Vocabulary, typologies and the IFF landscape

I begin every cohort with definitions, and I am aware that this bores people who came for the case studies. The reason I insist on it is that most of the weak files I have reviewed at SARS and elsewhere were weak because somebody charged the wrong thing at the outset. Money laundering is the process by which the proceeds of predicate crimes are made to appear legitimate. The legal definition adopted by the Financial Action Task Force (FATF) and reflected in almost every domestic anti-money-laundering (AML) statute has three constitutive elements. First, an underlying predicate offencePredicate offenceThe underlying crime whose proceeds are being laundered. — drug trafficking, corruption, fraud, tax crime, human trafficking, arms smuggling, environmental crime, cybercrime, or a schedule of comparable serious offences defined in national law. Second, proceeds derived from that offence — funds, property, rights, or any form of economic advantage traceable to the crime. Third, an act of conversion, transfer, concealment, acquisition, possession or use undertaken with knowledge, or reasonable grounds for suspicion, that the property is criminal in origin, and undertaken with the purpose of disguising that origin. The trilogy is deliberately broad: it captures the mule who deposits, the accountant who structures the invoice, the lawyer who conveys the property, and the buyer of the ultimate luxury car.

Money laundering is a stand-alone offence. This matters enormously. In classical criminal law, one prosecuted the underlying theft or trafficking. Under modern AML statutes, one can prosecute the laundering even where the predicate cannot be proved to criminal standard, occurred abroad, is time-barred, or was committed by an unknown person. This "autonomous offence" doctrine, endorsed by the FATF, embodied in the Palermo and Merida Conventions, and now standard in most jurisdictions — is what allows a national prosecutor to indict a domestic accountant for laundering the proceeds of a foreign bribe.

Sitting alongside money laundering is the related but distinct concept of Illicit Financial Flows (IFFIFFIllicit Financial Flows — cross-border money that is illegally earned, transferred or used.), as defined by UNCTAD and the UN High-Level Panel on IFFIFFIllicit Financial Flows — cross-border money that is illegally earned, transferred or used. from Africa (the Mbeki Panel of 2015). IFFIFFIllicit Financial Flows — cross-border money that is illegally earned, transferred or used. encompasses ALL cross-border movements of money that are illegally earned, illegally transferred, or illegally used. That definition is deliberately wider than the criminal AML frame. It captures commercial IFFIFFIllicit Financial Flows — cross-border money that is illegally earned, transferred or used., trade mis-invoicing, transfer mispricing, abusive use of hybrid instruments, artificial permanent-establishment avoidance, and base erosion — alongside the more familiar categories of criminal proceeds, corruption, and terrorist financing. It even reaches lawful commercial arrangements whose cumulative effect is to strip a developing country of tax base.

These four categories share a single laundering infrastructure. All of them move through the same plumbing: retail banks, private banks, correspondent networks, corporate service providers, trust and company service providers, notaries and conveyancers, real-estate agents, luxury-goods dealers, art auction houses, casinos, precious-metal traders, and — with rising prominence — virtual-asset service providers, decentralised exchanges, and crypto mixers. That shared plumbing is exactly why AML frameworks bind the four categories together operationally even when criminal law treats them separately: the same customer due diligence, the same suspicious-transaction reporting duty, the same politically-exposed-persons regime, the same enhanced monitoring for high-risk jurisdictions.

The distinction between the categories matters at three levels. At the level of prosecution, only categories 1, 2 and 4 sustain criminal charges without more; category 3 requires either a specific anti-abuse rule or a re-characterisation. At the level of quantification, IFFIFFIllicit Financial Flows — cross-border money that is illegally earned, transferred or used. estimates that lump the four together (as GFI does) produce very different headline numbers than tax-gap estimates that count only (3) and (4). At the level of policy, remedies diverge sharply: prosecutions and asset recovery for (1) and (2); tax reform, treaty renegotiation and transparency mandates for (3) and (4).

For developing countries this matters materially. IFFIFFIllicit Financial Flows — cross-border money that is illegally earned, transferred or used. has, in every recent estimate produced by the OECD, GFI, UNCTAD and the African Development Bank, exceeded inbound official development assistance and, in many years, inbound foreign direct investment. UNCTAD's 2020 Economic Development in Africa Report estimated USD 88.6 billion in annual capital flight from Africa alone, most of it via commercial IFFIFFIllicit Financial Flows — cross-border money that is illegally earned, transferred or used. channels. Every dollar successfully hidden or shifted is a dollar unavailable for schools, hospitals, infrastructure, and climate adaptation. The macroeconomic effect is asymmetric: the private beneficiary is typically resident or beneficially entitled in a low-tax high-institutional-quality jurisdiction, while the fiscal harm is borne by a state already fiscally stretched.

