Skip to content
RiskTemplates · The Daily Brief Thursday, November 19, 2026
Wire FINRA's Spartan Capital Case: The Private-Placement Diligence File That Wasn't There NOV 17

Feature Operational Risk

The SEC's FY2026 Scorecard: $100 Million for Monitoring That Didn't Fire

SEC FY 2026 enforcement results: the biggest penalties went to firms with working systems that never produced a reviewed alert. Here's the evidence file.

By Rebecca Leung · November 18, 2026 ·
Table of Contents
  • The SEC’s FY2026 release reports 348 actions and $1.095 billion in monetary relief. Read past that. The two penalties the release itself chose to highlight are $100 million against Western Asset and $20 million against UBS Financial Services — both for programs that existed on paper and did not function.
  • Western Asset was $100 million because it knew its co-CIO’s allocation practices diverged from everyone else’s and never built the step that turned that knowledge into a control. UBS was $20 million because its FX wire monitoring failed on the legacy system, then failed again on the replacement, and the lookback SARs it eventually filed were themselves late.
  • Same year, same release, a custody case: Northeast Financial Group, no fund audits for fiscal years 2015 through 2024, $75,000. A decade of nothing is cheaper than one broken surveillance scenario. That tells you where the SEC is pricing risk.
  • The evidence that saves you is not a scenario count. It is the alert-to-ticket chain with a named dispositioner and a supervisor sign-off.

The number the release buried

A hundred million dollars. Then twenty. Then seventy-five thousand.

Those three numbers sat in a single SEC press release on November 10, 2026, and they do not scale the way you would expect. Northeast Financial Group missed annual fund audits for fiscal years 2015 through 2024 — a decade — and paid $75,000. Western Asset Management failed to catch a cherry-picking scheme running inside its own trading desk and paid $100 million. Roughly 1,300 times more.

The gap is not about how bad the conduct was. It’s about what the firm had already promised. Northeast had a deficient practice. Western Asset had a policy, a known divergence, and no working mechanism between the two.

That distinction is the whole FY2026 release, and it is the most useful thing in it for anyone who owns a monitoring function.

What the year actually totaled

The Division of Enforcement’s FY2026 annual results report 348 actions: 235 standalone and 113 follow-on administrative proceedings for bars and suspensions, plus 13 trading suspensions. Monetary relief came to $1.095 billion — $752 million in disgorgement and prejudgment interest, $343 million in civil penalties. About $731 million went back to more than 250,000 investors. Two-thirds of standalone actions named at least one individual.

The release also confirms the Enforcement Manual got its first update since 2017, and that Rule 202.5(e) is gone — the policy that let settling defendants neither admit nor deny. That reversal has already changed settlement arithmetic, and if you have not walked your disclosure committee through it, the no-deny rescission playbook is the place to start.

But the release’s own structure is the tell. Under “Impactful Enforcement,” before any Ponzi scheme, the SEC leads with program failures. It is the third consecutive year that the priorities shift toward operating discipline has shown up in the numbers rather than the speeches.

Case one: the system failed, so they bought another system

UBS Financial Services settled on August 3, 2026 for $20 million, a censure, and a cease-and-desist order. The order covers January 2019 through June 2023 and finds that UBSFS failed to adequately monitor customer FX wire transactions for suspicious activity, as its own AML policies required.

Two things make it worth studying.

First, the failure mode migrated. The order attributes the gap first to flaws in the legacy AML transaction monitoring system, then to flaws in the new automated system brought in to replace it. A platform migration is supposed to reset your monitoring risk. Here it relocated it.

Second, the remediation made things worse. UBSFS hired a consulting firm to fix the data and run an FX wire lookback, and began filing lookback SARs in October 2023. The SEC found those SARs untimely anyway — thousands of transactions, roughly $250 million in value. The violation was charged as willful under Section 17(a) of the Exchange Act and Rule 17a-8.

Read that again as a practitioner: the firm did the expensive, defensible-sounding thing, and the clock still ran. Engaging the consultant bought no grace period. The only thing that would have helped was the monitoring working when the wires moved.

Case two: they knew, and knowledge wasn’t a control

Western Asset Management is the more instructive of the two, and it is not close. The SEC’s June 5, 2026 order found the firm failed to take reasonable steps to detect and prevent its former co-CIO’s cherry-picking scheme, which ran from January 2021 through October 2023 and steered trades with first-day gains to favored portfolios and first-day losses to disfavored ones.

