To: nextrade! who wrote (5673 ) 9/29/2002 8:29:55 AM From: nextrade! Read Replies (1) | Respond to of 306849 Automated Underwriting, Are Automated Underwriting Systems Really Helping? by David Reed realtytimes.com Would Professor Holmes please call? Sherlock Holmes? We might need a little detective work here. Lots of clues, lots of answers, trying to make a connection. Call as soon as able. What needs investigating? We need to find out why we're setting foreclosure records. Recently reported statistics show that residential home foreclosures have jumped to levels never seen before, and delinquent mortgage payments are at nosebleed levels. Before everyone yells "Duh!" let me suggest another influence other than layoffs. People who go into foreclosure typically don't do so out of boredom. Foreclosure happens when tragic events fall on otherwise happy homeowners. Loss of job, illness, or death of a spouse are usually at the bottom of the pile when people find themselves in the midst of foreclosure proceedings. Not only that, but fewer and fewer people save money like they used to. They can fall behind on a mortgage payment almost immediately after a layoff due to lack of cash reserves. But my question is this: Are people being approved for mortgages with unusually high debt loads? Just a few short years ago, both Fannie Mae and Freddie Mac introduced their latest innovation to increase the number of homeowners in the United States . . . Automated Underwriting Systems. These software programs allows for a lender to enter certain data about their borrower to obtain a loan decision. I recall that when these systems were first introduced my fear was that no longer would a human being be involved in underwriting mortgage loans. If there was a problem with a loan that could easily be explained it was just tough. No reasoning with a computer. But something happened. Not only did my fears not come true, these systems actually approved more loans than may have otherwise been made. Why? These systems attempt to take an overall snapshot of a borrower using historical decisioning models similarly used in other industries such as credit card approvals and automobile loans. Before such programs, if a borrower had a collection account, an underwriter would have the borrower not only pay off the bad debt but provide a written explanation as to what happened. An explanation that the underwriter would approve. Not so now. In fact, some automated decisions not only don't require an explanation from the borrower but the old debt doesn't have to be paid at all. Another factor in loan approval is the all-important Debt Ratio, or the amount of current debt divided by gross monthly income. A few years ago, this debt ratio ceiling was around 38%. Add up all your monthly credit payments along with your mortgage bill and divide that by your gross monthly wages. If it were beyond 38% then the borrower could be expected to provide other "compensating factors" to the underwriter to allow for higher ratios. Such factors could be an excellent credit history or lots of money laying around. And that might get you to 42 or 43%. Automated Underwriting systems can routinely approve debt ratios of 50% and beyond. Something a human being would rarely, if ever, do themselves. A "Fifty" ratio was unheard of. Such systems have come into full force since about 1997 and have been all the rage for lenders to approve loans where no man has ever gone before. But these programs really made headway during times of economic boom, when incomes were increasing, home prices were always rising and everyone had good jobs. Not to mention a chicken in every pot and a NASDAQ at 5,000. Current Automated Underwriting Systems have never really been tested in the current environment. Loans that were approved two to three years ago by an automated underwriting program may not have been approved by a human underwriter. And when, not if, economic times turn for the worse, are lenders doing borrowers a favor by approving loans in the name of expanding home ownership? Like I said, lots of clues. No conclusions. It could be nothing more than circumstantial evidence. But I suggest that we need to take a closer look. No one likes to turn down a home loan. Still fewer like to foreclose. Published: September 27, 2002