VAT Carousel Fraud (MTIC): Mechanics, Actors, Risk Zones

A VAT carousel does not need knowing accomplices — it needs honest traders who do not know its mechanics.

VAT carousel fraud — missing trader intra-Community (Missing Trader Intra-Community fraud: EU-wide VAT fraud in which one trader in the chain reports the tax, never remits it, and vanishes.">MTIC) fraud in its most organised form — is the heaviest single attack on the EU VAT system. It is also the phenomenon that most frequently pulls honest businesses into investigations they never saw coming. This page explains how a carousel works, which actors it uses and where the risk zones lie today — with verified figures, an EU-wide perspective and a German enforcement focus. Soberly, and without blanket suspicion against trade.

The mechanics: a system gap, not a force of nature

The EU VAT system has two design features which, combined, create the attack surface. First: intra-Community supplies between businesses are zero-rated — goods cross internal borders free of VAT. Second: input VAT deduction refunds a trader the tax it paid, regardless of whether its supplier ever remits that tax to the state. Foundational IMF research observed as early as 2007 that the system’s invoice trail creates audit points, but the buyer has no built-in incentive — and no inherent duty — to ensure that its supplier actually pays the tax over.

That is precisely where the carousel begins. The basic circuit:

1. The missing trader acquires goods VAT-free from another Member State, sells them domestically with VAT shown on its invoices, collects that VAT — and never remits it. Then it disappears. Economic analysis shows the leverage: a missing trader can pocket the VAT on the entire value added up to its stage of the chain.
2. Buffers — one or more intermediate traders — buy and resell the goods domestically. They lengthen the chain, create the look of normal commerce and complicate tracing. Crucially, the research literature itself describes buffers as frequently unwitting intermediaries. A position in the diagram proves no part in the plan.
3. The broker exports the goods VAT-free to another Member State and reclaims the input VAT it paid. The treasury thereby refunds money it never received at the missing trader’s stage.
4. The carousel closes when the goods — typically via a conduit company abroad — are sold back to the starting point and the circuit spins again. The same goods can rotate repeatedly; in a current Frankfurt investigation, goods were, according to the investigators’ findings, partly sold back to the original suppliers.

Two loops run in parallel: goods circle one way, money circles the other — and at exactly one point, tax leaks out of the system on every rotation. That leak is the entire business model.

Variants refine the pattern. Cross-invoicing and contra-trading offset fraudulent and clean trading flows against each other to make refund claims look unremarkable. For risk assessment, nothing changes at the core: the same mechanics, better camouflaged.

The actors: role labels are analytical tools, not verdicts

Missing trader, buffer, broker, conduit — these terms come from fraud analysis. They describe functions in a scheme. Nothing more. The international literature that coined this typology warns against overextending it: role typologies explain how a fraud model works; they do not prove that a specific trader knowingly performed such a role.

This is not an academic footnote — it is the pivot of every case. Organised fraudsters deliberately build their chains so that real companies doing real business form the links. The more normal the chain looks, the better the fraud works. The uncomfortable but clear consequence for honest traders: your integrity does not determine your risk of being drawn into proceedings — your ability to prove that integrity later does. What follows from the buffer position, and which case law protects it, is examined on the page “Caught in the chain”.

The scale: figures instead of drama

This page needs no dramatisation — the verified figures suffice. The European Commission’s current VAT Gap Report puts the EU compliance gap at roughly €128 billion (data year 2023), with organised MTIC fraud estimated at €12.5 to €32 billion per year. The European Public Prosecutor’s Office (EPPO) — the EU prosecution authority running cases through national investigators — reported 981 ongoing VAT and customs fraud investigations at the end of 2025, with estimated damage of €45 billion. In the preceding year, Germany alone accounted for 179 active VAT fraud investigations with an estimated €3.89 billion in damage.

