Revere Securities FINRA AML Fine: The Small-Cap IPO Red Flags Its Manual Review Missed

TL;DR

• FINRA fined Revere Securities $800,000 for AML and supervisory failures tied to foreign-based, small-cap IPOs and required an independent consultant-led remediation.
• FINRA says Revere relied on a manual daily trade-blotter review even after a 2017 settlement had identified that approach as unable to detect suspicious patterns across days, weeks, or months.
• The most useful lesson is operational: eight accounts with common referral and trading patterns, contradictory source-of-funds information, and liquidation instructions from unexpected countries still did not generate a suspicious-activity review.
• Broker-dealers should test account-linkage analytics, IPO allocation concentration, cross-account order timing, CIP exception handling, and evidence that independent AML testing occurred in the required calendar year.

The Revere Securities FINRA AML fine is what happens when a firm’s risk profile changes but its monitoring logic does not. Since 2022, Revere built a core business around small-cap initial public offerings, including roughly 60 IPOs and more than 40 involving issuers with operations in Hong Kong or China. FINRA found that the firm’s primary detection control remained a manual daily trade-blotter review.

That control could see trades. It could not reliably connect the accounts, addresses, referral sources, allocations, IP locations, order timing, or repeated activity that made the trades suspicious as a pattern.

FINRA’s October 6, 2026 announcement censures Revere, imposes an $800,000 fine, and requires a third-party consultant to complete a review and oversee remediation. Revere accepted the findings without admitting or denying them.

The amount is meaningful. The control history is more important. FINRA’s Letter of Acceptance, Waiver, and Consent, No. 2023080788201 says a 2017 Revere settlement had already identified the weakness in relying on a manual blotter to find multi-day or multi-account patterns in microcap activity. Nearly a decade later, the same basic design problem sat underneath a new product and customer mix.

What FINRA found in the Revere Securities AML case

FINRA’s findings cover activity from at least 2022 through the settlement. They are not limited to one bad customer or one missed alert.

FINRA cited violations of Rules 3310, 3110, and 2010. Rule 3310 requires a written AML program reasonably designed to comply with the Bank Secrecy Act, including suspicious-activity detection, customer identification, ongoing customer due diligence, and independent testing. Rule 3110 covers supervisory systems and written supervisory procedures. Rule 2010 requires high standards of commercial honor and just and equitable principles of trade.

This was not a surprise-risk case. FINRA had issued Regulatory Notice 22-25 in November 2022 about small-cap IPOs with characteristics associated with pump-and-dump or “ramp-and-dump” schemes. The notice specifically connected concentrated allocations, nominee accounts, foreign issuers, sharp post-IPO price moves, and suspicious-activity reporting obligations.

Eight accounts made the control failure hard to miss

The settlement’s clearest example concerns a September 2023 IPO by a Hong Kong-based fintech company. FINRA anonymized it as Issuer A.

The IPO raised $8 million. Revere allocated more than $1.5 million of shares—about 19.25% of the entire offering and nearly all of Revere’s allotment—to eight Australia-based accounts referred to the same representative by a foreign broker-dealer.

The onboarding data was already noisy:

• the customers opened their accounts over two days;
• four said they were unemployed while reporting annual income above $200,000;
• seven applications contained conflicting explanations for the source of deposited funds; and
• all eight expressed interest in the same IPO on the same date.

The trading added another layer. The stock had a $4 offering price, reached an intraday high of $12.74, and closed at $6.08. All eight customers emailed liquidation instructions on the IPO date. The requests came from IP addresses traced to countries other than Australia, where the account holders purportedly lived. FINRA also found later instances in which three of the accounts submitted similar buy-and-sell instructions on the same dates, sometimes minutes apart.

FINRA says none of that activity was flagged for potentially suspicious review.

The point is not that any one fact proves manipulation. It does not. The point is that the combined pattern should have created a reviewable case with linked evidence, an investigator decision, and a documented SAR determination. That is the difference between surveillance and a list of standalone alerts.

For teams rebuilding this control, the customer identification program documentation guide is the account-opening layer. The Revere action shows why CIP exceptions also need to flow into transaction monitoring and ongoing CDD rather than disappearing after approval.

The manual-blotter problem was already documented

The most damaging fact for a remediation team may be the oldest one.

In a 2017 AWC, according to the new settlement, FINRA found that Revere’s manual daily blotter review was not reasonably designed to detect suspicious patterns occurring over days, weeks, or months. That prior case imposed a censure and a $75,000 fine. The 2026 AWC says Revere did not begin implementing automated surveillance until late 2025.

That creates a basic issues-management question: what evidence showed the 2017 weakness was actually closed?

A closure package should have demonstrated more than revised policy language. At minimum, it should have included:

1. A requirements trace. Each cited risk—cross-day activity, linked accounts, low-priced securities, volume concentration—mapped to a surveillance rule or documented compensating control.
2. Scenario testing. Known historical patterns replayed through the new control to confirm they produced alerts.
3. Data validation. Proof that customer identifiers, addresses, referral source, IP or device data where lawfully collected, allocations, orders, executions, and wires reached the monitoring platform completely and accurately.
4. Alert-quality evidence. Sampled cases showing investigators considered the full linked pattern and recorded a defensible escalation or closure rationale.
5. Sustainable ownership. Named owners for tuning, model or rule changes, backlog monitoring, quality assurance, and annual independent testing.

