Bitcoin Tax Evasion Can’t Be Stopped With Just Anti-Money Laundering: IMF Report

Anti-money laundering (AML) rules are not a panacea for dealing with tax cheats and criminals who try to cover their tracks with crypto—but they’re a logical place to start, according to a paper published by the International Monetary Fund (IMF) on Wednesday.

The paper was authored by members of the IMF’s Fiscal Affairs Department: Katherine Baer, Ruud de Mooij, Shafik Hebous, and Michael Keen. It includes a disclosure that the views expressed by its authors do not necessarily reflect the views of the IMF.

From a taxation perspective, the paper’s primary concern is that digital assets provide a new and potent way for criminals and the rich to conduct transactions undetected. And its authors recognize tens of billions of dollars in potential tax revenue are at stake, with no consensus across the globe on how the issue should be approached.

The authors say plainly that they don’t intend to “provide policy prescriptions,” but also write that governments can look to regulations and laws present in the U.S. as a guide to stop financial crimes and illegal activity.

“Whether crypto withers or blossoms, the tax system still needs to deal with it,” the IMF authors said. “The first step for governments, nonetheless, is to apply AML rules and third-party reporting requirements where they can.”

In terms of AML rules, the paper references guidance released by the Financial Action Task Force in 2015—meant to serve as a global standard for combating money laundering—but acknowledges that not all jurisdictions fully comply.

The paper notes that centralized institutions like exchanges are in a unique position to help authorities gain information on digital asset ownership, often serving as the touch point where cash is exchanged for crypto, and capable of tracking activity beyond that.

The IMF paper estimates that a global 20% capital gains tax in 2021 might have raised around $300 billion from crypto-related transactions. Still, the authors said Know-Your-Customer (KYC) procedures—which help them remain compliant with anti-money laundering regulations—aren’t enough to paint the whole picture for tax authorities.

“KYC rules might enable the authorities to know, for instance, that some individual cashed out a certain amount of cryptocurrency,” the authors said. “But from the transactions prior to that recorded on the blockchain it will not be possible, without further information, to identify any associated capital gain or loss.”

In this sense, the technology underpinning many cryptocurrencies could actually be a boon for tax authorities, the authors wrote, stating blockchains are “remarkably transparent in the information they contain on the history of transactions.”

Artificial intelligence could be used in some capacity “to identify potentially tax-relevant behaviors” that take place on-chain, the IMF authors said, describing the vast amount of public information on networks as ripe for forensic analysis.

Some critics, however, like the whistleblower Edward Snowden, have critiqued exchanges, specifically Coinbase, in the past for over-compliance and called it a drag on the crypto space.

But in the event that governments embrace AML and KYC, the paper details other challenges that could emerge if it pushes bad actors in the direction of decentralized exchanges, where it would be harder to glean information with no one to abide by reporting requirements.

And the main obstacle outlined in the paper would still remain, which is the element of pseudo-anonymity associated with cryptocurrency transactions. Digital wallets are composed of public and private keys. If an entity doesn’t link a person’s name to a digital wallet, it becomes more challenging for authorities to know who it belongs to, unless a person divulges it themselves.

“Overcoming pseudo-anonymity is the core problem that tax administrations are now trying to address,” the IMF said. “In the old days, the tax authorities’ problem was that it knew who you were, but not your income; now the problem is that it knows your income but not who you are.”