The European Parliament’s research service draws the same picture from a neutral vantage point: its briefing on the fight against VAT fraud bundles the VAT gap, the Eurofisc early-warning network, transaction network analysis, CESOP payment data, the ViDA reform and the EPPO into one overall picture in which data analysis and prosecution interlock. For Germany, dedicated empirical findings now exist: in 2022, Germany counted 114 EPPO investigations — 66 of them VAT carousels, with damage of around €1.5 billion. Only one figure remains open: how large the carousel share of the VAT gap really is, not even the German Federal Government can quantify — yet the enforcement architecture keeps growing.

Behind the totals stand operations whose names the trading community knows. “Admiral” — at €2.9 billion the largest VAT fraud ever investigated in the EU, electronics traded via online marketplaces. “Goliath” — over €188 million, AirPods and small electronics, run out of Hamburg. “Mela” — €48 million, unlawful margin-scheme treatment of brand-new mobile phones. “Emily” — over €103 million, luxury-car carousels in which established dealerships stand accused as organisers. And the Frankfurt case of May 2026: an €18 million small-electronics carousel in which German suppliers allegedly acted as buffers.

These cases show two things. First: enforcement is massive, coordinated and data-driven — rightly so. Second: the profile of the accused is shifting towards real, established companies. That is exactly why this topic belongs on the management agenda, not only in the tax department.

Risk zones: where carousels prefer to run

Goods. Carousels favour goods with high value, low volume, low logistics cost and high fungibility — the research calls it the low-volume/high-value pattern: mobile phones, small electronics, chips, precious metals, carbon certificates, energy. It is no coincidence that mobile and electronics wholesale is the lead sector of European enforcement, with car trading and fuels close behind. But beware the reverse inference: recent investigations have reached household and hygiene products. Carousel risk is not confined to any product list.

Patterns. Certain commercial pictures recur across the case files: prices noticeably below market. Young suppliers with thin history and pressing deal speed. Chain transactions with unusual delivery routes. Payments to third-party accounts or via changing bank details. Margin-scheme invoicing on goods that are evidently new. Sudden, substance-free revenue growth at individual market participants.

And here is the necessary boundary — legal, not rhetorical: no single one of these features proves anything. A low price is, first of all, competition. A new supplier is, first of all, an opportunity. The Court of Justice held unambiguously in Mahagében (2012) that tax authorities may not offload their own investigative duties onto businesses, and that operating in a “risk sector” grounds no general suspicion. Warning signs are a reason to check — not evidence of participation. But the converse also holds: a warning sign that was visible and went unassessed becomes, two years later, the other side’s strongest exhibit. That is precisely the gap a traffic-light-based checking system closes.

System responsibility: the honest trader is not the system’s repair shop

Fighting VAT fraud is legitimate. But the context belongs in the picture: the carousel exploits a gap the system itself created. The law-and-economics analysis of the single market reaches a sobering conclusion: replacing physical border controls with administrative reporting and documentation duties raised compliance costs substantially without materially reducing the revenue risk. The border function migrated from customs posts to businesses’ documentation processes — and part of the risk migrated with it.

From this follows the standard the defence asserts in every case: system risk is not individual knowledge. The VAT gap, sector risk, the very existence of carousels — all of that explains the enforcement pressure. None of it proves anything in an individual case. What counts in the individual case is one question only: what did this business, in this transaction, at this moment know or have the means to know? How that test works — and who must prove what — is set out on the page “Knew or should have known”.

What honest businesses can do now

You cannot switch off the carousel’s mechanics — you can switch off your own vulnerability to them. Three steps carry far: know your own risk zones (goods, countries, deal patterns); assess warning signs systematically instead of ignoring them situationally; and document every check so that it still tells a coherent story two years later. How carousel-proof your supply chain is today is shown by the carousel risk self-check. If the German tax office, the tax fraud investigation unit (Steuerfahndung) or the EPPO has already made contact, the emergency page sets out the first steps — and they should not wait.

Prevention is the best defence. Not because it prevents every suspicion. But because it turns a suspicion into a question you can answer.

FAQ