Table of Contents
Introduction
The laundering of tax credits and VAT refunds describes the process by which proceeds from VAT refund fraud and tax-credit refund fraud are moved through the financial system to disguise their criminal origin. Value Added Tax (VAT) refund fraud targets the tax mechanism, allowing VAT recovery fraudsters to submit false or inflated claims to tax authorities, receive payments they are not entitled to, and then layer and integrate those funds until they appear legitimate. This form of financial crime spans carousel fraud, missing trader fraud, fictitious input-credit claims, and abusive monetisation of energy, R&D, or green tax credits.
It is essential to distinguish between two linked but separate offences: (a) committing VAT or tax refund fraud against the revenue authority, the predicate crime, and (b) laundering the illicit proceeds of that fraud through banks, payment service providers (PSPs), or other financial channels. The predicate offence generates the money; the laundering conceals it.
This article focuses primarily on VAT refund fraud and EU cross-border VAT typologies. Other tax-credit schemes are included as adjacent examples and may operate under different rules.
How does VAT refund fraud become a money-laundering risk for banks and PSPs? Once tax authorities pay out fraudulent refunds or credits, arriving as apparently legitimate tax rebates into corporate bank accounts, those funds are already inside the financial system. From there, they can be placed, layered through rapid transfers, circular trade flows, or crypto conversion, and integrated via real estate purchases, luxury goods, or reinvestment in legitimate businesses.
Why does this matter? Governments face billions in revenue losses due to VAT refund fraud annually. The EU VAT compliance gap in 2023 was approximately €128 billion in lost revenues, while the EU Parliament estimates carousel fraud costs €50 billion annually. The European Public Prosecutor’s Office (EPPO) had 488 VAT-fraud-related active cross-border investigations in 2024 with estimated financial damage of about €13.15 billion. Supervisory expectations are clear: FATF considers tax crimes predicate offences, and the EU’s “VAT in the Digital Age” initiative, CESOP, and Eurofisc are building new reporting and transparency requirements for regulated entities.
Key takeaways from this article:
The mechanics of VAT refund fraud and abusive tax-credit claims as predicate offences
How fraudulent refunds are laundered through business networks, trade flows, financial products, PSPs, and crypto
Key red flags and investigative approaches for banks, PSPs, and crypto firms to detect and respond
How to enhance monitoring, screening, and case management using RegTech solutions such as ZIGRAM’s tools
Understanding VAT Refund Fraud, Tax Credits, and Money Laundering
What Is VAT Refund Fraud and Tax-Credit Abuse?
VAT refund fraud refers to illegitimate or deceptive VAT refund claims: fraudsters reclaim VAT without paying it to authorities. These claims may involve Missing Trader Intra-Community (MTIC) schemes, fake exports, inflated input VAT on non-existent purchases, or fabricated suppliers. Fraudsters generate fictitious transactions and invoices to claim VAT refunds, causing tax authorities to pay out funds or forgo tax that should be collected. Fraudsters manipulate invoicing amounts to influence the size of VAT refunds requested, and duplicate VAT claims occur when a business submits the same invoice more than once.
Tax-credit abuse extends beyond VAT to include fraudulent claims for input tax credits, R&D tax credits, energy-transition credits, green subsidies, and regional tax reliefs. Companies may fabricate projects, overstate costs, or issue and sell credits where permitted. These credits, if monetised or refunded, produce proceeds for fraudsters comparable to direct refund fraud.
Core patterns are visible across European Union member states and other VAT/GST jurisdictions. The EU Commission reports that the EU’s VAT Gap was approximately €89.3 billion in 2022, about 7.0% of total VAT liability. The EPPO tracks fraudulent practices, including cross-invoicing with artificial VAT credits, e-commerce fraud, use of “paper mill” companies, and margin taxation abuse. At this stage, proceeds typically flow from the tax authority to the fraudster, usually via bank accounts, cheques, or offsets against other tax liabilities.
