ABS CDO pipeline as at 21/05/07
IssuerSize ($mln)Lead
Ischus Mezz CDO IV507Citi
Kilts CDO 07-11,000RBS GC*
Nordic Valley 07-11,000CS
Pine Mountain500Lehman
RFC CDO 5501Deutsche
Stockton 1,000WS
*Synthetic
Source: Dresdner Kleinwort

In mid-April, Lehman Brothers reckoned that there were $18 billion to $25 billion of mark-to-market losses residing in ABS CDOs – both deals that had already closed and those still in the works. It is therefore not surprising that activity in the ABS CDO market seemed to dry up overnight; and appetite for home equity loan ABS dried up along with it. But within a few weeks the underlying market seemed to have staged a remarkable recovery; spreads in HEL ABS were tightening by eye-watering increments thanks to buying by those same ABS CDOs – happily ramping up as before.

It is hard to believe that the CDO market can shrug off a problem as large as the US sub-prime mortgage mess as easily as this. According to Deutsche Bank, though, demand for sub-prime ABS recovered significantly during May and many deals were being oversubscribed by three to four times. Spreads at the triple-B and double-B level on HEL ABS deals had come in dramatically by mid-month; from 275 basis points to 175bp at triple-B and from 1,200bp to 500bp at double-B. And it is ABS CDOs that are doing the buying.

Scratch the surface

But scratch beneath the surface and all is not as rosy as it seems. "The market was so strong last year that when the problems emerged there were very large existing pipelines to work through," says Darius Grant, a managing director in Citi’s CDO group in New York. "Dealers have been running off deals on their books. Our belief is that the majority of the deals in the market now had already been ramped up and gone past the point of no return when the problems arose, so it made more economic sense to complete them rather than unwind the warehouses. A lot of the deals that had only just started ramping have either been unwound or put on hold."

So the question arises – what happens when this overhang has worked through? "CDO demand has been almost exclusively from existing CDOs that have been attempting to ramp up for a while," concurs Richard Parkus, research analyst at Deutsche Bank in New York. "When these existing CDOs eventually finish ramping, and if hedge funds back away from the market (after, say, sufficient spread tightening) it is not clear what will happen to demand for subordinate bonds."

In mid-April there were deemed to be $10.3 billion high-grade and $13.8 billion mezzanine ABS deals in the pipeline. But by mid-May the number of new issues on the horizon had shrunk to single figures – less than five by most estimates. "The area where we have seen some activity is Warf 1, triple-A only cashflow and market value deals," says Grant at Citi. (These are deals with high-quality collateral – a high weighted average rating factor – Warf.) These deals generally price very tightly and have small equity tranches, the driver behind them being retention of the super senior. According to Dresdner Kleinwort, on May 20 there were just four US mezz ABS CDOs in the pipe: Ischus, Nordic Valley, Pine Mountain and RSC CDO 5.

So is the current rally in ABS HEL a blip? There is a big difference between grabbing 2007 collateral to finish ramping up a deal that was nearly there and recommitting to the sector for the long haul. Even the most optimistic players in the market would struggle to state that all the problems are over, so the spread recovery in May was being seen as a short covering rally rather than a change in sentiment. But spreads might nevertheless continue to grind in for the foreseeable future.

Two questions therefore remain: how much worse will the performance of existing HEL tranches get, and to what extent will the CDO market get hit? At a recent presentation in London, Bear Stearns senior managing director Dale Westhoff emphasized that the volume of recent adjustable-rate loans approaching their reset dates is not the immediate concern. "So far this is not a reset issue, it is an underwriting issue," he emphasized. "Most [problem] loans are going delinquent long before the reset date. Reset may become a bigger issue next year when loans originated in 2006 that have seen little home price appreciation begin to reset in early 2008. These borrowers will have fewer refinancing alternatives given the recent tightening in underwriting standards. In contrast, borrowers resetting today have seen double-digit HPA." Lehman analysts have calculated that 15% of 2006 vintage full documentation borrowers and 43% of 2006 limited documentation borrowers will be locked out of refinancing.

The desire of CDO managers to avoid this collateral in deals they are ramping is very understandable, and has led to massive demand for 2007 loans. "There is now real improvement in collateral," says Citi’s Grant. "People are beginning to sniff around more. The standard of underwriting is now much better and originators are on a better capital footing. Investors will look for new origination collateral." But sub-prime mortgage origination having cleaned up its act is both good and bad news for CDOs – it means that there will be far fewer assets for mezz ABS CDOs in the future. "The question is whether we should be opening up warehouses for 2007 collateral," cautions Grant. "There is far less supply as there are fewer originators and, with tougher underwriting standards, fewer borrowers." Dealers are therefore rushing to buy 2007 collateral, which explains why these tranches have been four to five times oversubscribed. And many people who were short the market have been covering their positions with 2007 collateral.

Contraction

According to Lehman Brothers, sub-prime cash supply will be down 35% to 40% year on year. "The impact of tighter underwriting standards beginning in February is not yet reflected in housing statistics," Westhoff observed at the Bear presentation. The sharp contraction in collateral available should mean that asset spreads in mezz ABS CDO-land continue their inward trajectory – which is bad news for people trying to complete these deals in the current liability spread environment. "The question is whether there will be a snapping in of CDO liabilities?" asks Grant at Citi. "Asset spreads have come in and liability spreads are still wide – this is the worst-case scenario. IRRs are in the low teens so we need liability spreads to come in." Triple-B CDO spreads priced at an average of 468bp at the beginning of the year but by the end of April had widened out to 629bp. This is good news for CDO squareds, which have reappeared in the market with gusto, but bad news for ABS CDOs as asset spreads tighten in.

The worst is yet to come

Expected timing of mezzanine CDO losses

Note: Assumes a portfolio comprising 2005 and 2006 vintage HEL securities.

Source: Lehman Brothers.


"I think that the market will return to the kind of scenario seen at the beginning of 2006 with very tight asset spreads," muses Grant. "It will be tough to squeeze out an arb."


Move on up

This will get even tougher as CDO managers are expected to move up the capital structure with regard to the assets they buy. This will result in more managed high-grade ABS CDOs and fewer managed mezz ABS CDOs. The mezz ABS CDO market will gravitate towards static and bespoke transactions – as the corporate CDO market did after the disruption in 2001/02. But the market will certainly not disappear, as other asset classes have after poor performance.

"To make the analogy with other types of asset that have lost access to the securitization markets the experience in sub-prime would have to be so bad that it threatened triple-A," explained Gyan Sinha, senior managing director at Bear’s London presentation. "High-yield CDOs disappeared because when triple-A notes get downgraded you are striking at the heart of the securitization market. Manufactured housing loans disappeared because less than half triple-As maintained their rating. People do not write manufactured housing loans any more as the securitization experience was so bad." But even with the reset and performance risk that remains in 2005 and 2006 vintage home equity loan deals, this scenario remains very unlikely for mezz ABS CDOs.