Money Laundering Detection in Trading Analysis
Identify suspicious trading patterns and money laundering schemes with Sourcetable AI. Analyze transactions, detect wash trades, and monitor unusual activity automatically.
Andrew Grosser
February 24, 2026 • 18 min read
Understanding Money Laundering Detection in Trading
March 2022: FinCEN identifies 177 shell companies used to launder $1.3B through US real estate and crypto. Sourcetable helps compliance teams analyze suspicious transaction patterns before regulators do. Money laundering through trading markets represents one of the most sophisticated financial crimes facing institutions today. Criminals exploit complex trading strategies to obscure the origins of illicit funds, using techniques like wash trading, layering through multiple accounts, and structured transactions designed to evade detection thresholds. For compliance teams, risk analysts, and surveillance officers, identifying these patterns requires analyzing massive volumes of trading data across multiple dimensions—transaction timing, counterparties, pricing anomalies, and behavioral patterns.
Traditional surveillance systems rely on rigid rule-based alerts that generate thousands of false positives while missing sophisticated schemes. Excel-based analysis requires manual formula creation for each detection scenario, pivot tables that break with large datasets, and hours of data preparation before any actual investigation begins. When you're reviewing hundreds of accounts executing thousands of trades daily, these limitations create dangerous blind spots sign up free.
Sourcetable transforms money laundering detection by combining spreadsheet flexibility with AI-powered analysis. Upload trading data from any source and ask questions in plain English: 'Show me accounts with circular trading patterns,' 'Identify transactions just below reporting thresholds,' or 'Find counterparties with unusual relationship networks.' The AI understands financial crime typologies and automatically applies detection logic across your entire dataset. Get started with intelligent surveillance at sign up free.
This comprehensive guide covers detection methodologies for common money laundering schemes in trading, red flag indicators that distinguish legitimate activity from suspicious patterns, and how AI-powered analysis dramatically reduces investigation time while improving detection accuracy. Whether you're building a surveillance program from scratch or enhancing existing controls, you'll learn practical techniques for identifying the trading patterns that matter most.
Why Sourcetable Excels at Money Laundering Detection
Excel surveillance workbooks quickly become unmanageable when analyzing trading data at scale. A single day's trading activity across 500 accounts might generate 10,000+ transactions requiring cross-referencing against counterparty databases, historical patterns, and known typologies. Traditional spreadsheets force you to write VLOOKUP formulas across multiple sheets, create complex IF statements for each detection rule, and manually refresh calculations that take minutes to complete. When suspicious activity requires immediate investigation, these delays are unacceptable.
Sourcetable's AI assistant understands anti-money laundering terminology and trading surveillance concepts. Instead of writing nested formulas to detect wash trades (buying and selling the same security within short timeframes with no economic purpose), you simply ask 'Identify accounts with offsetting trades in the same security within 24 hours.' The AI instantly analyzes transaction timestamps, security identifiers, quantities, and account relationships to flag suspicious patterns. What would take 30 minutes of formula construction happens in seconds.
The platform handles the data complexity that breaks traditional spreadsheets. Link trading data with customer due diligence records, beneficial ownership structures, and geographic risk ratings without manual VLOOKUP chains. When you ask 'Show me high-risk jurisdiction clients with trades exceeding normal patterns,' Sourcetable automatically joins datasets, applies risk scoring logic, and calculates baseline deviations. The AI maintains these relationships as data updates, eliminating the broken reference errors that plague Excel surveillance tools.
Visualization capabilities make pattern recognition immediate. Ask for 'network graph of counterparty relationships for flagged accounts' and Sourcetable generates interactive visualizations showing trading connections that might indicate layering or integration schemes. Create time-series charts showing transaction velocity changes, heat maps highlighting geographic concentration, or scatter plots revealing pricing anomalies—all through natural language requests. Compliance officers can spot circular trading networks or structured transaction patterns at a glance rather than parsing thousands of spreadsheet rows.
Most importantly, Sourcetable adapts to evolving money laundering techniques without requiring technical reprogramming. When new typologies emerge—like using cryptocurrency derivatives for layering or exploiting low-liquidity securities for price manipulation—you can immediately create detection queries using plain English descriptions. The AI applies these concepts across your data without waiting for IT resources or vendor updates to traditional surveillance systems.
