Advances Continue To Reduce Risks While Cutting Costs And Increasing Speed

    0

    Secondary Marketing Executive, February 2006.

    In the mid-1990s, collateral valuation practices in the secondary market included random selection of loan files in a pool for re-evaluation of their original appraisals.

    Frequently, re-evaluations were supported by drive-by appraisals, but broker price opinions (BPOs) and traditional field reviews of the appraisals were also utilized. All these traditional appraisal methods required an industry professional to physically take a look at the subject properties under review.

    As a result, the high cost and turnaround constraints associated with these appraisal reviews resulted in the re-evaluation of only a small sampling of files. At that time, it usually took five to seven working days to receive results of a traditional valuation, and the cost ranged from $175 to $225 on average. Consequently, secondary marketing professionals buying large pools of loans could only re-evaluate collateral values for 10% to 20% of those files.

    Time for a change
    While a minimal number of appraisal re-evaluations were being performed in the mid-1990s, secondary marketing professionals were reviewing borrower credit more heavily. Largely driven by the increased credit risk associated with a rapidly expanding subprime market, it became customary for up to 100% of loan pools to go through a reunderwriting process to confirm borrower credit. This measure was seen as a critical step to defining and minimizing risk associated with a pool.

    Still, despite aggressive credit re-underwriting efforts prior to loan pool purchases, secondary marketing professionals experienced higher levels of default and foreclosures with subprime loans compared to prime loans. While an increased level of foreclosures was expected in the subprime market, what secondary marketing professionals did not expect were the issues that became evident on the collateral side.

    As foreclosures were analyzed, secondary marketing participants recognized that many of the thousands of files that did not get selected for the random appraisal re-evaluations indeed reflected overstated collateral values.

    When appraisals were overstated at origination, properties went into foreclosure more often and their losses were significantly larger. Secondary marketing executives quickly made the connection between the low number of files re-evaluated for collateral values and the increased level of losses in the subprime market. Investors were losing more than they expected, and as a result, focus on the collateral valuation problem intensified.

    It became clear that collateral value had as strong of an influence on bottom line losses as borrower creditworthiness. This new focus called for a different method or process that would enable every single file in a loan pool to be checked from a collateral standpoint.

    First waves of change
    Stemming from the need to find a better way to identify the underlying risk in collateral values, technology and data advancements were combined with traditional review processes around 1998 to create a more efficient appraisal re-evaluation process.

    This hybrid process still depended on review appraisers, but their primary role shifted to evaluating electronic public records information, like comparable sales, along with proprietary information, such as appraiser performance data. Review appraisers used this information specifically to identify risk indicators within appraisals.

    In this process, the primary focus centered on these risk indicators, and each re-evaluated appraisal was assigned a collateral score to help quantify each file's level of risk. Rather than go through the expensive process of a physical evaluation of each property, this process used the review appraiser's professional expertise to determine whether a specified appraisal should be placed in a low-, moderate- or high-risk category.

    Along with collateral scoring, this process also allowed secondary marketing executives to decide which original appraisals should be further scrutinized. Actions could range from accepting a loan when it was categorized as low risk, to completing a traditional drive-by appraisal or BPO for high-risk appraisals.

    By re-evaluating appraisals for risk indicators only and including professional recommendations, the overall process for appraisal re-evaluations was greatly streamlined. Before, each appraiser could only complete approximately 15 appraisal reviews a day. Using this risk indicator review process, each appraiser could increase daily production threefold.

    Clearly, the increase in review production minimized concerns regarding turnaround time, and the process also addressed cost issues, since risk indicator reviews averaged just $25 to $30 each. With time and cost constraints out of the way, secondary marketing professionals could feasibly review up to 100% of the collateral in loan portfolios being considered for purchase. They would also have remaining funds to implement recommendations provided in conjunction with the collateral scores.

    Around the mid-1990s, automated valuation models (AVMs) were becoming more reliable and starting to be used in tandem with other re-evaluation methods. However, while AVMs provided a quick turnaround time, the quality of data was not as robust as it needed to be for widespread use.

    For properties in some parts of the country, there might be no information available or only partial information for specific areas. Also, the underlying statistical models that powered AVMs were still being refined, and that fact discouraged many from relying solely on AVM results. Consequently, while AVM technology made great strides in the early to mid-1990s, it was still not customarily deployed in the secondary market.

    However, by the late 1990s, more county data became available to AVMs and the quality or detail of that data was improving, though still not perfected. At the same time, the statistical models and production methods associated with AVMs became more sophisticated.