Placement, Layering, and Integration in VAT Refund Schemes
In most financial crime typologies, placement is the initial step of introducing illicit cash into the financial system. VAT refund fraud often short-circuits this phase: funds originate directly inside the financial system because tax authorities pay refunds electronically into corporate bank accounts or PSP-held accounts. Placement-like features may still appear when intermediaries, nominees, family members, or third-party service firms receive funds as initial buffers.
Layering involves moving funds to obscure their origin. In VAT refund schemes, this takes the form of rapid wire transfers across jurisdictions, back-and-forth payments among corporate group entities, circular trade (Company A → Company B → Company C → Company A), sham loans among related entities, and the use of multiple PSP or electronic money institution (EMI) accounts to fragment value. Crypto conversion is increasingly common: legal-appearing funds from refunds are converted to virtual assets, mixed, then transferred to other wallets.
Integration is where funds re-enter the legitimate economy. This includes purchases of real estate, luxury vehicles, jewellery, and art; acquisition or investment in legitimate businesses; reinvestment as working capital; or even cash withdrawals. At this stage, funds appear to derive from lawful business operations.
In many carousel and MTIC frauds, the refund claim and layering happen almost simultaneously, with placement being brief or minimal. AML teams must therefore look at sequences of events: refund receipt, movement, and conversion, rather than waiting for a single red-flag transaction.
Understanding these mechanics sets the stage for examining concrete typologies and the network structures that enable them.
Typologies: From Carousel VAT Fraud to Refund Laundering Networks
Classic VAT fraud typologies such as MTIC and carousel fraud are often analysed from a tax perspective. Here, the focus is on how associated funds flow through banks and PSPs, the laundering dimension. The examples below draw on public enforcement cases and official reports.
Missing Trader and Carousel VAT Fraud as Laundering Engines
Missing trader fraud exploits VAT reclaim rules in transaction chains. The basic MTIC construct works as follows: goods are purchased VAT-free from another EU member state (intra-Community supply is zero-rated, then sold domestically with VAT charged to the buyer. The missing trader, typically a recently formed shell company, collects the VAT but disappears before remitting it to tax authorities. A missing trader collects VAT but does not pay it to authorities. Carousel fraud exploits VAT rules in EU transactions by circulating the same goods through multiple entities, sometimes across borders, repeatedly generating fraudulent VAT reclaim opportunities.
Funds are distributed through layers of participants: missing traders, buffer traders, brokers, logistics companies, and service providers, often using multiple bank and PSP accounts. Fraudsters often use multiple buffer traders to obscure transactions. Contra-trading utilizes parallel supply chains to manipulate VAT liabilities, adding further complexity. Carousel fraud often involves multiple shell companies set up across several EU member states.
Case example – “Fuel family” (Italy, EPPO, 2024–2025): A network of more than 40 missing traders imported fuel from Croatia and Slovenia using a tax warehouse for carousel fraud. The scheme generated simulated invoices exceeding €1 billion, with estimated VAT damage of €260 million. Over €35 million was laundered through bank accounts in Hungary and Romania. Assets seized included approximately 150 properties, a tourism resort, and vehicles.
Case example – “Vortex”: A cross-border VAT carousel fraud involving second-hand luxury cars across Europe, using hundreds of shell companies, forged documents, and fictitious invoices. Estimated losses reached €100 million across Bulgaria, Croatia, Germany, Italy, Latvia, Luxembourg, and Poland. Over €20 million in assets were seized, including bank accounts, luxury vehicles, and real estate.
In 2005–06, the UK lost an estimated €12.6 billion to carousel fraud – a figure that underscores the scale of the problem even in a single member state.
Fraudulent Tax-Credit Claims and Monetisation
Beyond missing trader fraud, a common form of tax fraud centres on credits rather than direct refunds. These include input tax credits on non-existent purchases, fake green or energy credits, R&D incentives, and regional subsidy-linked VAT relief – all exploited through false documentation or shell entities.