Here is the sentence that matters. The order finds Western Asset “was aware that its former co-CIO’s trading and allocation practices diverged from those of other portfolio managers at the firm, and that Western Asset knew or should have known about such practices.” The firm also failed to implement its own policies and procedures on reallocations.

So the firm had awareness. It had a written policy. It had a co-CIO whose behavior visibly differed from every peer. And it had nothing that converted that into an alert, a ticket, or a supervisory conversation. The SEC charged willful violations of Sections 206(2) and 206(4) and Rule 206(4)-7 — the compliance program rule — and failure to reasonably supervise under Section 203(e)(6). The $100 million penalty goes into a Fair Fund for investors.

Compliance program rules are usually treated as paperwork obligations. They aren’t. Rule 206(4)-7’s operative verb is “implement,” and this order prices the gap between adopted and implemented at nine figures.

The three ways this actually breaks

Every monitoring failure in these files resolves to one of three mechanics. They are worth naming because each has a different evidence answer.

Failure mechanicHow it looks in a fileThe artifact that proves it isn’t happening
Alert generation never firesScenario exists, thresholds set once at build, output volume collapses or hits zeroMonthly alert-count-by-scenario trend with an expected floor and a documented investigation when a scenario goes quiet
Alert fires, nothing opensAlert queue shows hits; ticketing system shows nothingAutomated alert-to-ticket reconciliation, exceptions treated as incidents, not housekeeping
Ticket opens, closure is unauditedClosures all coded the same, no supervisor review, average handle time implausibly flatStratified closure sample per quarter, disposition code taxonomy, supervisor sign-off above a materiality line

The middle row is the one that gets missed. Most firms can produce a scenario inventory. Far fewer can produce a reconciliation proving every alert in the queue became a ticket with an owner. That reconciliation is a two-hour SQL job if anyone has thought to build it, and a two-week archaeology project if nobody has.

Where ownership actually rots

Nobody writes a control objective that says “monitoring may be nominal.” It happens through a sequence that feels reasonable at every step.

Surveillance sits in Compliance. The scenario parameters live with a vendor or a data team. Tuning requests route through a change board that prioritizes production stability. Alert volume creeps down because thresholds get widened to cut false positives, and nobody restates the false-positive rate against caught outcomes, so the trade-off is invisible. The model validation team is scoped to models, not scenarios, so nothing independent ever looks. By year two, the number of alerts fired is a measure of system health in the reporting pack rather than a measure of whether the system is finding anything.

The UBS facts are that story with a date range attached. Legacy system flawed, replacement system flawed, data remediated by outside consultants, filings late. At no point was anyone lying.

What to pull before you need it

A 30-day build, in order:

Days 1–5. Reconciliation, one scenario, ninety days of history. Alert count versus ticket count. Any variance over zero gets an incident number and an owner. Do this before you tune anything, because tuning against an unreconciled pipeline hides the problem.

Days 6–12. Disposition taxonomy. If every closure is coded “no further action,” the coding tells you nothing. Minimum viable set: false positive, escalated to investigations, SAR or filing decision, referred to supervision, data quality defect. Second-line reviews a sample of twenty closures and grades them.

Days 13–20. Threshold change log with a reason field. Every parameter move needs a requester, a rationale, and the false-positive rate at the time of the move. When an examiner asks why the scenario went quiet in Q3, the log is your answer.

Days 21–30. Written known-divergence memo. This is the Western Asset control, and almost nobody has it. Once a quarter, supervision names every case where a desk, PM, or branch behaves differently from peers and states whether the difference is explained, monitored, or accepted. It is a one-page memo that takes an hour and would have had a name on the co-CIO’s allocation pattern two years before the SEC did.

The threshold numbers themselves should come from your own history, not a template. Pull the last six months of alert volume and outcomes, set a floor at the low end of normal variance, and investigate breaches of the floor. Then test for gaming by checking whether a scenario’s fall in volume tracks a fall in the underlying activity. If wires are flat and alerts are down 40%, someone widened a threshold.

So what

The SEC’s own release put two monitoring failures ahead of every fraud case in FY2026, and priced them at $100 million and $20 million. Meanwhile a decade of missing fund audits cost $75,000. The message isn’t that custody matters less — it’s that regulators have learned firms can manufacture the appearance of a control system cheaply, and are now pricing the difference between the appearance and the behavior.

If you own a surveillance function and cannot produce, today, an alert-to-ticket reconciliation for one scenario, you are carrying the same exposure Western Asset carried. It is a fixable exposure, and it is fixable in a month. The KRI Library’s 152 indicators include alert-volume floors, disposition-mix ratios, and closure-age bands you can wire straight into a monitoring report — grab it here and start with the three scenarios your last exam liked least.

