FNMA Big Data to Check Appraisals
FNMA New Collateral Underwriter to Check Appraisals Using Fannie’s Big Data
FNMA Collateral Underwriter Flyer showing info about the FNMA Collateral Underwriter process they will make available to lenders (NOT APPRAISERS) in January 2015. You should review it. It has to do with their Enhancement of Risk Controls.
This is what we know as Appraisal Quality Monitoring (AQM), which was announced almost 2 years ago. FNMA has already been using the ‘scope’ on your reports, but will soon allow the lenders to have access to the software so that they can do pre-submittal exams prior to uploading the loan file, and your appraisal, to FNMA.
Virtually everything is digital now in our real estate appraisal world. That makes it incredibly easy for ‘big data’ to be analyzed very quickly and efficiently. Hiding relevant property info under a rock, your clipboard [tablet?], or just ignoring it, is no longer possible. Discrepancies will be found fast, and you will be asked for explanations or corrections.
Note the examples from the flyer:
- Chain of property ownership
- Inconsistency in reported property data from your info compared to your peers (subject & comps)
- Checking adjustments made (or not made) – primarily the math
- Testing for comps in terms of location, characteristics, sales prices, etc.
FNMA’s news release about their Collateral Underwriter:
Introducing Collateral Underwriter
Collateral Underwriter™ (CU™) is a proprietary appraisal risk assessment application developed by Fannie Mae to support proactive management of appraisal quality. CU will:
- Provide additional transparency and certainty by giving lenders access to the same appraisal analytics used in Fannie Mae’s quality control process.
- Perform an automated risk assessment of appraisals submitted to the Uniform
- Collateral Data Portal® (UCDP®) and return a CU risk score, flags, and messages to the submitting lender.
- Be available at no charge so lenders can take full advantage of the application for quality control and risk management purposes.
The CU risk scores, flags, and messages will be available to all UCDP users in real-time beginning on Jan. 26, 2015 through UCDP. Find more information on the CU web page at FNMA Collateral Underwriter.
- Speed Regardless of Accuracy Under the Banner of Modernization - March 8, 2023
- Marin City Discrimination Case Settled - March 7, 2023
- Exactly How Are Property Data Collectors Professionally Trained? - March 3, 2023
I have no objection to legitimate internal checks and balances by any client.
What I DO have concerns about is being asked / directed to reconcile MY property data to some unknown “peers” data without having a full and complete copy of that “peers” data (appraisal report), so that I can see and understand the context in which that data was reported as opposed to my own.
There is a reason why appraisal review has its own specific compliance standard rule under USPAP. This proprietary CU does not conform to that USPAP Standard Rule (SR 3).
With respect to ‘Chain of Title issues’ THAT is not my responsibility. That is what title insurance is for. My ‘scope’ in this regard is researching within the normal course of business, as would be customary by my peers. Different commercial reporting services sometime provide different transactional histories.
Simple refinance transactions within a family can show up as “sales” due to title company recommended use of grant deeds versus quit claims, when a stronger credit family member is put on title to better qualify for a loan. Again, NOT my job to analyze prior refinances lending and title issues.
Checking adjustments. IF it is ONLY the math, I have no problem with it. IF it becomes a metric for measuring the propriety of an adjustment, or the amount of an adjustment, then FNMA first needs to demonstrate that THEIR software is a reliable market value indicator in the specific market being considered.
Checking comp applicability is a review appraiser function. It is NOT a software scoring function. Too may times appraisers have been told comps “closer” than those used should have been selected based only on proximity, when it turn out those comps were outside city (sub market area) bounds!
IF ‘big picture” databases were reliable, then all one would need is the national median price information attached to a local size related multiplier, in order to arrive at “value”.
FNMA and software gurus have never accepted that this cannot be done and keep reworking the software and its descriptions in an effort to achieve that which we already know is impossible.
I Understand that this Process is confidential, “… FNMA Collateral Underwriter Flyer showing info about the FNMA Collateral Underwriter process they will make available to lenders (NOT APPRAISERS)…” Why not? If their info is so extensive, so complex, so enigmatic why not run the analysis and give it to the appraiser up front? We’re in the age of sharing data open source freely, right? Why not? The assignment would read something like, “…we’ve run our analysis (by a bunch of high powered specialized statisticians not one of which knows everything there is to know about the process) and this is what we found out — now go out, take a look at the subject, the interior, the condition, the area, the nearby properties etc. and tell us what you think… ” A simple, direct, no nonsense approach where everybody that has an interest in the accuracy of the result contributes their bit. Why all the inane secrecy – unless there is a hidden agenda somewhere??