NewRez LLC entered a final multistate settlement and consent order on August 12, 2026 after regulators said the mortgage servicer imposed lender-placed, or force-placed, insurance on more than 4,200 borrowers who already had active homeowners coverage. The final settlement agreement and consent order totals $15.5 million and applies through 48 financial agencies covering 46 states and the District of Columbia. For billing-control owners, it creates an enforceable testing, refund and reporting regime for newly boarded loans.
The examination began in January 2022 and covered servicing from November 1, 2020 through October 31, 2021. NewRez remediated the borrowers identified through the examination and a subsequent audit. NewRez neither admits nor denies wrongdoing or violations. The company told HousingWire that the issues were identified several years ago, had been addressed and included remediation for affected borrowers. The order does not disclose NewRez’s current error rate, system design or whether the additional self-audit will find more borrowers.
What changed and what it means
A broken insurance-status-to-charge handoff can create improper borrower charges, refunds, remediation work, penalties and a reportable control failure.
- Decision affected
- Define and test the evidence gate that permits lender-placed-insurance charges on newly boarded loans, including the exception, cure and reporting path when monthly errors exceed 5%.
- Evidence in brief
- The final order requires monthly sampling of newly boarded loans with lender-placed insurance, manual evidence review, reporting of every monthly outcome and a failed metric when tested errors exceed 5%.
- What remains unresolved
- The order does not disclose NewRez’s current error rate, system or vendor design, or whether the additional self-audit will identify further affected borrowers.
- Next verification
- Map the insurance-status source, boarding handoff, charge authorization, sample denominator, cure evidence and regulatory-reporting owner before the 90-day implementation window closes.
Key takeaways
- The $15.5 million total includes $4.511 million of consumer relief already paid; the new penalty and cost components total $10.989 million.
- NewRez has up to 90 days to implement monthly testing, which then runs for one year from the implementation date.
- The sample is selected by state, and the reporting-period metric fails only when tested errors exceed 5%; exactly 5% is not a failed metric.
- An additional audit covers lender-placed-insurance fees on newly boarded loans from January 1, 2023 through the agreement’s effective date.
The $15.5 million total is not all a new payment
The CSBS announcement describes a $15.5 million settlement, more than $4.5 million in borrower remediation and nearly $11 million in costs and penalties. The order provides the exact components.
| Component | Amount | Status |
|---|---|---|
| Administrative penalty | $9,900,000.00 | Payable to participating states |
| Administrative costs | $1,088,757.84 | Payable to participating states |
| Consumer relief credit | $4,511,242.16 | Already paid for issues identified in the examination and subsequent audit |
| Total | $15,500,000.00 | Settlement amount |
The penalty and administrative costs add to $10,988,757.84. Those amounts are due within 30 calendar days after the effective date. For finance reporting, the settlement amount, new cash outflow and historical remediation are related but different measures. The New York DFS release, for example, separately reports the amount returned to affected New York borrowers and the state penalty.
The control starts before the insurance charge
Regulators described an evidence mismatch. A loan carried lender-placed insurance even though documentary evidence showed that valid homeowners insurance was already in force when the charge was assessed. The control point therefore sits before refund processing. The billing decision needs a reliable insurance-status record before a premium or fee reaches the borrower account.
The order requires monthly testing of newly boarded loans with lender-placed insurance in participating jurisdictions where NewRez operates. Testing can be performed by a third party. If NewRez performs it internally, members of internal audit or compliance who are independent of the applicable line of business must do the work. The review can include imaged loan documents, servicing-system notes or data, and management inquiries or interviews.
A control owner should preserve a traceable chain from the boarded-loan population to the insurance evidence, lender-placed-insurance initiation date, borrower charge, test result, exception owner, refund and regulatory report. That field-level design is Finance Circuit analysis, not a system specification in the order. It makes the required test reproducible instead of forcing an analyst to reconstruct the population after a charge posts.
Monthly testing creates a measurable failure trigger
NewRez must implement the enhanced standards no later than 90 days after the effective date. The one-year auditing period starts on the implementation date, so it does not necessarily run from August 12. The first month to be sampled is the first full calendar month after implementation; each sample draws from loans whose lender-placed insurance was initiated during the previous month.
Sample selection is state-based. Where a state has at least 20 newly boarded loans with lender-placed insurance in the review month, NewRez must randomly select 10%. Where there are 19 or fewer, it must test at least one loan. A sampled loan fails when valid documentary evidence shows that homeowners insurance was in place when lender-placed insurance was assessed.
The reporting-period metric fails when the total failed loans as a percentage of total loans tested exceeds 5%. Exactly 5% does not cross the written threshold. That does not make a 5% error rate acceptable, create a safe harbour or establish an industry tolerance. Each improper charge remains an error. The threshold determines when NewRez is deemed to have failed the specific metric in this order.
NewRez must report every monthly testing outcome to the regulators’ executive committee within 30 days after finalising the monthly report, including any above-threshold result. It can cure a failed loan by terminating the improper insurance and refunding all premiums and fees paid by the borrower. A defensible monthly package should therefore tie the source population, selected sample, evidence reviewed, failed loans, completed refunds and submitted report to one reporting period.
The additional self-audit extends the lookback
The order requires an audit of all lender-placed-insurance fees collected or refunded for newly boarded loans in participating states from January 1, 2023 through the effective date. NewRez must submit its methodology within 30 days. The executive committee then has 14 days to provide feedback or non-objection; silence in that window counts as non-objection. The audit is due within 180 days of that non-objection date, and its results are due within 30 days after finalisation.
If the audit finds that lender-placed insurance was placed in error and the borrower paid a premium or fee, NewRez must refund the full amount paid. The widely repeated description of “quarterly reporting” needs a narrower label. Updates every 90 days apply only to remediation for consumers identified by this additional audit, beginning 90 days after NewRez provides the audit results and continuing until those consumers are made whole.
What mortgage billing-control owners should lock now
- Population ownership. Fix who certifies the newly boarded loan population and the state-level counts before sampling begins.
- Evidence authority. Define which policy documents, servicing notes and status records prove that coverage was valid at the charge date.
- Charge authorization. Block lender-placed-insurance premiums and fees when insurance-status evidence is unresolved or contradictory.
- Sample reproducibility. Retain the monthly population, random-selection method, denominator and reviewer independence evidence.
- Exception closure. Link each failed loan to termination, refund amount, payment evidence and any corrective-action case.
- Regulatory reporting. Reconcile the report submitted to the executive committee to the tested population and completed cures.
The order does not prescribe a software product, vendor or data model. It does create an observable outcome: insurance evidence must support the charge, testing must be reproducible, every monthly result must be reported and identified errors must connect to remediation.
What the order does not establish
The settlement does not disclose NewRez’s current monthly error rate, how many loans will enter each state sample, whether a third party will conduct the tests, or whether the additional self-audit will identify more borrowers. The required reports are designated confidential supervisory information, so later operating results may not become public through routine reporting.
The 5% threshold is specific to this negotiated order and should not become another servicer’s default tolerance. Other firms still need to follow applicable law, their own risk standards and regulator-specific requirements. The next public verification points are implementation of the standards, any regulator disclosure from the additional self-audit and any later enforcement record.
