What lies beneath: evidence from leaked account data on how elites use offshore banking

By Matthew Colin

Wealth can travel around the world with relative ease, but a government’s ability to keep track of it is largely confined to its own borders. It is this information asymmetry that gives rise to a host of problems, the first of which is the ability of people to evade taxation by parking their wealth in other jurisdictions, away from the prying eyes of the tax authority. Recent evidence suggests that random audits of wealthy individuals routinely underestimate tax evasion by those who keep some of their money offshore. Not only is the amount of offshore wealth that goes undeclared to tax authorities likely to be substantial but the fact that the rich are more likely to engage in this behavior suggests that efforts to curb inequality through taxation be ineffective until authorities improve their ability to observe their taxpayers’ worldwide income and wealth.

A second, interrelated, problem is the fact that political elites and their associates who wish to abuse their public office or political power for private gain can pull this off more successfully when the gains are accrued offshore. It is easier to deny you have amassed a lot of wealth while in office when your government lacks the ability to observe what you own overseas and there is no local, conspicuous pot of money that someone can point to. This behavior is rife enough to be observable in international statistics: when high-corruption countries receive an influx of oil-revenue or foreign aid, offshore wealth owned by its residents surges.

The presence of offshore financial centers (OFCs) – colloquially known as tax havens – complicates the information asymmetry problem even more. Historically, many havens have offered both a low level of taxation and a high degree of financial secrecy: laws and policies that make very difficult for anyone, governments included, to understand who owns what. Ironically, the most concrete evidence that many OFCs have acted as havens for tax evaders has been generated by their gradual embrace of transparency: several studies have shown that foreign deposits held in these jurisdictions decline sharply after they sign agreements to share information on these deposits with foreign tax authorities.

Today, traditional small-island havens have been joined by some of the world’s biggest economies – such as the US – as the destination for untaxed and corrupt wealth. In both old and new destinations of illicit finance, multiple, overlapping legal structures that obscure the true owner of wealth (such as shell companies or trusts) are a common means to maintaining secrecy.

Researchers face the same information asymmetry problem that governments do: our attempts to understand how people evade tax and how they engage with the offshore financial center are hampered by the fact that most of this behavior is hidden by its very nature. Most research thus relies on either publicly-available, aggregate statistics or confidential data from tax authorities to understand how the offshore world functions and how taxpayers interact with it. These approaches are highly useful, but they are also incomplete: trying to understand what is happening to offshore wealth via these methods is analogous to sitting on a boat and inferring what is happening underwater by observing how the waves change (or sporadically interviewing people who swim to the surface).

Occasionally though we get a peek beneath the waves: over the past decade there has been a surge in high-profile leaks of data from banks based in tax havens and from the service providers that help people and companies organize their affairs offshore. The most high profile of these was the leak of customer information from the Panamanian law firm Mossack Fonseca (the Panama Papers). These offer an unprecedented opportunity to better understand how people hide their money in tax havens. Yet, possibly because they are often limited in both their scope and in what they can measure, few studies have relied on aggregated data from these leaks to directly understand the impact of policy.

On November 16th, 2019, the journalist collective Distributed Denial of Secrets (DDOS) announced it had made available data obtained from the Cayman National Bank and Trust, Isle of Man (henceforth CNBIOM) a subsidiary of the Cayman Island-based Cayman National Corporation. The data has purportedly been obtained by a hacker or hacker collective known as Phineas Phisher, who hacked CNBIOM and turned over the data to DDOS.5 DDOS dubbed the leaks “Sherwood” and made CNBIOM’s data available to the public in two formats: one was a searchable database of files and e-mails hosted on their “Hunter Memorial Library” platform, a search engine where files can be looked up via their contents. The second format was the leaked data in its entirety: the contents of several dozen hard drive images taken from CNBIOM’s servers.

Those hard drive images contained both working documents maintained by CNBIOM staff as well as snapshots, from 2016 and 2019 of the databases the bank used to maintain its customer records. From those databases I have reconstructed monthly and quarterly account estimates, in USD, for every customer from 2008 until late 2019. I can also observe the date of each account’s closure, as well as the date when any given customer closes their last account with the bank. I can also observe precise changes to account balances: for example, whether an increase in the balance is due to interest, capital gains, wire transfers in and out, or inter-account transfers. From this data, I establish three descriptive results.

The first is that the clients of offshore banks are likely to be rich, both in that they reside in rich countries and that they are likely to be from the upper-end of the income and wealth distribution in those same countries. While this is not a novel, earth-shattering finding, it reinforces growing evidence coming out of a number of countries that offshore banking is a luxury service and thus, to the extent that it is being used to evade or avoid tax, it has major implications for the ability of societies to fight inequality and raise revenue through progressive taxation.

The second is that a significant share of offshore deposits are controlled by clients who are connected to politically-influential people, such as current and former politicians, their family and friends. A number of these are from countries that score poorly on popular measures of corruption risk. These politically-exposed clients not only hold greater amounts of wealth than the average offshore client, but they are also more likely to have an address in a tax haven and are more likely to receive money into their accounts from accounts in other tax havens, all signs that significantly raise the risk that these clients have access to higher levels of wealth and are more willing to engage in activity that obscures their ownership of it. This is despite the fact that these clients were all successfully identified and were monitored by the bank in question.

The third result is that the legal arrangements that most customers of offshore banks used to organize their affairs, shell companies and trusts, interfere with the ability of popular measures of cross-border liabilities to measure offshore wealth. A growing amount of work on the level and determinants of offshore deposits relies on data from the Bank of International Settlements’ Locational Banking Statistics (BIS LBS), which assigns ownership of deposits based on the immediate counterparty. In practice, this means that deposits owned by shell companies based in tax havens will be assigned to those havens, rather than to the location of the ultimate owner, a limitation widely recognized (and occasionally exploited) by researchers working in this space. Using data on the ultimate owner of banks deposits from the leak combined with reports prepared for the LBS, I show that over half of deposits are controlled by entities based in tax havens and that many of these are incorrectly assigned as being haven deposits, despite being controlled by residents of non-haven jurisdictions. This not only implies that our measures of non-bank cross-border liabilities may be distorted by the presence of tax havens, it might also bias studies on the impact of transparency initiatives in ways that are hard to bound.

The data also highlights governments’ struggle to know where money flows outside their own borders, and the ability of individuals to park cash in the guise of an offshore company or trust makes this all the more difficult. It also makes it more difficult to discern the illicit from the licit. A single bank like CNBIOM can cater to many legitimate business interests and individuals who have honest reasons to keep their money offshore. But even when banks work hard to turn away or report on customers who are likely to be abusing the system (as was evidently the case here), they will inevitably host wealth which is illicit, either because someone should have paid tax on it or because it was stolen from someone else.

Shell companies and trusts form a kind of black-out curtain that hides what is a constellation of different actors with wildly different interests, including pro golfers, political elites being tried for corruption, and large German payment companies brought down by fraud. The sheer number of corruption scandals linked to large international banks over the past decade suggests that banks themselves are never going to be completely successful at self-policing, and that more needs to be done to help tax authorities and law enforcement agencies to understand – en masse – who controls offshore wealth, rather than in the piecemeal fashion that most authorities share information with one another.

The advent of automatic-exchange-of-information will have helped with this tremendously. But successfully lifting the veil of secrecy will likely require more substantial interventions, such as the establishment of beneficial ownership registries or better regulation of corporate service providers.

You can read the full report here.