Credits translate into cash or balance-sheet value through several mechanisms:
Direct cash refunds: Tax authorities pay the cash equivalent of claimed credits
Offsetting: Credits applied against other tax liabilities, freeing up company cashflow and reducing what a company would otherwise owe
Sale or assignment: In jurisdictions that allow tradable tax credits, credits may be sold or transferred to third parties
Financial instrument linkage: Connecting refunds or credits to invoices, then using those in trade finance or as collateral
The EU Commission report mentions “cross-invoicing with artificial VAT credits” as a common fraudulent practice. OECD analysis confirms that overclaiming refunds and input VAT in sectors with weak oversight – often through fake invoices – is a significant revenue and financial integrity risk. In 2020, Singapore proposed legislation to deny input tax credits in cases of suspected fraud, illustrating the global reach of this problem.
Complex Entity Networks, Trade Flows, and Professional Enablers
The network structures behind these schemes are deliberately complex:
Chains of shell or low-substance entities spread across multiple EU and non-EU jurisdictions, often with nominee directors, straw persons, and secrecy jurisdictions for ownership layering
Freight forwarders, customs brokers, and logistics companies – sometimes complicit, sometimes unwitting – provide shipping documentation that supports fraudulent invoices or export claims
Professional enablers – accountants, tax advisers, lawyers, and corporate formation agents – may certify documentation, facilitate VAT registration, or design abusive structures; most professionals are legitimate but can be misused
Key insights for the detection discussion ahead:
The involvement of many related or layered entities typically reveals common ultimate beneficial owners (UBOs) or overlapping management
Invoice fraud – fake or overstated invoices – is central to virtually all VAT refund fraud typologies
Trade documents and customs/shipping documentation are often mismatched or missing entirely
Bank accounts frequently act as buffers: receive large VAT-refund-linked inflows, then transfer out to many recipients domestically and across borders within days
How Fraudulent VAT Refunds Flow Through the Financial System
Business and Account Profiles Linked to VAT Refund Fraud
Typical business profiles associated with VAT refund laundering include:
Recently incorporated traders with high declared turnover but minimal staff counts and low organisational substance
Import–export or wholesale firms with rapidly shifting product lines
Companies frequently amending VAT registrations, bank mandates, UBO details, or directors – invalid VAT registration and newly formed entities are red flags for VAT fraud
Trade, Invoice, and Payment Patterns
Recurring patterns in invoices and trade data provide further detection opportunities:
High volumes of same goods transactions flowing between closely connected entities with minimal price mark-ups
Circular trade flows (Company A to Company B to Company C and back to Company A) completed over short timeframes
Invoices referencing large export consignments with minimal logistics or customs clearance evidence
Shipping costs inconsistent relative to the declared weight or value of goods
Payment-chain features observable by banks and PSPs include: large VAT refunds aligned almost immediately with outgoing payments for purported goods or services to offshore jurisdictions; chains of pass-through accounts where balances return to near-zero soon after VAT-related inflows; and the use of multiple PSPs or e-money institutions to fragment and obscure flows.
AML teams can exploit structured data from invoices, shipping documents, and customs records when available. Organisations dealing with trade-based money laundering patterns should integrate these data sources into their monitoring platforms.
Laundering Channels: Cash, Crypto, and Asset Acquisition
The main channels used to launder proceeds of VAT refund fraud include:
Cash: Large withdrawals or cash-intensive business turnover following receipt of VAT refunds; physical transport of cash among network participants
Asset acquisition: Purchases of real estate, luxury vehicles, jewellery, and art via bank transfers originating from refund accounts. In the “Fuel family” case, seized assets included properties, a tourism resort, and vehicles
Crypto: Refunds converted to virtual assets via exchanges, OTC brokers, or P2P platforms, followed by rapid coin mixing or transfers to unhosted wallets in jurisdictions with weak regulation
Detection and Response: Red Flags, Investigation, and Controls
This section moves from description to practical AML and FCC implications. The goal is to equip compliance teams with actionable indicators and investigative frameworks – not operational blueprints for committing fraud. Fraud detection requires rigorous controls and monitoring of transaction patterns.
