By Michael Detwiler

There is a critical choice facing lenders shopping for new technology to deliver efficiencies to their origination shop. It comes down to picking a loan origination system with an embedded rules engine or one that's got one bolted on.

Today's newest loan origination systems (LOSes) are powerful platforms that condense into a few minutes of data entry what once took many hours of hard labor. No one in the business long enough to have taken a 1003 by hand would ever want to go back. And yet, there is so much more that today's technology can do for the origination process that isn't being fully utilized. Because traditional mortgage companies operate on a set of processes that have been standardized over a number of years, these enterprises essentially have become trapped. Traditional or legacy LOSes were developed specifically to follow the workflow that lenders were forced to adopt as part of their manual, paper-based processes. Today's new LOSes or enterprise lending solutions (ELSes) are finally capable of allowing lenders to break out of those old inflexible workflow paths. Recently developed lending platforms are capable of doing virtually anything that a mortgage executive can imagine. Any kind of loan can be processed for borrowers of every description. Yet many lenders never take advantage of these capabilities because they assume that making system changes remains a costly and time-consuming endeavor. In some cases, lending professionals have just accepted that many of their requests are simply impossible.

Lenders have embraced business-rules technology to make changes to their lending platform that will positively affect workflow. Even so, operations executives run into roadblocks in their efforts to enhance their lending process due to the time it takes to coordinate with their information technology (IT) department. Complicated system integrations that require IT involvement slow down the process and make it difficult for lenders to implement changes quickly. This prevents them from effectively changing their automated workflow in order to get products to market rapidly and gain a competitive advantage.

Lenders realize that the mortgage market changes rapidly. While they are now investing in technologies flexible enough to serve their changing needs, they are finding it difficult to implement new business policy within their technology quickly and effectively. This prevents them from getting the return they should on their technology investments.

The BRMS solution
The generally accepted solution has been to provide software that will integrate with the lending platform and make the management of business rules easier for lending executives. Called a business rule management system (BRMS), this software interacts with the various lending platform components and the central database in order to establish processes that the automated systems would work through during the loan's life cycle.

According to Needham, Massachusetts–based TowerGroup, "A BRMS is essential in mortgage lending, which is one of the most complex account-acquisition processes in financial services. Mortgage industry firms manage thousands of internal and external policies and rules in mortgage credit evaluation, underwriting and loan pricing. Among these firms are mortgage lenders, credit-reporting companies, mortgage insurance companies and mortgage investors."

Due to continual changes in mortgage-related data and the vast amount of information involved in a mortgage transaction, I believe the only efficient way to implement policy, rules or pricing changes is by utilizing a BRMS of your own.

A typical BRMS consists of a database of business rules the company has established; a user interface for the creation and maintenance of the business rules; and an engine that uses logic to apply the rules to the data coming from the user and the lending platform.

Because the vast majority of lending platforms were not built with such a rules-management system embedded in the software, most of the BRM systems available today bolt onto the lending platform. Of those that can "bolt on," they were designed and developed by third-party vendors that have had little, if any, impact on the development of the LOS their engines are designed to complement. This has led to the current debate over the pros and cons of those lending platforms with embedded business rule management systems and those with external "bolt-on" business rule management systems.

Putting the power to work quickly
The first benefit of purchasing a lending platform with embedded rules management is there is a significant reduction in the total cost of ownership. On the implementation side, having an embedded BRMS eliminates the separate purchase, installation and integration of a disparate application. On the low end, license fees for a stand-alone BRMS a mortgage lender would utilize range from $100,000 to $300,000. When you tack onto that the price for implementing, integrating and configuring the rules to make them lending-specific, then fee increases of four times orgreater are evident. Based on a business-rules study performed by Framingham, Massachusetts–based IDC Research Inc., the total cost for implementing a third-party rules engine is between $500,000 and $2 million (including license fees, maintenance, training and business-rule development). One can conclude that going with an embedded BRMS would save you these costs.

Time saving is also a factor with an embedded BRMS. There is no need to go to the trouble of installing the loan origination system and then installing the business rule management technology. When configuring the lending application, the business rules are being configured at the same time. No additional time is required to make certain the two systems are configured properly or to test that they work well together. They were designed to work together seamlessly from the beginning.

Having the rules system embedded also saves time and money in the back office and for the IT department. Companies do not have to maintain separate maintenance and support schedules or manage relationships with separate vendors to ensure the continued performance of the lending platform. Everything is handled by the same vendor—therefore, accountability and responsibility for continued return on investment (ROI) rest with one party.

The IT department also only has to maintain one database of record, reducing the possibility for errors and, in some cases, decreasing lag time due to data transportation to and from the external system.

Building and maintaining the rules
Once implemented, the business rules repository becomes a key component to the system. The ease by which the business user can create, change and maintain the rules determines the real effectiveness of the BRMS. When the engine is embedded with the lending platform, it takes less time to formulate complex rules.

