Dec 01, 2006
Top mortgage lenders were the first to incorporate automated valuation models (AVMs) into their business processes in order to take advantage of the cost and time efficiencies that AVMs offer over traditional appraisals. Today, increasing numbers of smaller lenders are adopting AVM technology into their operations as well, presumably to enjoy the same benefits as, and better compete with, the larger lenders.
There are two major benefits to incorporating AVMs into the lending process:
For lenders, this compares favorably to the traditional appraiser environment, in which the preparer of the report is often subject to transactional pressures, and there is no way for users to test the statistical accuracy of the valuation. On the other hand, AVMs are not available for all properties and must be continuously tested and adjusted by AVM providers to ensure optimal performance.
Business Rules
Most lenders establish underwriting rules that determine how a property valuation decision will “cascade” through a predefined set of choices until an acceptable result is achieved. These business rules are unique to each lender, and rule enforcement can be automated through a collateral risk management system or manually executed through careful training and ongoing supervision.
One frequently overlooked requirement in establishing an AVM cascade is the need to establish a definition of minimum valuation accuracy, below which an AVM will not be invoked. Failure to establish minimal standards of valuation accuracy will result in AVM searches that have little or no probability of exceeding your minimal valuation standards.
While every lender needs to establish a reasonable variance tolerance for valuation performance according to its own lending guidelines, industry literature suggests that a variance of plus or minus six-to-twelve percent is acceptable. Once the variance tolerance has been determined, a lender may use one of several methods to evaluate AVM performance efficacy and build a cascade.
Evaluation Methods
One method for evaluating AVM performance is through the concept of “net efficacy.” Net efficacy looks at the percentage of observations that are accurate (for instance, plus or minus ten percent of a reference value) and subtracts the percentage of observations that are grossly inaccurate (for example, variances that are more than 25 percent too high, too low or both) to arrive at a scale for AVM effectiveness. Another method for evaluating AVM performance revolves around the concept of “centrality of the estimates,” where a lender is more concerned with median error or absolute median error than with net efficacy.
Adverse Selection
Whatever the definition of AVM performance, one fact remains: AVMs in the second and third position of a cascade will not perform nearly as well the AVM in the first position because of “adverse selection.” Adverse selection means that the AVMs in the second and third tiers are subject to the most difficult properties to value. In other words, the AVM in position three is less likely to achieve minimum valuation accuracy standards. In this situation it is helpful to understand how your tier-three AVMs are performing by county, property type and other variables. From there you can customize your AVM selection rules to avoid excessive no-hits.
Bottom Line
The answer to the question, “How deep should your cascade go,” is that it should go as deep as it can while still delivering your minimum acceptable performance. If your minimum acceptable criterion for AVM use is a median error of 15 percent and you are not achieving that range with three tiers, then you should consider ending your cascade after the second tier. After all, subsequent AVM use will simply not achieve your minimum AVM accuracy standards and will add unnecessary complexity and cost.
For further information please contact Robert L. Walker, CMB, CMT, executive vice president collateral solutions for First American Real Estate Solutions at 714-250-6684; or by email at robwalker@firstam.com.
About First American Real Estate Solutions
First American Real Estate Solutions is a member of The First American Family of Companies and America’s largest provider of advanced property and ownership information, analytics and services. First American RES’ databases cover more than 2,900 counties representing 99 percent of the United States population. With more than 600,000 users nationwide, First American RES products are used by companies to improve customer acquisition and retention, detect and prevent fraud, improve mortgage transaction cycle time and cost efficiency, measure the value of residential and commercial properties, identify real estate trends and neighborhood characteristics, track market performance and increase market share. First American RES is a joint-venture company 80-percent-owned by The First American Corporation and 20-percent-owned by Experian. More information about First American RES can be found on the Internet at www.firstamres.com.
Back to top