:: Home :: Magazine :: Free Newsletter :: Daily News :: Video Newscast :: Mortgage Focus Videos :: Archive :: Mortgage Tech Awards :: Apply :: Top Providers :: Mortgage Tech Glossary :: Executive Spotlight :: Tech Niches Blog :: Tech Innovation Blog :: Featured Story :: Tech Savvy Lenders :: New Buyer's Guide :: Conference Calendar :: Classified Related Sites :: National Mortgage News :: MortgageStats :: BrokerUniverse :: Mortgage Servicing News :: Mortgage University :: WeirdLoans :: Mortgage Grapevine








Archived Columns

Are Business Rules Engines for Everybody?

By Scott Kersnar, Editor, Mortgage Technology Magazine

I've talked in the past about application service providers and why I've long thought the hosted model would prove a good fit for many mortgage players. This time I'll tell you why I suspect that success for every bank and mortgage firm eventually will be pegged to success in automating rules management.

A caveat: If a mortgage lender gave me the keys to all their facilities and I had an entire week to rifle my way through all their systems by myself, the chances are slim that I could find out whether or not they deployed BRMS systems at all. I'm a journalist, not a systems analyst or engineer, but I've long been curious about what turns good ideas into complete thoughts and makes successful businesses tick.

I once came close to taking a position as an account exec with Southland Corp. where I would have acted, in effect, as a cog communicating corporate business rules to a group of 7-Eleven franchisees. A broad rule for convenience stores is never to let merchandise spoil or gather dust. Patrons have to know they can duck inside to get common items in a hurry with no worry about expiration dates.

If convenience store owners wanted to guarantee they'd never have quarts of milk spoil, Southland told me, then they put the oldest quarts in the front row on the right hand side of the shelf at eye level above the handle on the cooler door, because those are the milk quarts that get grabbed first. If a franchisee wanted to sell a lot of beer on the Fourth of July, then the franchisee would display the promotional materials exactly where the account exec told them to. The franchisee also absorbed a big discount on 12-packs -- and thereby enjoyed a big jump in business throughout the store. The Southland account exec was charged with having nearly perfect success in persuading franchisees to do things the way corporate has learned they are best done. If spoiled milk or weak Fourth of July numbers showed up in any of his stores, then the account exec was in immediate danger of getting fired.

Computers can derive if/then rules all by themselves. During the dot-com boom, that was how a Web-based grocery-delivery outfit called Pinkdot.com discovered that if a consumer went to the website around seven in the evening to order delivery of an off-the-shelf sleeping preparation, almost 100% of the time the answer was "yes" to an invitation to bring the consumer a fresh hot cup of coffee along with the sleeping pills. Contradictory? Maybe not, if it's dinnertime and the consumer is already thinking about how to get to sleep quickly after first slaving the evening away on an important project. In any event the issue was not why, but how often, consumers said yes.

A common problem for account execs in a franchise (substitute wholesale mortgage or net branch) network is that most franchisees like to think of themselves as entrepreneurs. They want to differentiate themselves from their fellows as well as their competitors. If they decide to sell hand-made jewelry up at the cash register, hey, no account exec is going to tell them to have a Fourth of July placard touting the beer promotion instead.

Nobody defers to a "best practice" they haven't seen work with their own eyes. That's a big reason why it has taken a long time for many underwriters to trust AU decisions and stick to handling exceptions instead of second-guessing every decision the computer makes.

As AUS has increasingly been deployed at point-of-sale, BRMS systems have accompanied them, first as product and pricing engines and then to orchestrate other functions. "An AUS essentially renders decisions that are ordinarily tightly focused primarily in automating underwriting decisions and evaluating credit worthiness," observed Skip Hulme, director of product development for Irvine, Calif.-based Portellus. "A BRMS, however, is capable of governing functions beyond underwriting such as data workflow validation, compliance, best execution, fee management, resource allocation, customer interaction, vendor policies, infrastructure procedures and more. While an AUS working alone really only interprets guidelines and examines credit worthiness to determine eligibility, a BRMS takes it a step further by delivering risk-based pricing, workflow tasks, task assignments" and so on.

What's more, explained Mr. Hulme, "a BRMS can be programmed or configured to govern all of an organization's policies, procedures and rules using predefined, structured descriptions of actionable business items. As a result, a BRMS enables a company to operate with precision, flexibility and agility, thus ensuring efficient and profitable processes across the entire enterprise, and provides it with a shared, consolidated view of the organization's collaborative business practices."

In the mortgage industry there are no quarts of milk to worry about or sleeping preparations to deliver, but leads are perishable, and closed loans need to get to the investor in good shape. Things work better if you can fine-tune where those leads go and how the loan file gets handled at every stage of the origination process.

As Fair Isaac's James Taylor often observes, separating business logic from the application code allow greater responsiveness to change.

Rules-based systems can make hundreds of lightning-fast either/or decisions and route them in a manner that fits a mortgage lender's shifting business priorities. When lenders can consistently deploy their best practices as rules, there is a very rich potential payoff -- in change management, in calculating the ROI for technology projects on an enterprisewide basis, in making corporate governance as Sarbox-compliant as possible, in agile launching of new loan products.

Most large companies maintain their internal metrics "at a very high level," notes Overture's Lee Decker. They can't afford to reconfigure their core systems on a whim, and they don't deploy BRMS systems to figure out when to cross-sell somebody a cup of coffee, though they may deploy smaller systems to automated decisioning down to that level of granularity. Nevertheless, by now BRMS technology is variously useful to the big players. Smaller companies have had to wait for vendors like Overture and Portellus to give them "the technology that would have been cost prohibitive five years ago," said Mr. Decker, who pointed out that business-rules engines are now getting cheap enough for small to midsized lenders to afford. With the aid of BRMS systems, he noted, they, too, can use automated underwriting to address the near misses and thus structure deals for a larger percentage of borrowers.

Mortgage lenders sold on the value of BRMS for converting best practices into automated rules are most likely to prosper from the evolution of lights-out processing and e-mortgages. A good working hypothesis, it seems to me, is that lenders will look for the best payoff from BRMS to the extent that they expect to thrive in an atmosphere of continual change.