Key Red Flags for Banks, PSPs, and Fintechs
Red flags should be assessed in context; they do not automatically indicate criminality but should trigger enhanced review.
Customer risk indicators:
Opaque or frequently changing ownership structures; nominee directors or straw persons
Mismatch between registered business activity (e.g., import/export) and actual trade documents or financial statements
Multiple companies sharing the same address, service providers, or management despite different jurisdictions
Minimal organisational substance relative to claimed turnover
Behavioural indicators:
Reluctance to provide VAT returns or tax documentation
Inconsistent or vague explanations for large credits from tax authorities
Frequent requests for refunds from jurisdictions known for lax VAT controls
Transactional indicators:
Sudden large credits from tax authorities followed by rapid scatter payments to multiple beneficiaries
Accounts cycling back to near-zero balances after refund inflows
Structured payments designed to stay below reporting thresholds
Cross-border transfers to or via jurisdictions with weak AML oversight
Trade-specific indicators:
Unsupported export claims; high-value goods with implausibly low shipping costs
Same goods appearing in repeated round-trip transactions among related entities
Routing through high-risk jurisdictions without clear commercial rationale
Crypto/PSP indicators:
Loading of prepaid instruments using tax-refund-linked funds
Quick conversion to virtual assets, especially via unregulated exchanges
Transfers to newly created or unverified wallets; rapid mixing of coins or tokens
Investigative Approach for AML and Financial Crime Teams
When investigating suspected VAT-refund-linked laundering, financial crime compliance teams can apply several analytical approaches:
Correlate tax payment references with customer KYC data, declared business models, and financials. A company claiming modest wholesale activity that receives disproportionately large treasury payments warrants deeper scrutiny.
Reconstruct account activity around refund dates: map inward flows against onward transfers, speed of dispersal, and destinations. Rapid dispersal to multiple jurisdictions is a strong layering indicator.
Compare customer-supplied invoices, contracts, and shipping documents with external data – corporate registries, customs records, and trade databases where accessible. Discrepancies in goods descriptions, weights, or values between trade and financial documents suggest fabrication.
Use network analysis to identify repeated links to known or suspected MTIC actors, shell entities, or high-risk PSPs. Mapping common UBOs, directors, or addresses across apparently unrelated entities often reveals the underlying fraud network.
Internal SAR narratives should clearly separate predicate suspicion (tax fraud indicators) from observed laundering actions (movement and integration of funds). Avoid legal conclusions about tax liability; instead, articulate risk indicators and factual observations. Training employees to identify VAT fraud indicators is vital for compliance.
Entity risk assessment, name screening, and adverse media tools – such as ZIGRAM’s PreScreening.io and Dragnet Alpha – can support this process by surfacing networked risks and cross-border linkages against known MTIC prosecutions or tax-fraud scandals.
Controls, Monitoring, and RegTech Support
Risk appetite: Formal statements explicitly covering tax-fraud-related money laundering and exposure to MTIC sectors, cross-border trade, and PSP-facilitated flows
Customer onboarding: Enhanced due diligence for high-risk VAT sectors and cross-border traders, including scrutiny of beneficial ownership, tax compliance history, expected VAT position, and effective supplier due diligence
Ongoing monitoring: Scenario-based rules tuned to detect suspicious VAT refund flows – alerting when refund size is disproportionate to revenue, when rapid outgoing flows follow treasury credits, or when unusual transaction corridors appear. Machine-learning approaches can complement rule-based detection
Thematic reviews: Periodic audits of portfolios disproportionately exposed to VAT-intensive supply chains
RegTech enhances detection at each stage:
Entity resolution and network analytics identify hidden linkages in VAT fraud chains, aligning with capabilities like ZIGRAM’s Entity Hero
Name screening and adverse media catch connections to known MTIC prosecutions or tax-fraud scandals
Transaction monitoring fine-tuned for tax-authority-originated flows and trade anomalies enables faster, more accurate alerting with fewer false positives
Cross-Border Cooperation and Regulatory Expectations
Information Sharing with FIUs, Tax Authorities, and Law Enforcement
Suspicious Transaction/Activity Reports (STRs/SARs) are critical for capturing proceeds of tax fraud. Quality narratives should include: clear description of account behaviour and transaction flows; explanation of why VAT refund or tax-credit payments appear atypical; any links to known typologies such as carousel patterns or named cases; and details of counterparties, jurisdictions, and timing. SARs should avoid legal conclusions about tax liability but must clearly articulate risk indicators and factual observations.
