AML compliance costs draw scrutiny as recovery rates remain low

Global financial institutions spent $206.1 billion on financial-crime compliance in 2023, while published estimates suggest authorities recover only a small share of laundered funds. In a Finextra opinion, Oleksandr Tseitelman, director at MONTOWIRE MSB LTD, argued that the economics of anti-money-laundering enforcement require closer examination by regulators, banks and policymakers.

Tseitelman cited LexisNexis Risk Solutions’ True Cost of Financial Crime Compliance report for the $206.1 billion figure. According to the same report, Europe, the Middle East and Africa accounted for $85 billion of that spending, while the US and Canada accounted for $61 billion. LexisNexis also found that 98% of surveyed institutions reported annual increases in compliance costs.

The scale of the activity being monitored is larger still. Tseitelman cited estimates from the UN Office on Drugs and Crime and the International Monetary Fund that money laundering represents 2% to 5% of global gross domestic product, or about $800 billion to $2 trillion a year.

Recovery rates remain contested

Measures of enforcement output vary by jurisdiction and methodology. Tseitelman cited a UNODC benchmark that authorities seize “much less than 1%” of laundered proceeds, probably around 0.2%. He also cited Europol’s estimate that 1.1% of criminal proceeds are confiscated in the European Union.

He pointed to a 2020 peer-reviewed paper by Ronald Pol in Policy Design and Practice, titled “Anti-money laundering: The world's least effective policy experiment?” Pol estimated that the system intercepts roughly $3 billion of about $3 trillion in annual criminal funds, equal to a 0.1% success rate, while compliance costs exceed $300 billion. Pol wrote that compliance spending exceeds recovered criminal funds by more than 100 times, according to Tseitelman’s account.

AML systems require financial institutions to screen customers, monitor transaction patterns, raise alerts and file suspicious activity reports. Those reports can support investigations, but they are also an input measure. Tseitelman argued that the regime often counts reports filed, alerts generated and analysts hired more readily than it demonstrates reductions in underlying criminal activity.

Data quality questioned

Tseitelman also questioned the reliability of both sides of the calculation. He noted that the Financial Action Task Force has said reliable global statistics are not available and that a definitive estimate of money laundered worldwide cannot be produced. He argued that broad estimates such as 2% to 5% of GDP rely partly on statistical models because the true amount is not directly observable.

On recoveries, he said the measurement can also be distorted when frozen accounts or suspicious business turnover are treated as interdicted criminal proceeds. He compared this with narcotics enforcement, where seizure values may be reported using estimated “street value” rather than realised proceeds.

The critique also referred to the volume of suspicious activity reporting. Tseitelman wrote that researchers have warned that large numbers of filings can create noise that makes it harder to identify laundering activity. He cited the US as an example, saying AML work costs several billion dollars a year and produces fewer than 700 money-laundering convictions.

Tseitelman said the argument was not against compliance itself, but for more rigorous assessment of its costs and results. He also cited remarks by David Lewis, then executive secretary of the FATF, at Chatham House in 2021, where Lewis said he was “fed up with protecting the integrity of the financial system” and that he did not care about that formulation of the mission.

The defence of AML frameworks, as Tseitelman described it, is that their value lies partly in deterrence and intelligence rather than seizures alone. His central contention is that, after more than 30 years, the regime should be judged against measurable outcomes as well as institutional activity.

This story draws on original reporting from Finextra Research.