For a practising official, the first analytical move on any new file is therefore triage: which of the four categories does this fit? That single question dictates the powers you can invoke, the counterparts you must engage, and the outcome you can realistically pursue. A drug case invokes the police and the FIU; a bribe case invokes the anti-corruption commission, the FIU, and (via UNCAC Chapter V) foreign counterparts for asset recovery; a transfer-pricing case invokes the tax administration, the exchange-of-information competent authority, and, if you have a GAAR; the tribunal that hears anti-abuse cases. Getting the triage wrong burns time and can jeopardise admissibility.

A final conceptual point. What is NOT laundering, and why the boundary matters. Simple concealment of lawful income (for tax reasons only) is tax evasion, not laundering, unless a predicate tax offence is defined and the subsequent movement of the untaxed funds meets the laundering elements. Handling stolen goods without the concealment purpose is a distinct offence in most systems. Cash-carrying across a border without declaration is a currency-reporting offence and, only where the origin is criminal, becomes laundering as well. Keeping these boundaries clean prevents the common prosecutorial error of loading multiple, overlapping charges that end up unravelling on appeal.

A colleague once told a seminar in Bochum that the three-stage cycle was "the only thing anyone remembers from AML training". He meant it as a criticism. I think he was half right: the model is remembered because it is useful, and misused because people mistake it for law. The classical three-stage model of the laundering cycle (placement, layering, integration) is the most useful mental scaffold in the field. It was first formalised by U.S. law enforcement in the late 1980s and adopted by FATF as a pedagogic device. It is not a legal test and not a doctrine, but the model organises typologies, red flags, evidentiary needs and enforcement responses so effectively that no practising investigator dispenses with it.

Structuring, or smurfing, breaks a single cash sum into multiple deposits, each below the mandatory reporting threshold (typically USD/EUR 10,000, sometimes lower). Smurfs may be recruited from vulnerable populations — students, unemployed workers, migrants, and paid a small fee per deposit. Cash-intensive front businesses; car washes, restaurants, night-clubs, casinos, taxi fleets, retail petrol, commingle illicit cash with declared takings, then bank it as legitimate revenue. Currency-exchange bureaux, especially in border towns and tourist zones, accept large cash swaps into foreign currency or into anonymous instruments (traveller's cheques where still available, prepaid cards, gold). Money mules — often recruited via social-media job scams — accept a deposit into their personal account and forward it minus a fee. Placement can also occur through the deliberate over-payment of tax, credit-card bills, or utility accounts, generating a refund from a legitimate counterparty that arrives as a clean cheque.

The cycle is idealised. Real laundering schemes skip stages, revisit them, or run multiple cycles in parallel. A drug trafficker whose customers pay in cryptocurrency may skip placement entirely. A corrupt politician whose bribe was paid by wire to an offshore account skipped placement and began at layering. A tax evader may need only integration; retrieving offshore accumulated income into a personally usable asset, without any prior placement in the classical sense. Nevertheless every investigator's mental map begins with these three stages, because red flags cluster very differently at each, because different sectors are exposed at different stages, and because the legal powers deployed against each stage differ.

Red-flag clustering is worth memorising. Placement red flags are almost all volumetric and behavioural at the teller line: sub-threshold structuring, unusual denomination mixes, unfamiliar customers depositing on behalf of others, refusal to complete a currency-transaction report. Layering red flags are almost all relational: unusual counterparties, high-risk jurisdictions in payment chains, pass-through accounts with negligible ending balances, complexity disproportionate to business turnover. Integration red flags are almost all valuation-and-provenance: purchases at odd multiples of market value, thin documentation of source-of-funds, rapid asset flipping to launder a spread.

The three-stage model also disciplines the enforcement response. Placement is where the retail bank, the money-remitter, the casino cage, and the customs post act. Layering is where the FIU, the correspondent-bank compliance function, and the corporate-registry investigator act. Integration is where the notary, the real-estate registrar, the tax administration and the asset-recovery lawyer act. A well-designed national AML regime therefore instruments each stage with the appropriate actor, the appropriate reporting duty and the appropriate onward-referral pathway. A regime that concentrates all attention on placement (as many first-generation regimes did) predictably fails as launderers migrate up the cycle. A regime that concentrates on integration alone (as some real-estate reforms attempt) predictably fails to prevent the injection in the first place. The professional investigator therefore reasons in both directions: given a placement pattern, what layering and integration is likely downstream? Given an integration event, what layering and placement upstream fed it?

Officials often arrive on this programme unsure which of the acronyms in their inbox actually binds them. Almost none of them do, directly. What follows is a map of who writes what, and how it reaches your desk. Four bodies dominate the international standards landscape on financial-crime, tax transparency and asset recovery. Every practising official must know their remit, their instruments, and the practical channels through which they touch domestic work.

