We are all aware that the mortgage industry is beginning to embrace technological delivery of services. What may not be obvious is the dramatic changes that should come about when these changes are fully implemented. This report is an attempt to get a handle on those changes and what they mean for industry participants.
To analyze the coming mortgage superhighway, we set up the "perfect" case of a mortgage that is originated largely without human intervention. The computer program that guides the mortgage is called HAL, after the computer in one of our favorite movies, "2001: A Space Odyssey."
Then, we discuss how reality is likely to differ from the perfect case. And finally, we describe the winners and losers among the various industry players.
* The potential borrower calls up HAL on a computer or TV and fills out an application form. Application data are verified partly by computer and partly by a human review of employment, income and net worth.
* HAL accesses a credit report on the borrower from one or more credit agencies.
* HAL requests an appraisal on the property to be purchased. All home sales are stored in a national data base, so comparable sales data can be accessed by computer. A human is needed to verify the existence and physical condition of the house and the property.
* HAL requests a title check. All title documents are stored in a national data base.
* HAL now underwrites the loan, adding a regional housing forecast to the borrower and property data. HAL not only approves or disapproves of the loan but, in the case of approval, calculates a credit risk score for the loan. In the case of disapproval, HAL suggests way for the borrower to work toward qualification, or names human financial counselors to contact.
Next, HAL requests price information, including quotes for private mortgage insurance if the loan is nonconforming, quotes from Fannie Mac and Freddie Mac for their own secondary mortgage insurance if the loan is conforming a required loan yield from investors, and loan servicing costs from loan servicers.
Finally, HAL presents the borrower with firm price quotes on one or more loan types. A button is pushed and the loan is dosed.
The sales function will not disappear. Not everyone will be comfortable borrowing from HAL, our computer. Many will want to sit down and review his or her personal situation and options with someone from a similar species. Others will desire at least phone contact with a human being.
Title insurance should not disappear that quickly. Only about 6% of homes sell each year and therefore require a check of relevant title documents, so it would take a long time for all the homes in the country to have a reason for computerizing their title certifications.
Computerized underwriting has a number of problems. For one, at least some borrowers will be difficult, if not impossible, to evaluate with a computer because of messy financial histories or unverifiable sources of income. At the minimum, some borrowers who do not currently qualify may need intensive counseling about how to qualify.
Data base ownership presents challenges. We assume that the superhighway works best with essentially public data bases; in other words, that the data bases are available to all potential providers of a service. That is the case today with credit histories. There are several consumer credit agencies that maintain personal credit data. Any lender can access the data for a fee.
The appraisal business seems to be developing in a similar manner. Specialty firms have sprung up to collect national home sales data.
A question mark is on mortgage default histories needed for underwriting analysis. Here, data is kept by lenders and mortgage insurers, including Fannie Mac and Freddie Mac. Other large players like MGIC, GE, and Countrywide Credit also have large data bases to use to develop and fine-tune automated underwriting programs. These companies have distinct advantages over smaller underwriters.
We do expect that over time, Fannie Mae and Freddie Mac will open their data bases to the underwriting industry at large or simply provide an underwriting program. However, there could still be an advantage for large underwriters of nonconforming loans.
This article was adapted from a recent research report by Mr. Gordon.