Unintended consequence of reviews

[DWiley]

Yes, it is aggravating to watch appraiser be scapegoats for the unprofessional and shortcut practices of others.

In a previous post you referred to "donkey" appraisers. Well guess what those "donkey" appraisers are? They are appraisers - appraisers who exhibit "unprofessional and shortcut practices" - and they are out there competing with the appraisers who do things correctly. And, more importantly, they are affecting the public perception of the value of appraisal practice.

Are those "donkey" appraisers just scapegoats? Or are they the real root of the problem for other appraisers?

Unprofessional conduct by others should indeed be called out. But the fact that their conduct is all too often facilitated by unprofessional conduct by an appraiser should not be ignored, as is often the case. There is plenty of blame to go around, but some in this community seems to only butter one side of the bread.

 

[alebrewer]

That’s a due diligence failure on the lender for contracting with those appraisers, or for contracting with those AMCs who contract those appraisers.

Man, in your world appraisers aren't responsible for anything. It's the big, bad lenders that are responsible for all the flaws in appraisals, right?

 

[sputnam]

I disagree with this. When the purpose of the appraisal is to assist the lender with a mortgage finance transaction, unintended users should have no input. That includes borrowers, agents, whomever.

If the borrower has an issue/question about the lender’s process about appraiser engagement, that should take place before the borrower signs with the lender.

I agree. The homeowner or buyer has the right to choose which lender they want to do business with... not the right to help choose the appraiser.

 

[RCA]

In a previous post you referred to "donkey" appraisers. Well guess what those "donkey" appraisers are? They are appraisers - appraisers who exhibit "unprofessional and shortcut practices" - and they are out there competing with the appraisers who do things correctly. And, more importantly, they are affecting the public perception of the value of appraisal practice.

Are those "donkey" appraisers just scapegoats? Or are they the real root of the problem for other appraisers?

Unprofessional conduct by others should indeed be called out. But the fact that their conduct is all too often facilitated by unprofessional conduct by an appraiser should not be ignored, as is often the case. There is plenty of blame to go around, but this community seems to only butter one side of the bread.

It's always been my opinion that ALL appraisal orders should be distributed on a rotational basis by a government entity from a panel of qualified appraisers. It should be a State entity, most likely the State Appraisal Board.

So I discussed with ChatGPT:

"...5.
The Core Dilemma
You have put your finger on the structural conflict of interest that still troubles the industry:
We let the party who benefits from a higher valuation choose the professional who is supposed to be independent.
The current system mitigates but does not fully eliminate that conflict.
Shifting all selection to a state board by rotation would remove that bias but runs up against federal statutory structure and concerns about competition and property-rights.

---

In Summary​

You are correct: lender choice creates inherent bias.
Existing independence rules restrain it but do not always prevent it.
A state-level rotational system would be cleaner from a neutrality perspective, but it would require a major federal-level change in how appraisal independence is conceived and enforced, and it would have to be designed to survive antitrust and due-process challenges."

"Would you like me to outline in more detail how a clearinghouse model might be governed (funding, data handling, oversight) so that it could credibly reduce lender pressure while remaining legally viable?"

​

Me: "Yes"

 

[RCA]

ChatGPT(1): "
Bert,
below is a concrete, legally cautious design for a neutral appraisal-order clearinghouse that reduces lender steering while preserving legitimate competency selection and avoiding antitrust pitfalls. I will treat this as a product + policy specification you could take to a legislature, regulator, or pilot consortium.

---

1) Mission and Scope​

Mission.
Allocate residential (and optionally commercial) appraisal assignments through a neutral conduit that (a) enforces independence, (b) documents an auditable trail, and (c) gives lenders bona-fide competency controls without covert value-shopping.

Scope (initial).

• Federally related 1–4 family transactions up to a complexity limit (expand later).
• Opt-in for lenders and AMCs; mandatory only if state statute subsequently requires.
• State-chartered, privately operated (public-benefit) utility under active state supervision.

---

2) Governance & Legal Guardrails​

Entity form.

• State-chartered public-benefit corporation (or nonprofit) with a statutory mandate narrowly tailored to independence and transparency—not price setting or market allocation beyond the matching function.

Board composition.