    As a result, AVM usage became more widespread in the secondary market, providing another cost-efficient solution for re-evaluating the appraisals in a loan portfolio. Yet, while this totally automated alternative provided the quickest results, the likelihood of getting solid results for 100% of a portfolio was still limited by a lack of data in some markets.

    Due to deficiencies with AVMs, another hybrid solution was developed in the late 1990s to leverage the strengths of AVMs, while offsetting some of their limitations. This solution combined AVM technology with public records information, appraiser performance data, and flip/fraud indicators.

    With this enhanced solution, a client was able to preview 100% of the files in a loan portfolio with an automated collateral score. In addition to receiving a value from an AVM, the approach provided a collateral score as well as detailed information on the subject property and surrounding neighborhood. In essence, this model was a totally automated collateral scoring solution.

    The next level
    Around 2000, solutions that competed directly with traditional field valuations began to emerge. These solutions could be easily deployed when appraisals were categorized as high risk by a collateral risk score and required further due diligence.

    One such solution combined the best of several proven methods by taking property information gleaned from a field inspection and incorporating that data into a valuation model.

    In this scenario, a BPO field inspection – less expensive than a drive-by appraisal, but containing less detail – was completed. However, by combining the information from a BPO with the appraisal logic and valuation methodology available through a modified valuation model, this standardized report rivaled drive-by inspections in both quality and cost.

    But what if a hybrid solution, drive-by appraisal or BPO fails to resolve risk questions about the collateral value of a property? In this case, a third phase of the appraisal re-evaluation process begins. This phase, referred to as the tie-out, begins when the potential buyer of a pool of loans sends the high-risk files back to the seller. The initial return of these loans gives the seller the opportunity to justify the original value before the potential purchaser requests they be removed from the pool.

    Prior to 2000, most secondary marketing professionals handled the tie-out process themselves. Around 2000, some providers of advanced valuations services began to offer third-party valuation tie-out services. These providers facilitated the reconciliation between the original appraisal value and the values derived from the re-evaluation approaches.

    The newest generation of valuation solutions nearly replaces the need for a drive-by appraisal or BPO at any stage in the valuation process. One solution involves an appraiser who evaluates a variety of information, including the risk indicators of an original appraisal, additional public and proprietary data, and AVM results.

    This comprehensive depth of information enables the review appraiser to develop a new value that can be compared to the original appraisal. As with other solutions that identify risk indicators, this solution also includes recommendations based on the final risk categorization.

    One of the latest valuation solutions also reduces the number of appraisals that have to be sent for a drive-by appraisal or BPO. This solution combines local sales comparable market data with a physical inspection of the subject property obtained via helicopter or drive-by photograph. An appraiser evaluates that combination of information and concludes by assigning a value.

    Changes to come
    In the valuation arena, the data that is available within current offerings continues to expand. In addition, data from other resources, such as Multiple Listing Service (MLS) platforms, will be directly integrated with valuation solutions, enhancing the quantity of data available for comparison.

    MLS data will provide sales prices that can be used for comparables in areas where no information was previously available from non-disclosure states or rural areas. Additionally, more data will be incorporated through analytical models into the valuation process, providing forward-looking valuation scores designed to improve the accuracy of current values.

    The driving force behind current valuation solutions, as well as those coming in the future, continues to be the ongoing need to minimize collateral risk for secondary market investors while deploying the quickest, most inexpensive means possible.

    By streamlining the review process so that fewer original appraisals require a costly field valuation process, collateral valuation solutions have made a tremendous difference in the quality of due diligence results. And as technology and data continue to improve, valuation solutions are expected to become even more sophisticated.

    As these products gain acceptance in the secondary market, originators are able to employ streamlined collateral scoring products to reduce costs, reduce turnaround times, and obtain better execution in the secondary market. More and more originators are able to drive these nontraditional products to the forefront of the origination process, which results in their organizations producing higher-quality loans, stronger valuation performance, and a more effective transition into the secondary market.

    Greg Hansen, president of Fidelity Hansen Quality, founded Hansen Quality in 1993 to provide appraisal review services. Since its merger with Fidelity National Financial, the company has focused on collateral risk assessment. Jon Davis is executive managing director of the Fidelity Hansen Quality Group. As senior vice president of product development from 1998 to 2002, Davis headed the Hansen Quality product development team, which created the next generation of property valuation products and services.

    Subscribe
    Notify of
    guest
    0 Comments
    newest
    oldest most voted
    Inline Feedbacks
    View all comments