With the increasing popularity and the rising concerns about IO loans, analysts at Merrill Lynch set out to separate fact from the fiction. What they found was that there is no major difference between IO mortgages and other types of mortgages. Their findings were published in a report entitled, Interest Only Mortgages: Separating Myths from Reality.
The number of IO mortgages saw a drastic increase from 2002 to 2005 - moving from less than 2% of all subprime originations to almost 26%. This rapid growth appears to have capped out at about 23% of the origination share in 2006. Still, at slightly less than a quarter, IOs are still a relatively popular option for home buyers looking for alternative options, Merrill analysts said. Specifically, subprime IO loans have become popular in recent years as a way for borrowers to cope with rising home prices and the declining affordability of traditional mortgage products. Previously, Merrill said that IO mortgages were primarily used only in the high net worth borrower segment of the population.
And with IO mortgages' increased popularity came rising speculation and concern over the various risk factors, including the issue of "payment shock" when the IO period expires and of borrowers stretching to afford expensive houses. In addition, since subprime IO mortgages are typically hybrid ARMS and the IO period and the ARM reset period coincided in many instances, there is also concern about "double payment shock," Merrill analysts said.
"The myth that IOs will under- perform amortizing mortgage is not supported by empirical data," Merrill analysts said. "Once we control for all the major factors driving credit and prepays, there exists no significant difference between IO and non-IO mortgages." They added that IO mortgages have undergone significant changes since becoming more popular. In their current form, IOs mirror more of the attributes of a subprime loan than a niche product, and accordingly, some of the risks cited above are slightly lowered.
Similarly, analysts found after utilizing three different approaches to analyze the IO and amortizing mortgage performance that if the two product types are properly compared, the difference in performance is not really considerable.
In their analysis, Merrill researchers began by comparing IO and non-IO mortgages over time without controlling for credit attributes. Initially, it seemed IOs actually outperformed non IO-mortgages in their short period of existence, but these results were not very telling as much of the differences were due to better credit and geographic profiles.
The approach that analysts viewed as the "true indication of performance variance" was to compare loans that are similar. This method involved the one-to-one mapping of loans with the closest set of attributes. They found that the difference in performance starts to fade even when only looking at LTV, FICO and loan size.
"The behavior patterns exhibited by the subprime borrowers with these mortgages are very similar (as we see when we analyze loans with similar attributes) - the performance difference comes from factors other than the choice of the product type (IO versus amortizing mortgages)," analysts said.
(c) 2006 Asset Securitization Report and SourceMedia, Inc. All Rights Reserved.