Enter your e-mail address and password below.


Forgot your password? New User? Register Now.
   ARTICLE   |   From Scotsman Guide Residential Edition   |   October 2005

Fighting Fraud

A combination of collateral assessment and AVMs can maximize mortgage-fraud management

r_2005-10_Schroefer_spotFraud is a growing issue, and the FBI’s “Financial Crimes Report to the Public” this past May revealed that the number of filed mortgage-fraud reports escalated from 6,936 in 2003 to 17,127 in 2004. California, Nevada, Utah, Colorado, Missouri, Illinois, Georgia, South Carolina and Michigan reported the most cases, according to the report, which also showed that 80 percent of the cases involved either overstated property appraisals or claimed nonexistent properties.

As the problem of overstated appraisals increases and attracts more attention within and outside the industry, more-accurate property valuation through automated systems is becoming central to the profitable management of mortgage banking. The role collateral-risk tools and automated valuation models (AVMs) play and the value they contribute to lending organizations are growing as more lenders realize the financial impact of avoided loss exposure.

AVMs are valuable tools for loan underwriting, but their primary function is to determine property value, not to catch appraisal fraud. While AVMs effectively streamline processes and lower costs, mortgage-underwriting fraud deserves special attention and treatment with specialized detection tools.

One strategy that a growing number of mortgage originators are using is to evaluate every loan’s potential for foreclosure with a collateral-risk tool at the front end of the lending workflow. This occurs before verifying the property appraisal with an AVM. Rather than randomly selecting loans for review or exhaustively investigating every loan with compliance tools, collateral-risk tools efficiently examine every loan’s agent, borrower and collateral history from the start.

First line of defense

No AVM alone can provide an adequate level of fraud management. An AVM should be used primarily in a second-tier screening for value verification, while the first tier includes an integrated assessment of the agent, borrower and collateral.

If the first-tier results appear acceptable in the risk-score output, then the loan should be streamlined, bypassing the need for a more-detailed investigation. If the collateral-risk tool determines that the property, borrower and/or agent scores exceed predefined limits, it is wise to consider further due diligence using an AVM.

AVMs are objective tools used to estimate a property’s value based on a variety of factors. They consider tax data, comparable sales records and geographical data in calculating the estimation. By removing human touch from the system, the opportunity to commit fraud is significantly diminished. The valuation is not a grade on the loan’s tendency for foreclosure, however. Appraisers and lenders are tempted to overstate estimates to increase their bottom line.

According to their April 2005 article in Secondary Marketing Executive, Michael G. Bradley and Mark A. Beardsell say that appraisers usually work as independent agents and collect fees. “The lender often resells the loan and may, therefore, not be overly concerned about collateral risk.”

Recently, this incentive has led to transaction bias. This occurs when appraisers and lenders accept inflated estimates to complete a transaction. From their study comparing the collateral protection of full appraisals and AVMs, Bradley and Beardsell say that AVM values tend to be less than appraised values for purchase and refinance transactions and in soft and not-soft markets. “Full appraisals provide relatively worse credit-risk protection in soft markets, a finding that is entirely consistent with the presence of transaction bias and the associated incentive incompatibilities appraisers face today.” In short, AVMs are immune to pressures that affect full appraisals, such as transaction bias.

In its June 2005 white paper “Systemic Risks in Residential Property Valuations,” the Collateral Assessment and Technologies Committee describes the benefits of AVMs: “High quality AVM systems are proven solutions that are capable of bringing incremental objectivity, consistency, and controls to collateral valuation and risk assessment processes in our evolved and modern marketplace. They can do so in an efficient and cost-effective fashion that even the best-intended appraisal industry reforms or new regulations may be hard-pressed to match.”

AVMs are not enough

Although there are many benefits to using AVMs, the methods also have weaknesses. AVMs have difficulty distinguishing fraudulent or fabricated markets. A fabricated market is created when an investor purchases a large number of properties (25 to 30) in an area and then flips them privately. After about six months, the area becomes its own market, and the AVM accepts the flipped values as accurate, though they are overvalued.

Initial collateral-risk assessments are necessary to prevent fraud with each loan because AVMs are vulnerable to fabricated markets. Scoring the loan’s foreclosure potential before using the AVM can avoid an inaccurate valuation of the property, should it be located in a fraudulent market. If an AVM is used upfront, however, it can worsen the problem of inflated appraisals by valuing the property according to overstated values on comparable properties.

As quoted in American Banker in August 2004, Donald Kelly, vice president of public affairs for the Appraisal Institute, describes this weakness of AVMs: “When [appraisers] see something like the algorithms in some AVMs, they recognize the weakness surrounding the availability of real data and unverified data used to come up with the ultimate value.”

Effective but not economical

Compliance tools are effective for catching fraud committed to qualify for housing because they identify misrepresentations on the part of the borrower and, in some cases, the broker. But they are inefficient for identifying fraudulent activities such as flipping.

As a compliance tool examines a growing number of criteria, more loans with high fraud-for-profit risk will be identified. As criteria thresholds increase, however, a greater number of easily correctable misrepresentations are caught that do not necessarily affect loss severity significantly.

For example, loans may be denied that could have been corrected and accepted easily. Unintentional misrepresentations, such as typographical errors, should be corrected, and the loan may be accepted rather than selected for additional investigation. In addition, the administrative cost associated with the breadth of this type of investigation is prohibitive.

The need for collateral-risk tools

Before using an AVM or a compliance tool, mortgage originators and investors should evaluate a loan’s fraud potential by analyzing the agent, the borrower and the collateral. They then should calculate an overall risk score based on a broad set of data, including multiple geographic conditions surrounding the subject property, the borrower’s financial and behavioral patterns and the agent’s production history.

After scoring the loan’s propensity to default, it then should fall into one of three categories. The first category, loans with no foreseeable problems, can be streamlined for processing. The second category, loans with some potential issues, requires more-careful evaluation, but these loans are likely acceptable. The third category, loans with high potential risk, requires further due diligence.

Automated-risk-management tools are now available to evaluate and score the borrower history, patterns, surrounding-area and subject-property characteristics and the agent or money channel itself. Typically, these front-end tools expedite workflow by collapsing a loan’s cycle time in half while saving on labor costs and increasing the company’s production capacity. 

•  •  •

Despite the effectiveness and efficiency offered by the combined use of collateral-risk tools and AVMs on the front end of loan production, this approach is only an interim solution to the growing epidemic of mortgage fraud. Fraud perpetrators are discovering new ways to beat the system, as software companies that serve mortgage originators and investors work vigorously beside their customers to stay ahead of the fraud curve.

Only through this type of ongoing innovation will the mortgage-banking industry provide the necessary line of defense to ensure a bright and profitable future.


Fins A Lender Post a Loan
Residential Find a Lender Commercial Find a Lender
Scotsman Guide Digital Magazine

Related Articles



© 2019 Scotsman Guide Media. All Rights Reserved.  Terms of Use  |  Privacy Policy