• 9–11 directors: 2 state officials (banking + consumer protection), 2 licensed appraisers, 2 lender risk/compliance officers, 1 AMC compliance, 2 public members, 1 data/AI ethicist.
• Recusal/COI rules; open meetings for policy decisions.

Active supervision (antitrust “state-action” doctrine).

• The authorizing statute must clearly articulate the public policy (reduce coercion, ensure independence, preserve competition) and require ongoing, substantive supervision by a state agency (periodic rule approvals, audits, power to veto/modify matching rules).
• No fee-fixing; no barriers to entry beyond licensure/competence; no exclusive dealing.

Separation of powers.

• The state appraiser licensing board continues licensing/discipline.
• The clearinghouse does not regulate; it only matches and logs.

---

3) Participation & Eligibility​

Appraisers.

• Must hold active state license/cert; in good standing; E&O on file.
• Maintain a structured competency profile (geography, property types, complexity tiers, languages, VA/FHA approvals, measurement standards, desktop/field capacity, turn-time bands, tech stack).

Lenders/AMCs.

• Execute participation agreement; submit objective filters per assignment (no named appraisers; competency/coverage only).
• Agree to transparency and audit provisions.

---

4) Assignment Taxonomy​

A shared schema ensures “apples to apples” matching:

• Geography: county; sub-market polygons; rural/urban flag.
• Property: SFR/Condo/2–4; manufactured; waterfront; luxury; acreage; ADU; complex features.
• Purpose: purchase/refi/HELOC; new construction; litigation/REO (if included).
• Product: 1004/1073/2055/hybrid/desktop; GSE overlays; FHA/VA.
• Constraints: target SLA window; borrower language needs; inspection constraints.
• Risk tier: lender-internal (Low/Mod/High) with definitions on file for audit.

---

5) Matching & Rotation Logic​

Core principles.

1. Lender sets filters, not names.
2. Clearinghouse selects from the eligible pool via transparent rotation with traceable exceptions.
3. Complex jobs weight to higher-tier competencies; simple jobs rotate broadly.

Algorithm (outline).

1. Build EligibleSet = Appraisers meeting all required filters (license, geo, product, complexity, approvals, SLA window).
2. Exclude temporarily unavailable or conflicted appraisers (disclosure flags; relationships with interested parties).
3. For the remaining pool, compute a priority score:
   • Base = inverse of recent share (to keep load fair).
   • Adjustments:
     • Competency depth score (experience hours, relevant samples) for complex assignments.
     • Past quality index (dispute rate, QC outcomes) within narrow bounds (to avoid covert favoritism).
     • – SLA risk penalty if recent late deliveries exceed threshold.
     • Never include fee level in the score (see §6).
4. Select the highest-priority appraiser; if they decline/timeout, next in order.
5. Log every decision factor for audit.

Pseudocode (simplified).

EligibleSet := Filter(Appraisers, Meets(Competency & Geo & Product & SLA & Approvals))
EligibleSet := Remove(EligibleSet, Conflicts OR Unavailable)

for a in EligibleSet:
base = 1 / (1 + RecentAssignmentsShare[a, rolling=90d])
adj = f(ComplexityWeight * CompetencyDepth[a])
+ f(QualityIndex[a], capped)
- f(SLAOverage[a], capped)
score[a] = base + adj

winner := argmax(score)
Assign(winner); Log(request, filters, EligibleSet, scores, winner)
"
(continued)

 

[RCA]

(continued)
ChatGPT: "

Graceful handling.

• Complexity escalation: if EligibleSet < N_min or scores cluster poorly, escalate to Specialist Pool review (human-in-the-loop) with reason logged.
• Rural scarcity: allow radius expansion and neighboring-market substitutions with borrower travel disclosures.

---

6) Fees, Bids, and No Price-Fixing​

To avoid antitrust issues and preserve market signals:

• Appraisers post dynamic fee bands by product/complexity/geo; lenders see bands only after eligibility is determined (no cherry-picking by price first).
• Matching occurs regardless of fee; price negotiation happens post-tentative match within posted bands (or lender standard schedules).
• Alternatively, support sealed short-bid windows for commoditized desktops with strict anti-race-to-the-bottom floors (e.g., below-cost flags).
• All fee interactions are logged; the clearinghouse does not set prices.