When both systems share the same data dictionaries, there is never a need to create a new data element in one of the systems in order to build the required workflow. When the rules management system is built from the same native code as the lending platform it serves, both systems have access to all of the same functionality without ever having to build new interfaces.

Not only does it take less time to build new rules, an embedded BRMS also generally will have mortgage-specific rules, events and actions that are already core to the system. Obviously lending organizations will have their own nuances when it comes to process. However, they will not be starting with a blank slate, as with a stand-alone BRMS. This typically allows companies to get up and running much quicker.

Additionally, those companies that do not have formalized processes in place (and there are still a number of companies that fit this description) may want to utilize the framework provided to help shape their business.

An additional benefit to sharing the same data set is that the IT department need not track the changes made to either system in order to update the other for future changes. With an embedded system the technology has a single database of record, so cross-referencing two locations is not needed.

Companies also save time when it comes to accessing information about changes to their central database. When the business rules engine is separate from the lending platform, the system must poll the LOS and its database in order to determine status changes that can affect the decisions it makes. It takes more time to access information from these remote systems than it does to access a shared database, as in the case of embedded systems.

Seeing the business in a new way
Perhaps the most important benefit, in my view, to having an embedded rules engine that is native to the processing software it controls is that it allows the business user to stop thinking in terms of processes and begin thinking in terms of data.

The mortgage business is famous for reducing everything to the documents. Regulation and competitive pressures have created an industry that runs by the document in order to guarantee compliance. While that may be necessary in some parts of the business, it becomes a severely limiting factor when the company begins to innovate.

In the final analysis, mortgage lenders don't succeed based on the documents they complete; their success is determined by the data elements they collect and sell. The secondary market buys loans, and most lenders now understand that the closing package is simply the documentation that certifies that a legal loan exists.

When lenders have the ability to think about the data in the transaction, they begin to see new ways to move and use them. They can then begin to write rules that will operate on those data for the good of the company, instead of just for the benefit of the legal department.

When the BRMS is embedded into the lending platform, it can begin to look at more than the documents in the transaction. Because it is native to the lending application, it can seek out individual data elements wherever they may be and populate needed information elsewhere in the loan lifecycle. Lenders can then analyze those elements separately to determine the best way to move the loan to completion.

Processes can be changed based on the combination of data available at the time, moving certain loans swiftly on to closing at the same time the system flags others for human intervention due to potentially fraudulent information.

Because every aspect of a stand-alone business rule management engine must be integrated into the lending platform in order to be effective, it is more difficult for these systems to reach this level of functionality.

BRMS in action
Homeowners Loan Corporation (HLC), Atlanta, is a nonprime lender. Using its prior lending platform, HLC's closing department took roughly 35 minutes to create a closed-loan package. "By utilizing Mortgage Cadence's embedded business rules engine [to automate our workflow processes], we decreased the amount of time to 20 minutes per closed-loan package," says Paul Sadler, chief technology officer at HLC. "At implementation we were generating 1,100 closed-loan packages per month, trimming roughly 275 hours out of the process—a significant efficiency gain for our corporation."

With the embedded business rules functionality of its new system, HLC was also able to achieve a savings from annual salaries in the closing department of roughly $200,000 and put those resources to better use within the organization. And utilizing these business rules has also helped HLC reduce the number of errors involved in processing a loan. It now has the ability to create rules within the system to review documents before they close, and to verify that all of the needed information is included and accurate. As HLC upgrades or migrates to newer versions of its lending platform, there will not be any versioning or capability issues with its business rules because they are native to the core application.

Furthering the argument for embedded business rules is the fact that HLC experienced a very minimal (60-day) end-to-end implementation period. During this time it was able to map out its processes, configure its business rules and put them into action. Had HLC chosen a lending platform that did not have an embedded BRMS, the possibility exists that the company 1) would not have the automation and workflow functionality necessary to cut its closing times and reduce manual errors or 2) would have needed to purchase, implement and integrate a third-party BRMS application, which would have taken longer than 60 days.

The power to expand, built-in
Depending upon how the embedded BRMS and the processing system are designed to work together, lenders that use such integrated applications could save significantly on future software purchases. If the software is only developed to handle the loan origination process, this may not be the case. However, if the LOS is designed to take full advantage of today's advanced architectures, the processing system and the right BRMS could be a powerful combination. Some stand-alone business rule management systems maintain that they are flexible because they were first designed for and used in other industries. But lenders have learned that functionality only becomes powerful when it is easy to use. Because these engines are not native to the lending software they are applied to, it is not often simple to get them to function in any way other than how they are preconfigured by the manufacturer.

Even if the desired functionality is available to these systems, few lenders will have the resources to get the software to actually work, and those that do have the resources will spend a great deal of time and money making that happen. In my mind, the question then becomes, is all of this effort worth the return, when one can have a system that includes the same functionality native to the lending application? Further study and analysis will give the industry the ultimate answer.