Following more than a year of market skepticism, the high-loan-to-value sector is poised for redemption and growth, as the performance figures continue to impress analysts.
According to study by the mortgage team at PaineWebber, default rates in the 1998 vintage deals are expected to level at 4%, compared to the whopping 8% in the 1996 vintage.
PaineWebber's study looked at more than 340,000 loans across 29 deals, which could be the largest high-LTV study to date, according to PaineWebber. Of significance, the study described the emergence of a credit curve, which until recently hadn't taken a definitive shape, because of the sector's youth.
"We weren't expecting to find some of the things we did," said Tom Zimmerman senior vice president and co-author of the report. "The story that sort of unfolded here is the changing credit characteristics in the high-LTV market, and there's really been an increase in the credit quality over the past four years."
One of the reasons for the rise in credit quality - also a finding of the study - is a significant improvement in the underwriting standards in hi-LTV lending from 1996 to 1998.
Zimmerman pointed out the irony, whereas in a lot of other sectors of ABS and mortgage-backed securities, particularly the home equity sector, underwriting standards actually worsened from 1996 to 1998.
"I think what happened was in retrospect, the initial underwriting standards were too loose, and everybody acknowledges that now," said Zimmerman.
"The better players learned from their mistakes, and the weaker players just aren't around anymore," said John Cerra, managing director of ABS at TIAA-CREF.
When loans are underwritten, it takes a while before delinquencies and losses start developing.
"It's going to take you six months or year before you really know what's going to happen," Zimmerman said. "When the delinquencies started to build up and then the losses, the guys who were in it for a long time - like [Empire Funding Corp.] and [FirstPlus Financial] - really started to change their underwriting standards."
In addition to expected findings - such as the importance of disposable income in evaluating a consumer's credit - certain non-quantitative characteristics of underwriting standards emerged, when certain pools performed better than others despite having similar average Fair Isaac & Co. (FICO) scores and debt-to-income statistics.
Cited from the study, "In discussing the [Residential Funding Corp.'s] performance, with RFC and other industry members, it became clear that even though RFC's quantitative criteria were similar to other issuers, their non-quantitative underwriting standards were more stringent."
Examples of non-quantitative criteria are 30-day mortgage lates, time since last bankruptcy, and simultaneous closure on first and second mortgages.
According to the study, high-LTV underwriters have made loans to borrowers with bankruptcies as recent as four years back. RFC, however, will not lend to a borrower with a bankruptcy on record. A bankruptcy, however, will not appear on a borrower's record if it happened more than seven years prior.
"It's a pretty simple concept," said Dean DiBias, a managing director at RFC. "Do you want to lend to an individual who has proved their lack of willingness to repay their obligation, as defined by the fact that they've had a very, very recent bankruptcy?"
"One of the things we do know that we didn't go into that much in this report, is that roughly 50% of all the defaults in this product are bankruptcies," Zimmerman said. "So if you can do anything all-priority ahead of time, if it's weed out any borrower that might have a propensity toward bankruptcy."
Giving that the 1998 loan pools are performing well, and that they were underwritten using looser standards than what are used today, portfolio performance should continue to improve going forward, Zimmerman added.
"I think the 1999 figures and the 2000 figures will look even better," he said.
As to how the product will price as the sector evolves:
"I expect that, as time goes by, the spreads between home equity and 125 paper will narrow," said DiBias. "And ultimately it's our belief, that 125 paper, given a superior prepayment and convexity profile, over subprime lending as an example, will ultimately trade at or through home equity product.