Benefits of AI-Powered Money Laundering Detection
Effective money laundering surveillance delivers critical value beyond regulatory compliance—it protects institutional reputation, prevents financial losses from facilitating criminal activity, and creates competitive advantage through superior risk management. Organizations that detect suspicious activity early avoid the multi-million dollar fines, remediation costs, and business restrictions that follow enforcement actions. Sourcetable's intelligent approach delivers these benefits while dramatically reducing the resource burden of surveillance operations.
Automated Pattern Recognition Across Multiple Typologies
Money laundering schemes rarely present as single suspicious transactions—they manifest as patterns across time, accounts, and transaction types. Wash trading might involve 50+ offsetting trades spread across three accounts over two weeks. Layering schemes create complex webs of transfers designed to obscure audit trails. Sourcetable's AI simultaneously monitors for multiple typologies without requiring separate analysis for each scheme type.
Ask 'Identify accounts showing wash trading, structuring, or rapid movement patterns' and the AI applies detection logic for all three typologies at once. It recognizes that an account executing offsetting trades (wash trading) while keeping individual transactions below $10,000 (structuring) and moving proceeds within 48 hours (rapid movement) represents significantly higher risk than any single indicator. This multi-dimensional analysis mirrors how experienced investigators think but applies it consistently across thousands of accounts. Traditional rule-based systems require separate alerts for each pattern, creating alert fatigue when the same account triggers multiple rules.
- Structuring (Smurfing): Breaking large cash transactions into amounts just below the $10,000 CTR reporting threshold; detection requires identifying customers with multiple same-day transactions summing to $9,500-$9,999, often across multiple branches.
- Layering via Shell Companies: Tracing beneficial ownership through chains of LLCs; FinCEN Beneficial Ownership rules require banks to know the natural persons behind legal entity customers; automated entity graph analysis identifies circular ownership structures.
- Round-Tripping: Funds sent offshore and returned as foreign investment; US bank receives $5M wire from a Cayman Islands entity owned by a US person who originally sent $4.9M offshore last year - the structure creates apparent foreign investment from domestic funds.
- Trade-Based Money Laundering: Over- or under-invoicing trade transactions; exporter invoices $100K for goods worth $500K, transferring $400K in value internationally; detecting requires comparison to industry average transaction values for similar trade categories.
Baseline Deviation Analysis for Behavioral Monitoring
Static thresholds miss sophisticated laundering that stays just below detection limits. An account that normally trades $50,000 monthly suddenly executing $200,000 in a week represents significant behavioral change even if individual transactions seem unremarkable. Sourcetable automatically establishes behavioral baselines for each account and identifies statistically significant deviations.
Simply ask 'Show me accounts with trading volume exceeding 3 standard deviations from their 90-day average.' The AI calculates historical means and standard deviations for each account, then flags outliers requiring investigation. You can refine this with contextual factors: 'Exclude accounts with documented business changes in CRM notes' or 'Prioritize accounts from high-risk jurisdictions.' This behavioral approach catches money laundering that evolves to avoid static detection rules. A compliance analyst reviewing 500 accounts can immediately focus on the 12 showing genuinely unusual patterns rather than investigating every account exceeding arbitrary thresholds.
- Transaction Velocity: Normal retail customer: 20-30 transactions/month averaging $150; suspicious: 200+ transactions/month averaging $9,500 - velocity change of 10x with amounts just below reporting thresholds is a classic structuring signal.
- Peer Group Benchmarking: Compare customer transaction behavior to a peer group (same industry, company size, geography); a restaurant generating $50K/month cash deposits vs peer average of $15K warrants SAR investigation - 3x deviation from peer median.
- Geographic Anomalies: Customer based in Ohio making consistent wire transfers to Panama, Belize, and Cayman Islands; geographic risk scoring flags jurisdictions with weak AML controls; FATF blacklist and grey-list countries receive automatic enhanced due diligence.
- Rapid Asset Accumulation: Customer with $30K annual income purchasing $800K in real estate cash in 18 months without clear wealth explanation; asset-income ratio alert triggers wealth of funds documentation requirement.