---

7) Independence, Conflicts, and Reconsiderations​

• Conflict checks: cross-reference appraiser disclosures (prior service on the subject within X months; relationships with parties; ownership interests).
• Reconsideration of Value (ROV): standardized portal for evidence uploads; anonymized routing to the appraiser; second-opinion policy for material factual errors; full audit trail.
• Second appraisals: permitted only under predefined triggers (QC fail; new scope; investor requirement)—not for value shopping. Every second order is flagged.

---

8) Transparency & Auditability​

Public dashboards (aggregate, privacy-safe).

• Assignment counts by county, product, complexity tier.
• Average turn-time, QC dispute rates, ROV frequency.
• Distribution fairness indices (Gini/Herfindahl across qualified pools).

Regulatory audit pack (access-controlled).

• Full per-order decision trail (filters, eligible list, scores, reasons for exception, timestamps).
• Periodic fairness tests (no systematic under-assignment to protected geographies or cohorts).
• Lender-specific steering flags (e.g., abnormal ROV rates, high second-order incidence).

---

9) Data, Privacy, and Security​

• Security: SOC 2 Type II; encryption in transit/at rest; least-privilege access; immutable logs (WORM).
• Data minimization: no borrower PII beyond what is essential; hashed identifiers; retention schedules aligned with banking regs.
• Explainability: each assignment produces a machine-readable Decision Card describing inputs and the rotation outcome.

---

10) Coexistence with AMCs​

• AMCs may use the clearinghouse as their order router; they retain post-match file processing, scheduling, QC, and communication functions.
• The system levels the playing field on selection while letting AMCs compete on service, analytics, borrower experience, and post-selection workflow.

---

11) Funding Model​

• Per-transaction clearing fee (small, capped) paid by the ordering party.
• Tiered membership for lenders/AMCs (API volume, analytics).
• Grant/appropriation support in the pilot phase to prevent fee shocks.
• Absolutely no revenue share with appraisers (avoids kickback optics).

---

12) Metrics & Compliance KPIs​

• Independence: % of orders with named-appraiser requests (should be ~0 except bona-fide specialty); second-appraisal rate; ROV rate by lender and product.
• Fairness: assignment distribution vs. eligible supply; time-to-first-offer; declination reasons.
• Quality: post-close defect rates; investor repurchase triggers tied to valuation.
• Service: median/95th percentile turn-times; borrower complaint rates.

13) Implementation Roadmap (12–18 months)​

1. Statute & rulemaking (0–6 mo).
   • Mandate; supervision structure; data-sharing safe harbors; privacy.
2. MVP build (3–9 mo).
   • Eligibility registry; filters; rotation engine; audit log; lender/appraiser portals; pilot APIs.
3. Pilot (6–12 mo).
   • 3 lenders + 1 AMC; 3 diverse counties (urban, suburban, rural); desktop + 1004.
   • Independent evaluation partner (university/think-tank).
4. Scale-up (12–18 mo).
   • Add FHA/VA; expand geos; introduce sealed short-bid option; publish first annual transparency report.

14) Risk Register & Mitigations​

• Antitrust challenge: tightly drafted statute; no fee-setting; open eligibility; active supervision; periodic independent competition audits.
• Under-matching specialists: competency tiers + human escalation + curated specialist pools.
• Data breach: SOC 2, red-team tests, incident response playbooks.
• Lender pushback: preserve competency filters; opt-in pilot; demonstrate lower repurchase/defect risk via metrics.
• Small-shop burden: simple onboarding; free profile assistance; standard data connectors.

15) Opinionated Bottom Line​

• A pure state-board rotation is unlikely to survive the combination of banking-regulator expectations, antitrust scrutiny, and operational complexity.
• A state-chartered, actively supervised, neutrality-first clearinghouse—with lenders limited to objective filters and with full transparency/audit trails—can materially reduce steering pressure while keeping lenders accountable for competency and credit risk.
• The key is narrowly tailored authority: rotate among the eligible (objectively defined), log everything, and keep pricing and post-match workflow out of the clearinghouse’s hands."

 

[BRCJR]

If reviews by non-appraisers and appraisers are both called Appraisal review in USPAP or called appraisal review and review, the terms are still confusing and misleading regarding differentiating between the two versions. There should be a clearly distinct name for a technical /clerical "review " of an appraisal, especially it can be performed by non-appraisers.

