Bearer Financial Instruments in the Offshore Leaks Database
Welcome to the first post of DATAINT, the blog that dares to ask the question, is there any topic more intrinsically exciting than obscure financial instruments and offshore accounting?
If you’re like me, and I hope for your sake that you aren’t, the answer is no. Tax evasion is in fact an evasion of one’s responsibilities under the intangible but still very real social contract. Evading taxes doesn’t make you smart, it makes you an asshole, someone who’s willing to call the police, but unwilling to pay for them. According the IRS the U.S. government was out about half a trillion (with a “tr”) dollars per year between 2008 and 2010. [Forbes, Chris Matthews, 04/2016]
Offshore accounting, stashing one’s money in low-tax low-accountability nations, is one of the major ways that rich and powerful people hide their wealth from the societies that support them. Thanks to the Offshore Leaks Database, a consolidation of four major leaks of offshore accounting information between 2013 and 2018 maintained by the International Consortium of Investigative Journalists, the general public has a view into how offshore accounting works. [ICIJ] The OLD is not a documents database, it is a network database. A network database is made up of nodes and edges. The nodes are entities, individual names, or locations. The edges are the connections between those nodes.
Network analysis software takes that data and allows you to make the node/edge visualization that we all know and love. I will creating such visualizations in subsequent entries, but for this initial exploration I will be sticking with pie charts, which by themselves reveal quite a lot.
Bearer Financial Instruments
More specifically, for this post I will be focusing on those nodes marked “the bearer” or some variation thereof. These nodes were created from financial documents relating to stocks and bonds that are legally owned by whoever physically holds a specific piece of paper. This means that someone could be the ultimate beneficiary of a corporation worth millions without having to mark their name down on any piece of paper. This is not the only or most common path to anonymity, shell companies with nominee directors are more common. But a shell company with a nominee director is usually created by a law firm specializing in such scuzzy arrangements, and that law firm will write your name down somewhere. Oh sure they might promise to keep it secret no matter which government or ex-spouse leans on them to do otherwise, but the loose end is still there. It’s confidential ownership, not anonymous, and the OLD proved there is a big difference between confidential and anonymous. Just ask Vladimir Putin, or Bono, or Shakira. [ICIJ, Scilla Alecci, 11/2017]
BFI have been falling out of favor in recent decades because they really are only good for three things: (a) tax evasion, (b) money laundering, or (c)movies about money laundering. [Investopedia, James Chen, 04/2019] Supranational organizations such as the OECD and the Financial Action Task Force (FATF) (which is essentially an arm of the OECD) have been trying to shame their member states into following certain transparency procedures regarding bearer shares namely either (a) banning them outright, or (b) making bearer shares “immobile.” [Offhsoreworld, Nida U., 05/2013] Immobilization means that a bearer share in kept with an accredited financial institution, and the name of the beneficiary owner is registered with that institution and thus subject to subpoena by angry governments or soon-to-be ex-spouses.
So that’s enough background information for one post (perhaps enough for several posts), let’s start digging into the data.
The Shape of the Data
When trying to get the “lay of the land” regarding BFI in the OLD we need to get an idea of which lands these bearer shares are actually from. In the OLD, most of the nodes representing people or entities don’t have a country associated with them, but some of them do. About 11% of the “bearers” in the OLD have countries associated with them (figure 1). The ICIJ says that it associates a person or entity with a country based on its registered address (if any).[ICIJ, “FAQ”] So the country of a “bearer” will match the country of the address that bearer is attached to.
Out of the 216 jurisdiction in the OLD, 120 of them have “bearers” associated with them. The mean value of the number of bearers associated with a country is 76, but the spread is so wide (1 at the low end and over 1000 at the high end) that a box plot is better to get sense of where most of the values lie (see figure 2).
So we can see that despite the mean value, most of the values cluster around a dozen. This is important to understand because it emphasizes the outlier-ness of the countries that have a lot of bearers associated with them. Those are jurisdictions that do something to attract those who wish to use BFI (e.g., those that really wish to hide their ownership or involvement in of a particular tranche of wealth), and that something works.
So who are those outlier countries? A pie chart (figure 3) shows that they aren’t quite who you’d expect. The name “offshore leaks database” brings to mind places like Monaco, sunny paradisiacal places without much economy aside from sheltering the money of the rich and/or powerful and/or criminal. But the two top jurisdictions in the OLD by a pretty hefty margin are Switzerland and Luxembourg, the paradisiacal island nations of Europe.
Swiss banking, as you may have heard, is protected by particularly strong privacy laws. For years the OECD, and its post 9/11 attack dog the FATF, have been pressuring the country to reform its banking laws so it’s not so easy for terror groups and tax evaders to hide their money there.[STEP, “Switzerland Bows to Pressure”, 11/2018] Currently bearer shares are permitted with some restrictions, so they are not immobile. [World Bank, “Guide to Beneficial Ownership Information in Switzerland,” 05/2018] A company can issue bearer shares to a person who can keep them wherever they chose, but the company must keep the holder’s name on file. The bearer share holder can transfer the shares to whomever they chose, but the recipient must report to the company that they are now the shares’ beneficiary within one month, or they lose the rights to the benefits of those shares.
If that seems less than iron clad to you, the OECD agreed, and in its 2016 peer review of Switzerland strongly hinted that if Switzerland didn’t tighten up even more on bearer shares then it risked being labeled as a “non-cooperative jurisdiction.” [OECD, Global Forum for Tax Transparency 2016 Peer Review Switzerland, 07/2016] The OECD’s complaints regarding bearer shares boiled down to Swiss law still having too many opportunities for beneficial owners to foot-drag on reporting their receiving of bearer shares. The Swiss government is, as I write this, passing laws to address the OECD’s complaints.
And that is it for this first entry of DATAINT (OSINT with graphs). Stay tuned for next week when I delve even deeper into the topic of BFI, a topic I am in no way weirdly obsessed with. I’m going to look into the question of whether or not jurisdictions with immobile bearer shares are storing those shares in reputable financial institutions like they’re supposed to. The answer is, they are, except when they’re not.
For each post on this blog that requires it I will be creating a Git Hub where you can reproduce the process for making the data frames and charts I used in the article. I do this because (a) I desperately want people to critique my code, and (b) I think it should be standard practice as complex analysis and visualization becomes more common.
Note on citations:
I will be experimenting around with different citation formats. I generally dislike the practice of linking in the sentence because I think it encourages the reader to wander off somewhere else instead of finishing the peice. I’m trying to find something that strikes a good balance between convienience of information for the reader, and maintaining the integrity of the text I’m trying to write.