Network Analysis for Relationship-Based Schemes
Layering schemes deliberately spread transactions across multiple accounts controlled by the same beneficial owner or criminal network. Individual accounts might appear unremarkable, but collective analysis reveals coordinated activity. Sourcetable's AI identifies these networks by analyzing counterparty relationships, shared identifying information, and coordinated transaction timing.
Upload trading data with counterparty information and ask 'Map trading networks for accounts flagged with other suspicious indicators.' Sourcetable generates network graphs showing which accounts trade with each other, identifies tightly connected clusters that might represent layering networks, and highlights accounts serving as hubs in suspicious transaction chains. You can then query 'Calculate total volume flowing through this network' to assess the scheme's magnitude. This relationship analysis is nearly impossible in traditional spreadsheets, which lack native network analysis capabilities and require manual mapping of connections.
- Correspondent Banking Networks: Shell company A wires to Shell B wires to Shell C, all owned ultimately by the same beneficial owner; network analysis identifies ring structures where funds return to originator after 3-5 hops - a classic layering structure.
- Common Signatories: Multiple accounts with different business names but the same authorized signatories; a single person controlling 12 LLCs all receiving wires from the same foreign source is a beneficial ownership red flag requiring SAR filing.
- Shared Address Risk: Multiple accounts with the same registered address (often a registered agent address housing hundreds of shell companies) - not automatically suspicious but combined with other risk factors triggers enhanced scrutiny.
- SAR Filing Threshold: BSA requires SAR filing for suspicious transactions of $5,000+ ($2,000 for MSBs); penalties for non-filing can reach $1M per violation; automated pattern detection reduces false negative rate and ensures consistent application of reporting standards.
Temporal Pattern Detection for Timing-Based Schemes
Money launderers often exploit timing to avoid detection—executing trades just after market close to delay reporting, structuring transactions across statement periods, or coordinating activity with known enforcement gaps. Sourcetable's AI recognizes these temporal patterns by analyzing transaction timestamps, frequencies, and distributions.
Ask 'Identify accounts with unusual concentration of trades near reporting deadlines' or 'Show me transaction patterns that avoid business day detection windows.' The AI analyzes timestamp distributions, calculates clustering metrics, and flags accounts whose timing patterns differ significantly from legitimate trading behavior. For example, it might identify an account where 80% of trades occur within 30 minutes of market close compared to a portfolio average of 15%—a red flag for potential reporting evasion. You can visualize these patterns with 'Create hourly distribution chart for flagged accounts' to immediately see timing anomalies that would require complex Excel formulas to detect.
Pricing Anomaly Detection for Market Manipulation
Money laundering through trading sometimes involves intentional mispricing—executing trades significantly above or below market rates to transfer value between accounts. These pricing anomalies indicate potential layering or integration of illicit funds. Sourcetable automatically compares transaction prices against market benchmarks to identify suspicious pricing.
Upload trading data with execution prices and ask 'Flag transactions executed more than 5% away from contemporaneous market prices.' The AI compares each trade against market data, calculates percentage deviations, and identifies outliers. You can refine this with 'Exclude illiquid securities with wide bid-ask spreads' to reduce false positives from legitimate market conditions. For OTC or thinly-traded securities without reliable market prices, ask 'Identify trades between related parties with unusual pricing compared to other transactions in the same security.' This catches value transfer schemes that exploit securities lacking transparent pricing.
How Money Laundering Detection Works in Sourcetable
Effective surveillance combines multiple detection techniques applied systematically across trading data. Sourcetable's AI-powered approach follows a structured methodology that mirrors best practices in financial crime detection while eliminating the manual effort typically required. Understanding this workflow helps you implement comprehensive surveillance that catches sophisticated schemes while managing investigation workload.
Step 1: Data Integration and Enrichment
Money laundering detection requires combining trading data with contextual information—customer due diligence records, beneficial ownership structures, geographic risk ratings, and historical behavior profiles. Start by uploading your core trading data: transaction dates, account identifiers, security symbols, quantities, prices, counterparties, and transaction types. Sourcetable handles CSV, Excel files, or direct database connections from trading systems.
