By Gabe Minton

No one, I'm sure, would doubt— given the current market environment— that fraud is at an all-time high and trending up, not down. There are multiple methods for attacking this growing problem. They include legal remedies such as the legislation passed in Georgia in May 2005, which produced a noteworthy drop in mortgage fraud. Then there are technical methods, which include all the databases that have been and continue to be created to combat fraud. Other technical fraud-fighting tools include new types of process automation, which allows lenders and service providers to examine whether they have the optimum event flow to ensure that "red flags" are raised during the origination process in order to detect fraud before (or while) it happens instead of after the fact. Industry experts believe that in at least 50 percent of mortage fraud-related cases, the red flags go undetected.

This column focuses on technical aspects and potentially some challenges to the technical options available to fight fraud. (The column will not focus on legislative/legal issues and remedies.)

They say if you are armed with the right data, you can make the right decision. A lot of times, when it comes to mortgage fraud, there is a very coordinated and potentially sophisticated effort playing out to take advantage of the many gaps that still exist within the mortgage origination process. Industry experts estimate that 90 percent of fraudrelated cases involve someone in the industry (the "coach").

As with an aviation accident, from any one occurrence of fraud, there is usually a path that can be followed that will trace to other forms of fraud—often perpetrated by the same individuals/teams. Sometimes the same individuals even after they have been caught and prosecuted will come back to commit more fraud.

Under the premise of knowing everything in real time, one would argue that if we had the right database, or collection of databases, we could catch fraud as it was happening. There are solutions today that address this in many constructive ways. By using these emerging data repositories to augment your or igination process, you can better prepare your shop to catch fraud during the process instead of after the fact. However, there are challenges with this model, which include the timeliness and integrity of the data available.

Process cannot be overstated in terms of importance and usefulness in identifying and capturing fraud when it occurs. Did you know that the very position that has arguably the highest turnover and one of the lowest salaries in the mortgage industry is usually entrusted with the details of the loan files?

The loan processor is responsible for making sure documentation is complete and accurate. This employee follows up on inconsistencies in the file—yet these individuals are often promoted out of the processor position or burn out and leave within a year.

A solution could be to automate processing to the point where it is exceptionbased, and the system then escalates and re-escalates issues such as incomplete loan information and/or inconsistent values. This would identify common patterns in fraud activities (i.e., increased scrutiny on a [scanned] bank statement if the income values are inconsistent or questionable), which would help eliminate simple oversights that should be red flags. Keep in mind that you also want to preserve an expeditious process—bogging people down with unnecessary checks is one of the primary reasons they start skipping steps. The trick is to push far enough to raise the flags without inundating originators and processors with red tape.

Emerging technologies such as eMortgages and intelligent image documents will also help address this issue, because systems will increasingly be able to validate and perform consistency and integrity checks on the files themselves. It is just too easy to make mistakes when you are looking at hundreds of documents a week (onscreen or not).

But what do we do in the meantime, while we are waiting for the industry to fully embrace eMortgages and other "intelligent" document technologies?

If you speak with industry experts, they will tell you that attacking mortgage fraud often comes down to increased sophistication in your automated workflow. The more of your workflow that is automated to force things to be checked and rechecked and to automatically raise flags to underwriters based on steps being skipped, the farther along you will be in attacking the problem.
For example, instead of manually operating from a printed checklist, you utilize technology to manage it. Most origination systems today have some level of this functionality (process and workflow automation). But are you using it to its fullest potential? Are there other "packaged workflows" that include increased fraud checks (to external databases or on loan data during the origination process)that you could create or add?

There is no doubt that fraud will continue to increase. Given today's market environment and on into the future, the scrutiny of loan files and the quality of data will be of greater focus. I recommend you look at the plethora of databases and tools that are available to you to augment the origination process. Focus on the timeliness and purity of the information being provided, and weigh that with the risk you are willing to assign to red flags raised by these systems.

Second, take a step back to evaluate your business process flows. Think about what exception-based fraud workflow you could add to your processes to take them to the next level in terms of fraud and loan-data checking. But in doing so, make sure not to inundate your personnel with too many steps. The automation you put in place should augment what your employees do, making them more productive, while catching potential red flags earlier in the process.

An example would be to add additional scrutiny to the bank statements submitted for borrowers reporting self-employment income. These statements could also be scrutinized by computer through optical character recognition (OCR) technology, to minimize the impact of adding this step on personnel. You should also properly empower the fraud-checking group within your organization.

There needs to be a balance between the goal of outright production volume and the goals being pursued by people who manage the checks and balances to catch inconsistencies and ask additional questions, when necessary, on loan files. Be mindful of "false positives" and build these into your process model as well. These occur when potential fraud is flagged in a file yet everything turns out to be fine. If the expectations are set properly, this will become part of the process instead of a new area of angst for originators and processors.

Lastly, keep an ear to the ground about what the industry is doing on an ongoing basis about fraud. There are many activities going on in the financial industry on this topic, such as MISMO's Fraud Detection Workgroup. Stay tuned in.