Legal frameworks enabling cooperation include domestic gateways between FIUs and tax authorities; the EU’s Eurofisc network for VAT fraud administrative cooperation; and EPPO investigations of cross-border VAT cases above specified thresholds. The EPPO’s “Money Cat” probe, for example, involved estimated tax damage over €60 million, with over €11 million in invoices for non-existent goods and services passed through international banking channels.
Supervisory Focus and Emerging Regulatory Initiatives
Supervisory expectations are intensifying across multiple fronts:
FATF: National risk assessments must include risk from VAT fraud and tax refund fraud; tax crimes are confirmed predicate offences
European Commission: The “VAT in the Digital Age” (ViDA) proposal introduces real-time digital reporting and e-invoicing, while CESOP imposes payment data reporting obligations on PSPs for cross-border transactions. The EU adopted a Quick Reaction Mechanism on 1 August 2012 to respond faster to emerging fraud patterns
National enforcement: Cases like “Fuel family” and “Vortex” demonstrate the scale of seizures, asset freezing, and cross-border bank account restrictions. The reverse-charge mechanism reduces opportunities for certain types of VAT fraud – the UK introduced a reverse charge mechanism on 1 June 2007 to combat MTIC fraud
Governments spend substantial resources on enforcement against VAT fraud schemes, and supervisors increasingly expect banks to maintain controls for tax-fraud-related ML exposure
Common Challenges and Practical Solutions
Even sophisticated institutions struggle with VAT-refund-linked typologies due to data gaps, complex structures, and rapidly evolving fraud methods.
Challenge 1: Distinguishing Aggressive Tax Planning from Criminal Tax Fraud
The grey area between lawful tax optimisation, taking advantage of available reliefs, and criminal fraud can be hard to navigate. Criminal fraud involves false statements, fake invoices, and non-existent supplies. AML teams aren’t tax auditors, but they must recognise patterns that clearly suggest deceit.
Solutions:
Establish internal escalation channels to tax specialists or external advisers when suspicious VAT patterns emerge
Define practical indicators that trigger higher scrutiny: unexplained refunds, mismatches between VAT refunds and underlying trade volumes, supplier names unknown in commercial registries, and invoices from jurisdictions with known fake-invoice risks
Challenge 2: Fragmented Data Across Systems and Jurisdictions
Customer, transaction, trade, and tax-related data often sit in different silos – sometimes across multiple countries – hindering holistic risk assessment. KYC data, trade data, customs records, tax-authority refund data, and PSP transaction flows may never be connected.
Solutions:
Invest in data integration and entity-resolution tools that unify KYC, transaction, trade, and adverse media data. ZIGRAM’s entity-risk and screening capabilities can assist in linking fragmented datasets
Enforce consistent use of common identifiers – tax IDs, VAT numbers, company registration numbers – across internal systems to aid linkage
Leverage legal mechanisms for cross-border data sharing, including FIU cooperation, MOUs, and Eurofisc channels
Challenge 3: Keeping Monitoring Scenarios Current with Evolving Typologies
Fraudsters rapidly pivot to new products, sectors, and payment channels. Organised crime groups shift from electronics to services or digital goods, exploit new tax incentives, or adopt crypto and emerging PSPs. Businesses involved in VAT fraud may face severe legal penalties and reputational damage – but the schemes keep evolving.