Next, enrich this data by linking customer information. Upload KYC records with customer risk ratings, jurisdiction codes, business descriptions, and expected activity profiles. Ask Sourcetable to 'Join trading data with customer records using account ID.' The AI automatically performs the merge, matching on account identifiers and creating a unified dataset. Unlike Excel's VLOOKUP which breaks when data updates, Sourcetable maintains these relationships dynamically. You can then add beneficial ownership data: 'Link accounts sharing the same beneficial owner or control person' to enable network analysis across related accounts.
- Money laundering detection requires combining trading data with contextual infor.
- "Join trading data with customer records using account ID."
- "Link accounts sharing the same beneficial owner or control person"
Step 2: Baseline Establishment and Behavioral Profiling
Before detecting anomalies, establish what constitutes normal behavior for each account. Ask Sourcetable to 'Calculate 90-day baseline metrics for each account: average daily volume, typical transaction size, common trading hours, and primary securities traded.' The AI computes these statistics automatically, creating behavioral profiles that serve as comparison points for detecting deviations.
You can refine these baselines with contextual factors. Request 'Segment accounts by customer type and calculate separate baselines for retail, institutional, and high-net-worth categories.' This prevents false positives from comparing accounts with fundamentally different legitimate trading patterns. For accounts with insufficient history, ask 'Apply peer group baselines to new accounts based on customer profile similarity' so you can monitor even recently opened accounts effectively. Sourcetable maintains these baselines automatically as new data arrives, ensuring your detection logic stays current without manual recalculation.
Step 3: Multi-Typology Screening
Apply detection logic for common money laundering typologies simultaneously. For wash trading detection, ask 'Identify accounts with offsetting buy and sell transactions in the same security within 48 hours where net position change is less than 10%.' Sourcetable analyzes transaction sequences, matches offsetting trades, calculates timing gaps, and flags accounts meeting these criteria. You'll see results showing the account, security, trade pairs, timing, and net position change for each suspicious pattern.
For structuring detection, request 'Flag accounts with multiple transactions between $9,000 and $10,000 within a rolling 7-day period.' The AI identifies transaction clustering just below reporting thresholds that suggests deliberate structuring to avoid currency transaction reports. For layering detection, ask 'Show accounts with more than 10 transactions in 48 hours followed by complete position liquidation and withdrawal.' This catches rapid movement patterns designed to obscure audit trails. You can run all these queries simultaneously, and Sourcetable will flag accounts triggering multiple typologies—the highest risk cases requiring immediate investigation.
- Apply detection logic for common money laundering typologies simultaneously.
- For structuring detection, request 'Flag accounts with multiple transactions bet.
Step 4: Relationship and Network Analysis
Sophisticated schemes involve multiple accounts working in coordination. Ask Sourcetable to 'Identify trading networks where accounts frequently trade with the same counterparties or show coordinated transaction timing.' The AI analyzes counterparty relationships across all accounts, identifies tightly connected clusters, and flags networks with characteristics suggesting coordination rather than coincidental market activity.
Visualize these networks by requesting 'Create network graph showing trading relationships for flagged accounts.' Sourcetable generates interactive visualizations where nodes represent accounts and edges represent trading relationships, with edge thickness indicating transaction volume. You can immediately spot circular trading patterns, hub accounts serving as transaction intermediaries, or isolated clusters of accounts that only trade with each other—all red flags for layering schemes. Click on any node to see detailed transaction history and ask follow-up questions like 'What's the total value flowing through this network in the past 30 days?'
Step 5: Risk Scoring and Prioritization
Not all suspicious indicators carry equal risk. Accounts triggering multiple red flags from high-risk jurisdictions with inadequate due diligence require immediate investigation, while single anomalies from established low-risk customers might warrant monitoring. Ask Sourcetable to 'Calculate composite risk scores based on: number of typologies triggered, customer risk rating, jurisdiction risk, and deviation magnitude from baseline.'
The AI applies weighting to each factor and generates overall risk scores for every flagged account. Request 'Rank accounts by risk score and show top 20 requiring investigation' to immediately focus on the highest-priority cases. You can refine scoring logic by asking 'Increase weight for accounts from FATF high-risk jurisdictions' or 'Add bonus score for accounts with incomplete beneficial ownership information.' This risk-based approach ensures investigation resources focus on the most significant threats rather than chasing low-risk anomalies that generate false positive alerts.