This is where firms often get trapped: an issue is marked closed because a procedure was approved, while the production control still cannot detect the conduct described in the finding. The site’s analysis of repeat findings and weak closure validation provides a more defensible standard for proving that remediation worked.

FINRA’s small-cap sweep makes this broader than Revere

Revere is one firm, but the review perimeter is wider.

FINRA launched a targeted examination of small-cap offerings in October 2025. Its request list covers underwriters, bookrunners, syndicate and selling-group members, placement agents, firms handling initial or secondary trading, and firms with relevant omnibus accounts.

The request is unusually useful as an exam-readiness checklist. FINRA asks firms to produce:

• offering-specific due-diligence and approval procedures;
• Regulation M and Rule 5210 procedures;
• a tabular inventory of offerings, participants, compensation, allocations, and investors;
• all versions of the AML program;
• procedures for manipulation, spoofing, layering, wash trading, and coordinated trading; and
• a complete inventory of surveillance tools, including purpose, frequency, period of use, and whether each tool is proprietary or third-party.

FINRA’s 2026 Annual Regulatory Oversight Report adds the operating expectation: surveillance thresholds should reflect the product class and business model, patterns should be reviewed across customers and multiple days, external red flags should be incorporated, and alert decisions should be timely and documented.

Broker-dealers should read the Revere action alongside FINRA’s other 2026 low-priced-securities cases. The earlier Pictet Overseas and Blue Ocean ATS enforcement analysis shows the same theme through omnibus-account visibility and thin manual surveillance. Revere adds underwriting allocations, customer identity discrepancies, and a prior unresolved control warning.

Five control tests to run this week

Do not start by rewriting the policy. Start with production data and try to recreate the risk pattern.

1. Link accounts before reviewing alerts

Have AML Operations and Market Surveillance produce a population of accounts sharing any of these attributes:

• residential or business address;
• introducing broker or referral source;
• beneficial owner, authorized trader, phone number, email domain, device, or IP geography where available and permissible;
• same-day account opening or funding;
• same IPO indication of interest; or
• materially similar order timing and size.

The evidence artifact is a reproducible linkage report, not an analyst’s memory that two names looked familiar.

2. Measure allocation and market-volume concentration

For every small-cap offering, calculate the share allocated to the top accounts and to linked-account groups. After trading begins, calculate each linked group’s portion of daily volume.

Do not copy a universal trigger from this case. FINRA noted Revere customers exceeding 25% of market volume on multiple occasions, but a firm’s threshold should be calibrated to its securities, liquidity, historical alerts, and false-negative testing. Treat 25% as a fact from the settlement, not a safe harbor.

3. Turn CIP contradictions into stop conditions

Account Opening should define which discrepancies prevent approval until resolved: conflicting source-of-funds descriptions, implausible employment and income combinations, shared addresses with no explained relationship, recently formed entities with no apparent operations, or documents that do not support the claimed jurisdiction.

The approval record should show the discrepancy, evidence obtained, reviewer, decision, and any resulting customer-risk adjustment. “Business approved” is not a rationale.

4. Test surveillance across time and customers

Replay known patterns from the AWC: multiple accounts selling at the same limit price within minutes, coordinated liquidations on IPO day, repeated trading on identical dates, and orders sent from unexpected geographies. Confirm that scenarios aggregate activity across accounts and days.

Then reconcile source records to alerts. A sophisticated scenario is useless if referral-source or allocation data never reaches it.

5. Put independent testing on a non-negotiable calendar

Revere’s settlement says the firm did not complete annual AML testing in 2022 or 2023 and instead had a consultant address both years in 2024. The CCO and Internal Audit lead should maintain an obligation record showing scope approval, tester independence, fieldwork, final report, findings, and completion date for each required cycle.

If testing is overdue, log the breach as an issue; do not quietly roll it into next year’s scope. The site’s BSA/AML independent testing guide covers the evidence package in more detail.

The remediation terms tell firms what “done” looks like

Revere had already engaged a consultant before the AWC was issued. Under the settlement, the consultant must complete a comprehensive review of compliance with specified parts of Rule 3310 and submit an initial report within 60 days of FINRA’s acceptance notice. Revere must adopt the recommendations or propose alternatives designed to achieve the same objectives. The consultant evaluates those alternatives, and Revere must ultimately follow the consultant’s determination if they cannot agree.

The firm must then submit an officer-certified implementation report supported by narrative and exhibits. FINRA may request more evidence.

That last point deserves attention. Remediation is not complete when management says the recommendation was implemented. It is complete when evidence can show the control exists, operates, covers the cited risk, and produces reviewable results.

For a team facing the same problem, the first Monday-morning task is straightforward: pull the last 12 months of small-cap allocations, linked-account indicators, and post-offering orders into one dataset. If Compliance cannot reconstruct the patterns FINRA described, the firm does not yet have evidence that its surveillance would find them.

Use the Issues Management Tracker & Template to map each finding to an owner, action plan, due date, validation test, and closure evidence before the next regulatory request arrives.