Solutions:
Conduct regular typology refresh cycles informed by FIU reports, tax authority alerts, FATF typologies, and EPPO case summaries
Use flexible, configurable transaction-monitoring platforms that allow quick deployment of new scenarios without long IT lead times
Train analysts to recognise emerging patterns and feed insights back into model tuning and rule calibration through structured FRAML feedback loops
Continuous improvement and collaboration between tax, AML, and risk functions are essential for keeping pace with this evolving threat.
Conclusion and Next Steps
VAT refund fraud and tax-credit abuse are among the most significant sources of illicit proceeds in the European Union and globally. The schemes – from carousel fraud to fictitious credit claims – generate billions in criminal proceeds each year. VAT fraud undermines public confidence in tax systems and fairness of tax regulations. Banks, PSPs, fintechs, and other regulated entities play a critical role in disrupting the laundering of those proceeds by detecting suspicious patterns and reporting them effectively.
Practical next steps for AML leaders:
Review customer and sector risk assessments for exposure to MTIC and VAT-intensive supply chains
Evaluate whether current transaction monitoring rules capture key VAT refund and tax-credit laundering indicators, including refund-triggered dispersals and circular flows
Enhance collaboration between compliance, tax, and finance teams on suspicious refund patterns – cross-functional escalation channels are essential
Benchmark existing tools and data sources and consider augmenting them with RegTech capabilities such as ZIGRAM’s screening, transaction monitoring, and entity-risk platforms
Plan staff training focused specifically on tax-crime-related money laundering typologies and red flag recognition
Looking ahead, the rollout of mandatory e-invoicing under ViDA, expanded payment data reporting through CESOP, and deepening international cooperation through EPPO and Eurofisc will substantially increase data availability. Institutions that invest now in the systems, skills, and governance structures to exploit these datasets will be far better positioned to detect and disrupt VAT refund laundering as it evolves.
Additional Resources and References
The following authoritative public sources and selected RegTech resources provide deeper reading on VAT refund fraud and related money laundering:
FATF: Guidance on tax crimes as predicate offences and related ML typologies
European Commission: VAT fraud publications including VAT Gap reports and the “VAT in the Digital Age” initiative
EPPO: Annual report and case studies relating to cross-border VAT carousel and refund fraud investigations
OECD: Tax Administration 2024 – analysis of tax gaps and refund over-claiming
ZIGRAM resources: Transaction monitoring solutions, name screening tools, fraud monitoring and detection, and financial crime compliance guidance
FAQ for VAT Refund Fraud
Are banks expected to detect underlying tax fraud, or only the laundering of its proceeds?
Institutions are primarily responsible for identifying suspicious financial activity that may involve proceeds of crime – including tax fraud – not for re-performing tax audits. However, strong indicators of fraudulent VAT refunds or tax-credit abuse, combined with suspicious movement of funds, should trigger SARs even if the exact tax loss cannot be quantified. Supervisors expect banks to have scenarios and controls that recognise when tax-authority payments are disproportionate to a customer’s profile and when onward flows suggest laundering rather than legitimate business activity.
What sectors and customer types pose the highest VAT refund laundering risk?
Historically high-risk sectors include cross-border wholesale trading in electronics (particularly mobile phones and computer chips), carbon credits, commodities, fuel, and second-hand luxury vehicles. Payroll companies and labour-intensive services firms can also be exploited, particularly where input VAT is claimed on fictitious subcontracting. Risk is dynamic and jurisdiction-specific: classification should be tailored to local VAT and tax rules and informed by recent enforcement trends. Newly incorporated entities with disproportionate turnover and frequent changes in ownership or directors merit particular attention.
How can we enhance detection without overwhelming our teams with false positives?
Combining rule-based scenarios with customer risk rating and network analytics allows teams to prioritise genuinely suspicious patterns. Rather than generating alerts on every tax-authority credit, layer contextual factors: customer risk score, speed and pattern of onward transfers, counterparty profiles, and trade-document consistency. Feedback loops from investigators into scenario tuning are essential – analysts who review false positives should feed learnings back to model calibration. RegTech solutions that support explainable, risk-based alerting help reduce noise while maintaining coverage.
