djd09
Elite Member
- Joined
- May 20, 2009
- Professional Status
- Licensed Appraiser
- State
- Ohio
‘The Hidden $27 Billion Lending Risk’
White paper details how AI is helping expose flaws in home appraisals
There’s a significant blind spot in the real estate appraisal process — one that could be costing lenders over $27 billion annually in repurchase risk. That’s according to a new white paper from AI-driven computer vision firm Restb.AI, “The Impact of Condition and Quality on Appraisal Accuracy.”
Appraisers traditionally assess a property's condition and quality using the Uniform Appraisal Dataset (UAD) 6-point scale. But that method often leads to clustering of homes in the middle categories — 86.1% of properties were rated as C3 or C4 for condition, and 97.0% as Q3 or Q4 for quality, Restb.AI found.
Such “clustering” can obscure meaningful differences between properties. Further, inconsistencies arise when adjustments are made even when both the subject and comparable properties share the same condition or quality ratings.
The study found that appraisers still applied condition adjustments in 11.8% of such cases and quality adjustments in 5.3%, raising questions about the consistency and transparency of these evaluations.
Financial Implications: A $27 Billion Risk
By analyzing 1,271 appraisals and 6,495 comparable properties, Restb.AI identified that 33.6% of appraisals had a “high risk” of inadequate or missing adjustments, leading to potential repurchase risks exceeding $27 billion annually. What’s more, 73.9% of appraisals exhibited a medium risk, suggesting that the issue is actually more widespread and systemic.
‘The Hidden $27 Billion Lending Risk’
There’s a significant blind spot in the real estate appraisal process — one that could be costing lenders over $27 billion annually in repurchase risk. That’s according to a new white paper from AI-driven computer vision firm Restb.AI, “The Impact of Condition and Quality on Appraisal Accuracy.”
C3 and C4 = Average