FAQ

What were the SEC’s FY2026 enforcement totals?

The Division of Enforcement’s November 10, 2026 release reports 348 total actions — 235 standalone and 113 follow-on administrative proceedings — plus 13 trading suspensions. Monetary relief totaled $1.095 billion, split between $752 million in disgorgement and prejudgment interest and $343 million in civil penalties. About $731 million was returned to more than 250,000 investors.

Why is a monitoring failure enforcement rather than an exam finding?

Rule 206(4)-7 requires advisers to adopt and implement written policies reasonably designed to prevent violations; Exchange Act Rule 17a-8 does the equivalent for SAR filing. Western Asset was penalized for failing to implement its own reallocation policies. UBS paid $20 million after its monitoring systems, legacy and replacement, failed to surface suspicious FX wires.

What evidence proves a monitoring system works?

A dated sample that runs the whole loop: the scenario that generated the alert, the ticket, the dispositioner, the closure rationale, the supervisor signature. Plus the parameters file, the threshold change log, and the latest recalibration analysis. A scenario count proves the system exists. A disposition sample proves it produces judgment.

Do lookback SARs get any timing relief?

No. In the UBS order the firm hired a consultant and started filing lookback SARs in October 2023, and the SEC still charged the filings as untimely for thousands of transactions worth roughly $250 million. Remediation effort doesn’t toll the clock.

What’s the cheapest starting point if alert volumes are too low to sample?

Alert-to-ticket reconciliation for one scenario over ninety days. If the counts don’t match, you’ve found the gap before an examiner does. Then add disposition codes beyond “closed” and require supervisor sign-off on closures above a materiality line.

◆ Need the working template?

Start with the source guide.

These answer-first guides summarize the required fields, evidence, and implementation steps behind the templates practitioners search for.

◆ Immaterial Findings · Weekly

Sharp risk & compliance insights. No fluff.

◆ FAQ

Frequently asked questions.

What were the SEC's FY2026 enforcement totals?
The Division of Enforcement's November 10, 2026 release reports 348 total actions — 235 standalone actions and 113 follow-on administrative proceedings — plus 13 trading suspensions. Monetary relief totaled $1.095 billion: $752 million in disgorgement and prejudgment interest and $343 million in civil penalties. About $731 million was returned to more than 250,000 investors, and two-thirds of standalone actions charged at least one individual.
Why is a monitoring failure an enforcement matter rather than just an exam finding?
Because Rule 206(4)-7 requires advisers to adopt and implement written policies reasonably designed to prevent violations, and Exchange Act Rule 17a-8 does the same for SAR filing at broker-dealers. Once a firm writes the procedure, 'implement' is the operative word. Western Asset was penalized $100 million for failing to implement its own reallocation policies; UBS paid $20 million after its monitoring systems, old and new, failed to surface suspicious FX wires.
What evidence proves a monitoring system actually works?
A dated sample that runs the full loop: the rule or scenario that generated the alert, the ticket it opened, the analyst who dispositioned it, the closure rationale, and the supervisor who signed. Plus the parameters file showing thresholds, the change log when a threshold moved, and the last recalibration analysis against outcomes. A scenario count proves the system exists. A disposition sample proves it produces judgment.
How fast do lookback SARs have to be filed?
Untimely is untimely even when the lookback itself was voluntary. In the UBS order, the firm engaged a consultant and began filing lookback SARs in October 2023, but the SEC still found those filings late for thousands of transactions totaling roughly $250 million. Hiring the consultant does not toll the clock.
What is the cheapest fix if our alert volumes are too low to sample?
Start with alert-to-ticket reconciliation for a single scenario and pull the last ninety days. If the alert count and the ticket count do not match, you have found the gap before an examiner does. Then add one disposition code beyond 'closed' — false positive, escalated, SAR filed, referred to supervision — and require a supervisor sign-off on every closure above a set materiality line.
Rebecca Leung

Author

Rebecca Leung

Rebecca Leung has 8+ years of risk and compliance experience across first and second line roles at commercial banks, asset managers, and fintechs. Former management consultant advising financial institutions on risk strategy. Founder of RiskTemplates.

◆ Related framework

KRI Library (152 Key Risk Indicators)

152 KRIs — including 20 emerging-risk KRIs for AI-enabled fraud, scams and AI governance — with thresholds, owners and a calculating dashboard.

Immaterial Findings · Newsletter

The brief, in your inbox.

Enforcement of the week, a framework breakdown, and the prompts that are actually worth running. Delivered to your inbox. Free.