Step 6: Investigation Support and Documentation
Once you've identified high-risk accounts, Sourcetable accelerates the investigation process. For any flagged account, ask 'Show me complete transaction history, counterparty details, and comparison to baseline behavior.' The AI generates comprehensive investigation packages with all relevant data pre-assembled. Request 'Create timeline visualization of suspicious activity' to see how the pattern developed over time, making it easier to document in suspicious activity reports.
You can also perform comparative analysis: 'How does this account's activity compare to similar customers in the same jurisdiction?' helps determine if patterns might have legitimate business explanations. For regulatory reporting, ask 'Summarize key facts: total suspicious volume, time period, typologies identified, and deviation from expected activity.' Sourcetable generates narrative summaries that form the foundation of SAR filings, dramatically reducing the time from detection to reporting. All analysis is documented automatically, creating audit trails showing your detection methodology and investigation steps.
Real-World Money Laundering Detection Use Cases
Money laundering schemes vary by institution type, product offerings, and customer base. These practical scenarios demonstrate how different organizations apply Sourcetable's AI-powered surveillance to detect suspicious activity specific to their risk profiles. Each use case shows actual detection techniques that have identified real money laundering patterns in trading environments.
Broker-Dealer Wash Trading Surveillance
A mid-sized broker-dealer with 5,000 active accounts processes 50,000+ trades monthly across equities, options, and fixed income. Their compliance team of three analysts struggled with Excel-based surveillance that required overnight batch processing and generated 200+ alerts weekly, of which 95% were false positives from legitimate market-making or hedging activity. Investigation backlog grew to 600+ pending alerts, creating regulatory risk and obscuring genuine suspicious activity.
They implemented Sourcetable for wash trading detection by uploading daily trading files and asking 'Identify accounts with offsetting trades in the same security within 24 hours where the account has no market-making designation and position changes by less than 5%.' This refined query excluded legitimate market-making while flagging accounts with no economic purpose for offsetting trades. The AI immediately identified 12 accounts showing consistent wash trading patterns—buying and selling the same low-volume securities repeatedly with minimal price variation.
Further investigation using 'Show counterparty networks for these accounts' revealed that 8 of the 12 accounts were trading exclusively with two counterparty accounts, creating a circular trading network. The compliance team asked 'Calculate total volume and commission generated by this network' and discovered $4.2 million in circular trading generating $85,000 in commissions over three months—a classic wash trading scheme to create false trading volume while generating commission rebates. They filed SARs within 48 hours of detection and implemented ongoing monitoring with 'Alert me when accounts in this network resume trading.' Detection time dropped from weeks to hours, and false positive rates fell to 15%.
Cryptocurrency Exchange Layering Detection
A cryptocurrency exchange with international operations faced unique challenges detecting layering schemes where criminals rapidly move funds through multiple crypto-to-crypto trades before converting to fiat currency. Traditional blockchain analysis tools tracked on-chain transfers but missed the trading layer where criminals exploited exchange wallets to obscure transaction trails. Their surveillance focused on withdrawal monitoring but missed the intermediate layering happening through trading activity.
Using Sourcetable, they uploaded trading data including deposit sources, trade sequences, and withdrawal destinations. They asked 'Identify accounts with deposits from external wallets followed by more than 15 trades across different trading pairs within 24 hours, then complete liquidation to fiat and withdrawal.' This query targeted the classic layering pattern: deposit illicit crypto, execute numerous trades to create complexity, convert to fiat, and withdraw to bank accounts.
The AI flagged 34 accounts matching this pattern. The compliance team refined results by asking 'Exclude accounts with trading history longer than 90 days and filter for deposit sources from mixing services or high-risk exchanges.' This focused on new accounts using known laundering infrastructure. They identified 18 accounts that deposited funds from mixing services, executed 20-50 rapid trades creating convoluted audit trails, then withdrew to bank accounts in jurisdictions with weak AML enforcement. Total suspicious volume exceeded $12 million. By asking 'Map the trading sequences for these accounts and identify common patterns,' they discovered most accounts traded through the same intermediate cryptocurrencies (privacy coins like Monero) before converting to fiat—a clear layering methodology. This intelligence informed enhanced monitoring rules and led to 18 SARs filed with detailed trading pattern documentation.