Doesn't the SOW take care of your concern?

 

[RCA]

ChatGPT Legislative Brief Proposal:

Legislative Brief (2 pages)​

Title​

Establishing a State-Chartered, Actively Supervised Appraisal-Order Clearinghouse to Reduce Steering and Strengthen Independence

Purpose​

Create a neutral, auditable conduit for allocating appraisal orders among qualified, licensed appraisers using objective competency filters and transparent rotation—thereby reducing lender steering and “value shopping” while preserving lenders’ federal duty to select competent appraisers.

Problem Statement​

Current market practice allows lenders/AMCs to select specific appraisers. Incentives to close loans can bias selection toward pliable appraisers or those who reliably “make the number,” despite federal and state independence rules. Enforcement is slow and largely retrospective. Appraisers fear removal from panels if they resist pressure. Borrowers and investors lack a transparent audit trail of how the appraiser was chosen.

Policy Solution​

Authorize a state-chartered, privately operated, public-benefit clearinghouse to route appraisal orders by objective filters (competency, geography, product, accreditation) and transparent rotation among the eligible pool. The clearinghouse is not a regulator and does not set prices; it operates under active state supervision to qualify for state-action antitrust immunity and to protect competition.

Key Features​

1. Active State Supervision
   • Statute clearly articulates public policy (independence, transparency, consumer protection, competition).
   • Supervising agency (e.g., Department of Financial Protection) approves rules, audits operations, and may veto or modify matching policies.
2. Narrowly Tailored Mandate
   • Matching only; no licensure/discipline (that remains with the state appraiser board).
   • No price setting; no exclusive dealing; open to any licensed/certified appraiser in good standing.
3. Objective Filters, Not Named Picks
   • Lenders/AMCs specify property/product type (e.g., 1004/1073/desktop), geography, complexity tier, required approvals (FHA/VA), SLA band, and other bona-fide competencies.
   • The system computes an eligible pool and rotates among them with light quality/SLA stabilizers (caps to prevent covert favoritism).
4. Full Auditability
   • Immutable logs: request filters, eligible set, scores, chosen appraiser, decline/timeout reasons, second-appraisal triggers, reconsiderations of value (ROVs).
   • Public aggregate dashboards (privacy-safe) on distribution fairness, turn-time, dispute/ROV metrics.
5. Coexistence with AMCs
   • AMCs may submit through the clearinghouse and continue post-match workflow (scheduling, QC, communications). The clearinghouse levels selection while preserving competitive service layers.
6. Due-Process & Competition Safeguards
   • No barriers beyond licensure/competence attestation and E&O.
   • No discrimination by fee levels during eligibility/rotation.
   • Appeals/complaints process for lenders and appraisers; sanctions only for integrity violations (e.g., falsified competencies).

Benefits​

• Independence: Reduces steering and perceived coercion without eliminating lender responsibility for competency.
• Transparency: Creates a contemporaneous audit trail useful for bank examiners, investors, and courts.
• Fairness: Distributes opportunities more equitably among qualified appraisers, especially independents and rural practitioners.
• Quality & Risk: Encourages assignment-to-competence alignment, reducing repurchase and valuation defect risks.

Implementation Outline​

• Phase I (0–6 months): Statute and rulemaking; governance; data-sharing safe harbors; privacy and security baselines (SOC 2 Type II).
• Phase II (3–9 months): MVP build—registries, eligibility filters, rotation engine, portals/APIs, immutable logging.
• Phase III (6–12 months): Pilot in three counties with at least three lenders and one AMC; external evaluation partner.
• Phase IV (12–18 months): Scale statewide; publish first transparency report; consider optional adoption for selected commercial segments.

Funding​

Per-transaction clearing fee paid by ordering party (capped), volume-tiered memberships for API access/analytics, and one-time seed grant or appropriation to prevent fee shock during the pilot.

Drafting Notes​

• Include explicit legislative findings on lender incentives and the insufficiency of purely retrospective enforcement.
• Specify active supervision mechanisms (approval of matching rules, periodic audits, remedial directives).
• Include antitrust/competition savings clauses (no fee-setting, no exclusive dealing, open eligibility).
• Provide data-privacy and cybersecurity minima; mandate immutable audit trails and retention periods aligned with banking regs.

