CoreLogic's Case For Improving Property Tax Estimates

By: Jann Swanson •

The Truth-in-Lending/RESPA (TRID) rules implemented in late 2015 have placed increased pressure on the accurate forecasting of estimation of non-mortgage related costs for the loan estimates that must be provided to borrowers.  Dom Lalisse, CoreLogic's Director of Product Management, says this is especially problematic for property tax amounts.  He likens the current methods to the days, pre-standardization of credit scores and credit reports, when the determination of credit worthiness was more art than science and relied on "specialized knowledge and personal interpretation of financial information."

Writing in the February edition of CoreLogic's MarketPulse, Lalisse works to make a case for a change in the system.  He says with the increased scrutiny around the preparation of the TRID Loan Estimate, lenders must increase the quality and consistency of the process they use to project tax amounts.  These estimates also need to be validated during the underwriting stage to qualify the borrower's ability to support all of the housing costs and then finally to complete settlement documents.  He says asking the real estate agent or the borrower for tax information "is analogous to requesting a homeowner to estimate his own credit score.  A consistent procurement process is the only way to ensure access to the best information available."

A standardized version of an estimation process conforming to TRID requirements would follow these steps:

  • Validate address
  • Determine the tax agency or agencies involved;
  • Identify the source of information from each agency
  • Determine Tax Identification for each agency. This is complicated for new construction loans or apartment units where final per-unit allocations haven't been made. Additional steps are needed here.
  • Analyze tax amounts to determine expected future taxes.

This, Lalisse says, demonstrates the complexity of the process and the knowledge and experience required to make estimates into the future.  He estimates that getting reliable information takes about 20 minutes per loan on the part of each of three parties - the loan officer, underwriting, and settlement agent.

It becomes even more challenging in those jurisdictions where legislation caps tax increases for existing residents.  For example, in California, where the 1978 passage of Proposition 13 amended the state constitution to limit annual increases to assessed values.  However, when a property changes ownership or there is new construction then the assessor must reassess the property at current market value.  At a time of large increases in market values, there can also be abnormal changes to the tax.  This is true in other states with similar legislation including Massachusetts, Maryland, Michigan, Texas, and Oklahoma.

He says access to historical data and knowledge of the property tax structure and legislation are critical foundation blocks for new products that will quickly and consistently generate high quality property tax estimates throughout all stages of loan origination.  He appears, in other words, to be suggesting a FICO type product for property taxes.

Such an estimation solution should at a minimum facilitate the following steps:

  • Collection of basis property information by the loan officer
  • A call triggered by technology triggers requesting an estimate also with automatic delivery of the basic information to the data provider
  • Automated address standardization and search of a property database followed by extraction of information.
  • Automated calculation of value-based tax amounts, excluding exemptions.
  • Automated standardized return of information to the user.

Such a process, Lalisse estimates, would reduce the total tax amount estimation process time to no more than 3 minutes.  Such an estimation tool eliminates the need for any specialized skills as well as quality issues and time variations across property complexity.

He concludes that "Originators lacking the processes and controls to help ensure accurate property tax data collection, analysis and estimation, will not be positioned to move forward with those adopting a comprehensive property tax estimated solution as part of their lending process.

We conclude by cynically predicting that CoreLogic will soon announce the availability of such a product under its auspices.  Still, we thought it was an interesting discussion.