Toll Free
1.855.PURVIEW
Help Center
User ID:
Password:

Automated Valuation Model FAQs

  • What is an AVM?

    An Automated Valuation Model (AVMs) is a mathematically based computer program that produces an estimate of the market value of a residential property based on the analysis of public record data, property location, market conditions and real estate characteristics at a specified date.

    Typically, the street address of the subject property is entered as input into the Automated Valuation Model system. The system in turn identifies the property, checks its location, size and other key features and searches its systems database for the best set of comparables or indices to generate a value for the subject property.

  • Why is an AVM used?

    Advantages to using an AVM include:

    • Inexpensive when compared to appraisals which can cost in excess of $200
    • Time savings: AVMs are generated instantly and in real time therefore significantly decreasing the application cycle time
    • Cost effective: AVMs performed during the front end minimize needless administration expenses
    • Optimize the use of appraisals: AVMs can flag potential risks and help to determine the extent of further due diligence
    • Convenience: AVMs are accessible through the web and are displayed quickly and often available 24 hours a day. Stored configurations and simple design allow for rapid display of results in one convenient report
    • AVMs are objective: No opinion or exceptions are possible. Valuations cannot be swayed or influenced by emotions or personal prejudices
  • What is the role of valuations within the lifecycle of a mortgage?

    • Valuations play a significant role in the lending process by assisting mortgage lending financial institutions throughout different stages within the lifecycle of a mortgage. AVMs are commonly used by some lenders where the loan to value ratio is not high and to support most home equity loans (e.g. line of credit or second mortgage based on existing property owner home equity) and during the preapproval stage, valuations determine the viability of the mortgage application at the time the loan is processed
    • At mortgage origination, AVMs can be used by mortgage brokers to understand the value of a home and can be used to optimize the LTV of a customer’s loan
    • During the collateral adjudication stage, valuations and integrated fraud checks help to assess collateral risk related to the mortgage
    • Post funding, lenders use valuations to provide an ongoing value assessment of their overall lending portfolio (book)
    • During the course of servicing a loan, valuations help a lender to determine the client’s available equity and potential for other home-equity products
  • AVMs or Appraisals?

    • An AVM’s primary advantage over a traditional appraisal is that it saves both time and money
    • An AVM is unbiased versus the human bias which can occur with an appraisal – an AVM cannot commit fraud or make mistakes with calculations
    • An AVM does not replace a full appraisal since an appraisal, unlike an AVM, does take into account the external and internal condition of the property
    • Depending on the situation, an AVM can be used to supplement and/or audit the traditional appraisal process
  • How Accurate are AVMs?

    There are recommended tests for evaluating residential AVMs, these include:

    • Hit Rate: The number of times the AVM will return a result for a given population sample
    • Comparison of AVM Output to the Actual Sale Price: The comparison of the AVM return for the subject property and the recent sale transaction of the subject property
    • AVM Percentage Error: The difference between the AVM return for a sold property and sale price, expressed as a percentage of the sale price
  • What types of AVM models exist?

    An Automated Valuation Model may be based on one model or the combined outputs of multiple models. The most common types include:

    • Price Indices: Multiple repeat sales are used to create and establish house price indices for a specific geographic area. This index is then applied to a past transaction price or valuation of the subject property to provide a current valuation
    • Hedonic: Largely based on statistical models using some form of linear regression. Property attributes such as location, property size, and nature of improvements are data requirements of this model type. Essentially, the attributes of a subject property are compared with other comparable properties using a radius search pattern or other logical search parameters, over a pre-determined time period
    • Tax Assessed Value Model: The Current Market Value is estimated by updating the valuation assessed for tax purposes at a past date. The statistical relationship between past assessed values and subsequent price data is measured to create a ratio which is then applied to update the assessed values