THE FINANCIAL ACTION TASK FORCE (FATF). Founded at the 1989 G7 Paris Summit in response to the drug-trafficking crisis, FATF is the global standard-setter for AML/CFT/CPF (anti-money-laundering / counter-terrorist-financing / counter-proliferation-financing). Its principal product is the 40 Recommendations, revised in 2003, 2012, and repeatedly updated since (notably Rec. 15 on virtual assets in 2018–2019 and Rec. 24 on beneficial ownership in 2022). The Recommendations are supplemented by the 11 Immediate Outcomes used to score effectiveness. FATF evaluates member and observer countries through a rigorous four-year Mutual Evaluation cycle covering technical compliance (do you have the rule?) and effectiveness (does it work?). Countries with strategic deficiencies are placed on the "grey list" (Jurisdictions under Increased Monitoring) or, in extreme cases, the "black list" (High-Risk Jurisdictions subject to a Call for Action). Grey-listing has measurable negative macroeconomic effects — the IMF has documented capital-inflow reductions averaging 7.6 percent of GDP for grey-listed countries — which is why FATF's soft-law influence functions in practice like hard law. FATF also runs typologies projects (with the Egmont Group), produces guidance for sectors (real estate, dealers in precious metals and stones, virtual-asset service providers), and works through nine FATF-Style Regional Bodies (FSRBs): APG, CFATF, ESAAMLG, GAFILAT, GABAC, GAFI-MOAN (MENAFATF), Council of Europe Committee MONEYVAL, EAG, GIABA. Every developing-country official interacts primarily with their FSRB, whose mutual evaluations shape the domestic reform agenda.

THE ORGANISATION FOR ECONOMIC CO-OPERATION AND DEVELOPMENT (OECD). Where FATF governs the AML channel, OECD governs the tax-transparency channel. Its instruments include the Common Reporting Standard (CRSCRSCommon Reporting Standard, automatic exchange of financial account info between tax authorities.), which mandates automatic exchange of financial-account information between over 120 participating jurisdictions on an annual basis; the Convention on Mutual Administrative Assistance in Tax Matters (MAAC), the multilateral treaty that provides the legal basis for CRSCRSCommon Reporting Standard, automatic exchange of financial account info between tax authorities., spontaneous exchange, exchange on request, joint audits and mutual assistance in collection, and the BEPS Project (Base Erosion and Profit Shifting), a 15-Action programme adopted in 2015 that tightened rules on transfer pricing, treaty shopping, hybrid mismatches, interest deductibility, controlled-foreign-company regimes, and country-by-country reporting. BEPS 2.0 (2021 onwards) added Pillar One (partial re-allocation of taxing rights on the largest MNEs) and Pillar Two (a 15% global minimum effective tax rate for MNE groups above EUR 750m). The Global Forum on Transparency and Exchange of Information for Tax Purposes runs peer reviews on the exchange-of-information standard analogous to FATF Mutual Evaluations. The OECD's Anti-Bribery Convention (1997) criminalises the bribery of foreign public officials by nationals and companies of party states and is enforced by the OECD Working Group on Bribery.

FOR A DEVELOPING-COUNTRY PRACTITIONER, the operational reality is this: your legal AML regime borrows heavily from FATF templates and is scored against them at your next FSRB Mutual Evaluation; your tax administration's cross-border information rights depend on OECD instruments, chiefly the MAAC and the CRSCRSCommon Reporting Standard, automatic exchange of financial account info between tax authorities.; your asset-recovery capability depends on the UNCAC framework, StAR support, and bilateral MLATs; and, if you send trade or capital to the EU, your firms are indirectly subject to EU rules through their counterparties. These regimes are increasingly interoperable, the FATF now cross-references OECD BEPS work in its tax-crime typologies; the EU AMLA will coordinate with FATF; the StAR Initiative works with FATF, UNODC and the Egmont Group — but their operating cultures are distinct. FATF is technical, adversarial, and evaluation-driven. The OECD is policy-driven and negotiation-heavy. The UN is politically representative and consensus-slow. The EU is legislative and directly binding on member states.

A final practical point. Standards are only as good as the domestic transposition and the operational implementation. Country after country presents FATF-compliant statutes and near-zero convictions, or CRSCRSCommon Reporting Standard, automatic exchange of financial account info between tax authorities.-participating status and near-zero use of the data by tax administrators. The reforming official's job is not to import a standard — it is to implement it. That means budget, staff, IT, inter-agency memoranda, judicial training and, above all, political cover for prosecutions that hurt the powerful. The four bodies above give you the vocabulary and the template; your national politics decides whether the template works.