# AML Model Governance: SR 11-7 & ATL/BTL Tuning (2026)

> Validate rule-based and ML/AI AML transaction monitoring models under Fed SR 11-7 / OCC 2011-12 governance, ATL/BTL sampling, and SHAP explainability.

Canonical: https://regtechcodex.org/briefings/aml-transaction-monitoring-model-validation-sr-11-7-backtesting
Last updated: 2026-11-14T19:46:17.413Z

**Publication:** RegTech Codex
**Category:** Fatf AML KYC Transaction Monitoring
**Published:** 2026-11-14T02:30:25.915Z
**Author:** Victoria Ashworth, LL.M, CAMS (LL.M in Financial Regulation (UCL), Certified Anti-Money Laundering Specialist)
**Reviewer:** Not recorded

## Answer-first summary

Supervisory consent orders routinely strike down "arbitrary threshold adjustments" tuned solely to shrink analyst headcount—requiring independent SR 11-7 conceptual soundness validation, Above/Below-the-Line sampling, and SHAP explainability.

## Key takeaways

- Under Federal Reserve SR 11-7 / OCC Bulletin 2011-12 (and the 2021 US Interagency Statement on Model Risk Management for BSA/AML), changes to AML thresholds must be justified by documented empirical risk effectiveness—NEVER analyst staffing constraints.
- Above-the-Line (ATL) testing measures alert-to-SAR precision above the active threshold, whereas Below-the-Line (BTL) stratified random sampling tests the "near-miss" band (e.g., 75%–99% of threshold) to quantify undetected false negatives.
- Complete end-to-end Data Lineage reconciliation—proving 100% of core banking, MT4/MT5/FIX trade, card, and blockchain ledger codes map accurately into the TM engine—prevents silent "unmonitored product" enforcement fines.
- Deploying gradient-boosted or graph machine-learning AML models requires local feature-attribution explainability (SHAP values) so L1/L2 investigators and regulatory examiners can understand and defend every alert disposition.

## Related financial entities

- **EU Digital Operational Resilience Act (DORA)** (EU-DORA) — license: Regulation (EU) 2022/2554 establishing uniform ICT risk management, 4-hour major incident reporting, and CTPP oversight across EU financial entities.
- **FATF Recommendation 16 & EU TFR (2023/1113) Crypto Wire Travel Rule** (FATF-TRAVEL-RULE) — license: FATF Recommendation 16 & EU TFR (2023/1113) Crypto Wire Travel Rule (FATF-TRAVEL-RULE) refers to an institutional financial benchmark and regulatory standard monitored by FATF / European Banking Authority (INTL), currently calibrated at EUR 0 (Zero De Minimis CASP Threshold) as of October 2026.

## Analysis

Reviewing major BSA/AML enforcement actions issued by the [Federal Reserve](/regulators/u-s-federal-reserve-system-156), [OCC](/regulators/office-of-the-comptroller-of-the-currency-u-s-155), [FDIC](/regulators/federal-deposit-insurance-corporation-u-s-158), [NYDFS](/regulators/new-york-state-department-of-financial-services-182) (`Part 504`), and the [UK FCA](/regulators/financial-conduct-authority-uk-121) reveals two recurring multi-hundred-million-dollar failures: **unmapped upstream transaction codes** silently omitted from surveillance feeds, and **undocumented threshold hikes** engineered to keep alert volumes inside analyst headcount budgets. Governed by `SR 11-7` (`Guidance on Model Risk Management`), the `2021 Interagency Statement`, and [EU DORA ICT & Algorithmic Governance](/definitions/eu-dora-regulation-2022-2554), surveillance teams at firms such as [E*TRADE](/licence-holders/e-trade), [Coinbase](/licence-holders/coinbase), and [CMC Markets](/licence-holders/cmcmarkets) enforce three-lines-of-defense model governance. Consult our [FATF AML/KYC Compliance Playbook 2026](/briefings/fatf-aml-kyc-cdd-transaction-monitoring-sar-playbook-2026) inside the [FATF AML/KYC hub](/topics/fatf-aml-kyc-transaction-monitoring).

## Four-Stage SR 11-7 Lifecycle for Rule & Machine-Learning AML Scenarios

Every scenario—whether a deterministic peer-percentile velocity rule or an XGBoost/Graph-Neural-Network typology scorer—must maintain an auditable **Model Inventory Dossier** verified by an independent Second-Line Model Risk Management (`MRM`) unit:

| MRM Lifecycle Pillar | Core Engineering & Statistical Deliverable | Fatal Supervisory Red Flag | Examination Pass Criterion |
| :--- | :--- | :--- | :--- |
| 1. Typology Coverage & Data Lineage | Enterprise Risk Assessment (`EWRA`) -> Scenario map + GL-to-TM row/dollar recon | New instant-payment or crypto product launched with unmapped `txn_code` | Automated daily `100.00%` source-to-TM ledger hash match |
| 2. Peer-Group Segmentation | Cluster homogeneous peers (Retail vs SME vs Prop Desk vs Institutional `CASP`) | Grouping `$5k` retail savers inside same percentile distribution as `$50M` hedge funds | Kolmogorov-Smirnov intra-cluster homogeneity + documented features |
| 3. `ATL` & `BTL` Threshold Tuning | Sensitivity grid + blinded `BTL` stratified sampling (`[0.75*tau, 0.99*tau]`) | Raising `$25k` threshold to `$100k` solely to clear a 12,000-alert backlog | Upper `95%` Clopper-Pearson bound on `BTL` productive SARs `<= 1.5%` |
| 4. Ongoing Drift & `SHAP` AI Audit | Monthly `PSI` (`<0.20`), champion-challenger shadow runs & top-5 `SHAP` drivers | Black-box ML probability score (`0.87`) with zero investigator reason codes | Human-interpretable feature attributions archived per case |

![SR 11-7 AML transaction monitoring model validation workflow covering data lineage, peer segmentation, ATL/BTL sampling, and PSI monitoring](/api/figure?type=flow&title=SR%2011-7%20Independent%20AML%20Scenario%20Validation%20Pipeline&subtitle=Proving%20data%20completeness%2C%20conceptual%20soundness%2C%20BTL%20false-negative%20bounds%2C%20and%20ongoing%20PSI%20stability&steps=1.%20Source-to-TM%20Recon%3A%3AReconcile%20100%25%20of%20core%20ledger%2C%20cashier%20%26%20FIX%20executions%20into%20TM%20staging%7C2.%20Sensitivity%20Grid%3A%3ASimulate%20parameter%20grid%20%28p90..p99%29%20on%2012m%20history%20%26%20known%20SAR%20seeds%7C3.%20Blinded%20BTL%20Audit%3A%3ASample%20near-miss%20%5B0.75*tau%2C%200.99*tau%5D%20band%20blind%20to%20bound%20missed%20SARs%7C4.%20Live%20PSI%20%26%20SHAP%3A%3AMonitor%20monthly%20Population%20Stability%20Index%20%28%3C0.20%29%20%26%20log%20SHAP%20drivers&metrics=PRODUCTIVE%20SAR%20LIFT%3A%3A3.8x%3A%3A6.9%25%20vs%201.8%25%20Legacy%20Alert-to-SAR%20Rate%7CUNMAPPED%20TXN%20CODES%3A%3A0%20Allowed%3A%3AAutomated%20Schema-Change%20CI%2FCD%20Gate%7CINDEPENDENT%20MRM%3A%3A2nd%20Line%3A%3ASeparate%20Reporting%20Line%20From%20Scenario%20Dev%7CANNUAL%20RE-VALIDATION%3A%3A12%20Months%3A%3AMandatory%20High-Materiality%20Tier-1%20Cadence)

## Preventing Silent "Unmonitored Product" Disasters With Automated Schema Gates

Why did multiple Tier-1 banks suffer `$100M+` fines after core-banking or brokerage upgrades? Upstream engineering teams added a new transaction type code (e.g., `TX_INSTANT_SEPA_V2` or `TX_STAKING_UNBOND`) without updating the ETL filter feeding the AML engine—leaving billions in volume completely unmonitored for years! Enforcing a daily **General-Ledger-to-TM Completeness Control** (blocking any unknown `txn_type_code` into an immediate P1 compliance incident) eliminates unmonitored-rail risk.

## Below-the-Line (`BTL`) Upper-Bound & Population Stability (`PSI`) Equations

Independent model validators evaluate proposed threshold hikes using exact binomial upper bounds on blinded `BTL` samples alongside monthly **Population Stability Index (`PSI`)** drift tracking:

```formula
PSI = sum_(b=1..10) ( Actual_Share_b - Baseline_Share_b ) * ln( Actual_Share_b / Baseline_Share_b )
SAR_Rate_BTL_hat = k_sar_btl / n_sampled_btl
Estimated_Missed_SARs = N_population_btl * Upper_95CI_ClopperPearson(k_sar_btl, n_sampled_btl)

Approval Gate for Threshold Increase (tau_old -> tau_new):
 Require k_sar_btl / n_sampled_btl <= 0.005 AND Upper_95CI <= 0.015 AND 100% Prior Known SARs Retained or Captured by Companion Rule

Worked Validation (Proposed Velocity Hike Creating N_btl = 4,000 Suppressed Alerts; n = 250 Blinded Sample Finds k = 0 Missed SARs):
 Rule-of-3 Upper 95% CI = 3 / 250 = 0.012 (1.2% <= 1.5% ceiling); PSI = 0.07 (< 0.20 stable) -> Approved by MRM Committee
```