 

[alebrewer]

Here's a way to make the system even more broken: hand it over to the politicians.

 

[RCA]

2) API Sketch (Order Intake & Matching Receipt)​

Base URL (example): https://api.appraisal-clearinghouse.state.us/v1

Authentication​

• OAuth2 client credentials (lenders/AMCs).
• mTLS for server-to-server integrations.
• All endpoints JSON over HTTPS.

2.1 Submit Order​

POST /orders

Request (example)​

POST /orders

{
"client_order_id": "ABC-2025-000123",
"lender_id": "LEND-4412",
"loan_purpose": "purchase",
"geography": {
"county_fips": "06081",
"subject_address_hash": "a6c5... (sha256)"
},
"product": {
"form": "1004",
"inspection": "full",
"program": ["FANNIE_MAE"]
},
"property": {
"type": "SFR",
"living_area_sqft": 2100,
"acreage": 0.15,
"features": ["ADU:false", "waterfront:false"]
},
"complexity_tier": "standard",
"approvals_required": ["FHA:false", "VA:false"],
"sla": { "target_days": 7, "latest_delivery_utc": "2025-10-15T00:00:00Z" },
"constraints": { "borrower_language": "en", "availability_window_days": 3 },
"filters": {
"geo_radius_miles": 25,
"min_experience_hours": 2000,
"licenses_required": ["CA_CERTIFIED_RESIDENTIAL"]
},
"attachments": {
"loan_number_hash": "b9e1...",
"prior_appraisal_hash": null
},
"notes": "Condo-like amenities; new subdivision nearby; avoid weekdays 12–2pm for access."
}

Response (Accepted)​

{
"order_id": "ORD-9c2e7a8c",
"status": "intake_pending",
"intake_timestamp": "2025-09-30T14:12:03Z",
"next": { "poll_after_seconds": 5, "href": "/orders/ORD-9c2e7a8c" }
}

2.2 Get Order Status / Matching Receipt​

GET /orders/{order_id}

Response (Matched)​

{
"order_id": "ORD-9c2e7a8c",
"status": "matched",
"match": {
"appraiser_id": "APR-CA-002913",
"tentative": true,
"expires_utc": "2025-09-30T14:27:03Z",
"contact_channel": "clearinghouse_messaging",
"handoff_href": "/assignments/ASN-7a31f1b9"
},
"decision_card": {
"filters": {
"geo_radius_miles": 25,
"licenses_required": ["CA_CERTIFIED_RESIDENTIAL"],
"min_experience_hours": 2000,
"form": "1004",
"inspection": "full"
},
"eligible_pool_size": 37,
"scoring_inputs": [
{ "appraiser_id": "APR-CA-002913", "recent_share_90d": 0.021, "competency_depth": 0.81, "quality_index": 0.94, "sla_penalty": 0.00, "score": 1.83 }
],
"selection_rule": "rotation_with_capped_quality_adjustments_v1",
"conflict_checks": { "prior_service_12m": false, "relationship_flags": [] },
"audit_hash": "7f1a... (sha256 over request+filters+eligible_set+scores)"
}
}

2.3 Decline/Accept by Appraiser​

POST /assignments/{assignment_id}/decision
{ "decision": "accept", "message": "Can inspect within 48 hours." }
(or)

{ "decision": "decline", "reason_code": "capacity", "message": "Booked through Friday." }

2.4 Reconsideration of Value (ROV) Intake​

POST /assignments/{assignment_id}/rov
{
"submitted_by": "lender",
"evidence": [
{ "type": "sale", "address_hash": "913f...", "close_date": "2025-07-15", "price": 1125000, "notes": "Similar condition per photos" }
],
"claim": "omitted_comparable",
"explanation": "Two sales within 0.3 mi not analyzed; request review."
}

2.5 Transparency (Aggregate)​

GET /analytics/distribution?county_fips=06081&amp;period=2025Q3
{
"orders": 1832,
"eligible_pool_mean": 42.7,
"distribution_gini": 0.19,
"median_turn_days": 6.3,
"rov_rate": 0.041
}