Investment Advisor Structuring and Smurfing Detection
A registered investment advisor managing 200 client accounts noticed unusual cash deposit patterns but lacked systematic surveillance for structuring schemes. Individual deposits appeared unremarkable—$8,000 here, $9,500 there—but a client relationship manager suspected coordinated activity designed to avoid $10,000 currency transaction reporting requirements. Manual Excel analysis of six months of transactions took a full week and found suggestive patterns but couldn't definitively establish structuring across the entire client base.
They loaded transaction history into Sourcetable and asked 'Identify accounts with multiple cash deposits between $7,000 and $10,000 within any rolling 30-day period where the total exceeds $50,000.' This targeted structuring patterns while filtering out accounts with occasional large deposits. The AI flagged 7 accounts meeting these criteria. The compliance officer then asked 'Show deposit timing patterns for these accounts—are deposits clustered on specific days or spread evenly?' and discovered that 5 accounts made deposits every 3-4 days in remarkably consistent amounts ($8,200, $8,500, $9,000) suggesting deliberate structuring rather than normal business cash flow.
Deeper analysis using 'Check if these accounts share beneficial owners, addresses, or contact information' revealed that 4 of the 5 accounts had the same business address despite being registered to different entities—a smurfing scheme using multiple accounts to structure deposits while staying below per-account thresholds. By asking 'Calculate aggregate deposits across these related accounts,' they discovered the network deposited $340,000 in cash over 6 months, all structured to avoid CTR filing. The firm filed SARs and implemented ongoing surveillance with 'Monitor all accounts at this address for similar deposit patterns.' Sourcetable's AI reduced investigation time from one week to 90 minutes while providing definitive evidence of coordinated structuring that manual Excel analysis couldn't establish conclusively.
Hedge Fund Trade Allocation Manipulation Detection
A hedge fund administrator providing back-office services to 50 funds noticed concerning trade allocation patterns at one client fund. The fund manager had discretion to allocate block trades across multiple investor accounts post-execution—a legitimate practice for operational efficiency that also creates opportunities for abuse. Suspicious activity would involve systematically allocating profitable trades to preferred accounts (potentially including the manager's personal accounts or those of associates) while allocating losing trades to other investors. Traditional surveillance focused on performance dispersion but struggled to distinguish intentional manipulation from legitimate allocation methodologies.
The administrator uploaded trade allocation data showing block trade executions, subsequent allocations to individual accounts, and profit/loss outcomes. They asked Sourcetable 'For each account, calculate the percentage of allocated trades that were profitable versus unprofitable, then identify accounts with statistically significant deviations from the portfolio average.' The AI computed win rates for each account and flagged 3 accounts where 78-82% of allocations were profitable compared to the portfolio average of 54%—a statistically improbable result suggesting preferential allocation.
They then asked 'Show the timing of allocations—are profitable trades allocated more quickly than unprofitable ones?' and discovered that profitable trades were allocated to the flagged accounts within minutes of execution, while unprofitable trades were allocated hours later after price movements were clear. This timing pattern strongly suggested the manager was waiting to see price direction before making allocation decisions. By requesting 'Compare allocation patterns for these accounts versus others with similar risk profiles and investment mandates,' they confirmed the preferential treatment wasn't explained by different investment strategies. The administrator reported findings to the fund's board and regulatory authorities, leading to manager replacement and investor restitution. Sourcetable's ability to perform complex statistical analysis across allocation timing, profitability, and account characteristics revealed manipulation that would have required custom programming in traditional systems.
Frequently Asked Questions
If your question is not covered here, you can contact our team.
Contact UsWhat are the three stages of money laundering and how are they detected in financial systems?
How does anti-money laundering (AML) compliance work in financial institutions and what are the costs?
How do real estate transactions enable money laundering and what reporting requirements apply?
How do cryptocurrency transactions complicate AML enforcement and what regulatory responses have emerged?
What machine learning techniques do financial intelligence units use to detect money laundering patterns?
What are the global financial penalties for AML violations and how do they affect institutional risk management?
How do trade-based money laundering (TBML) schemes work and what detection methods are most effective?