![Comparison of alert-to-SAR conversion rates and false-negative capture across legacy rules, tuned peer percentiles, and explainable ML overlays](/api/figure?type=bars&title=Alert-to-SAR%20Precision%20Across%20Surveillance%20Architectures&subtitle=Empirical%202026%20benchmark%20comparing%20unsegmented%20static%20thresholds%20vs%20SR%2011-7%20peer%20%2B%20SHAP%20ML%20models&steps=Static%20Global%20Rules%3A%3A1.4%25%20SAR%20conversion%20%20%2014.2%25%20BTL%20leakage%20on%20SMEs%2FWhales%7CPeer-Segmented%20Rules%3A%3A4.1%25%20SAR%20conversion%20%20%203.1%25%20BTL%20leakage%20across%20cohorts%7CRules%20%2B%20SHAP%20ML%20Scorer%3A%3A7.6%25%20SAR%20conversion%20%20%200.8%25%20BTL%20leakage%20%28Explainable%20AI%29%7CGraph%20%2B%20ML%20Typology%20Net%3A%3A11.2%25%20SAR%20conversion%20%20%20Catches%20multi-account%20smurf%2Fmirror%20rings&metrics=HIBERNATION%20GOVERNANCE%3A%3AMRM%20Sign-Off%3A%3ARetiring%20Redundant%20Rules%20Requires%20Proof%7CALERT%20TRIAGE%20SLA%3A%3A%3C%3D%2015%20Days%3A%3APrevents%20Unworked%20Backlog%20Citations%7CSYNTHETIC%20RED-TEAM%3A%3AQuarterly%3A%3AInjects%20Simulated%20Structuring%2FTBML%20Seeds%7CNYDFS%20504%20ALIGNMENT%3A%3ACertified%3A%3ASupports%20Annual%20Board%20Attestation)

## NYDFS 23 NYCRR Part 504 Annual Board / Senior Officer Certification

For New York-chartered or licensed institutions—including global correspondent banks, broker affiliates, and BitLicense virtual-currency firms—`23 NYCRR Part 504` elevates transaction-monitoring and sanctions-filtering governance to personal executive accountability: requiring an annual **Board of Directors resolution or Senior Officer Compliance Finding** backed by documented data-lineage, pre-/post-implementation testing, and threshold-validation workpapers.

## Connecting Model Risk Management to Typologies, AI Governance & SAR Quality

Link `SR 11-7` model validation directly to [structuring and smurfing bunching models](/briefings/structuring-smurfing-ctr-threshold-evasion-detection-models), [brokerage mirror-trade and money-pass rules](/briefings/mirror-trading-wash-trades-capital-flight-brokerage-aml-rules), and [FinCEN SAR / goAML XML conversion feedback loops](/briefings/fincen-sar-goaml-xml-narrative-quality-completeness-audits) so investigator dispositions continuously refine scenario precision.

## Frequently asked questions

### Does every deterministic AML rule have to be treated as a formal "Model" under SR 11-7?

The April 2021 Interagency Statement (Fed, FDIC, OCC, FinCEN, NCUA) clarified that simple deterministic single-criterion rules (such as an exact $10,000.01 CTR trigger) need not automatically be classified as statistical "models," though they still require rigorous change-control, data-completeness testing, and periodic threshold validation.

### Why did regulators fine banks hundreds of millions for "capacity-driven" threshold tuning?

In multiple landmark OCC, FinCEN, and NYDFS consent orders, institutions raised transaction-monitoring dollar cutoffs specifically to cap monthly alerts at the number their existing analysts could review—suppressing thousands of genuinely suspicious SARs without conducting Below-the-Line impact studies.

### How is Below-the-Line (BTL) testing executed statistically when true false negatives are unknown?

Independent validation teams temporarily run scenarios in shadow mode across a 10%–25% lower parameter band (e.g., lowering a $50,000 velocity trigger to $37,500), draw a statistically sized stratified random sample (95% confidence level), and have senior QA investigators review those near-miss cases blind to measure missed-SAR rates.

### Can machine learning replace AML rules entirely, or how do supervisors expect AI to be governed?

FinCEN/Interagency AI Innovation statements, UK FCA, and EU DORA/AI Act frameworks permit hybrid ML anomaly scoring or alert-prioritization models provided firms maintain champion-challenger benchmarks, Population Stability Index (PSI) drift monitors, zero protected-class bias, and human-readable SHAP explanations.


## Source URLs

- https://www.federalreserve.gov/supervisionreg/srletters/sr1107.htm
- https://www.fincen.gov/news/news-releases/agencies-issue-statement-how-risk-management-principles-described-model-risk
- https://www.occ.gov/news-issuances/bulletins/2021/bulletin-2021-19.html

## Citation

- APA: Victoria Ashworth, LL.M, CAMS. (2026). AML Transaction Monitoring Model Governance: SR 11-7 Validation, ATL/BTL Tuning & AI Explainability. RegTech Codex. https://regtechcodex.org/briefings/aml-transaction-monitoring-model-validation-sr-11-7-backtesting
- Canonical article: [AML Transaction Monitoring Model Governance: SR 11-7 Validation, ATL/BTL Tuning & AI Explainability](https://regtechcodex.org/briefings/aml-transaction-monitoring-model-validation-sr-11-7-backtesting)
- Site research index: [RegTech Codex](https://regtechcodex.org/llms-full.txt)
