Finance and banking: the sector that avoids tax and builds the structures for everyone else
ByLoopholeKiln EditorialPublished
Figures current as of·Corrections
← The sector playbooks: each industry's signature tax move, in one place
Banking sits in a unique spot in this story, because it plays two roles at once. A bank uses avoidance techniques on its own profit, like any other multinational. But a bank is also the infrastructure through which every other sector runs its structures: it holds the offshore accounts, builds the special-purpose vehicles, designs the hybrid instruments and books the intra-group loans. To understand banking and tax you have to look at both roles, and this page does. The signature move on its own account is the "ghost operation"; the bigger story is that finance is the plumbing.
The tools underneath are debt and interest, hybrid mismatches and conduit jurisdictions, all explained in general elsewhere. What is specific to finance is that money is the product, which makes internal lending uniquely powerful, and that the sector is the enabler for everyone else.
The signature move: booking profit where nobody works
On its own account, the bank's signature move is the ghost operation: booking large profits in jurisdictions where it employs no staff at all. Transparency International EU examined the financial data of 39 major EU and UK banks from 2015 to 2019. Of those, 31 were using low-tax or zero-tax havens, and 29 appeared to be declaring high profits in places where they employed nobody. Profit generated by lending and capital-markets activity in high-tax countries, the UK, Germany, France, is reallocated on paper to Luxembourg, the Cayman Islands, Jersey or the Bahamas through internal pricing.
The headline number for what the sector books offshore: European banks reported an average of about EUR 20 billion a year of profit in tax havens between 2014 and 2020, which is 14% of their total profits. The spread was wide: the most concentrated bank in the study booked on average 58% of its pre-tax profit in low-tax jurisdictions over that period, much of it in a single home market.
Why internal lending is the bank's natural lever
Every sector can load a subsidiary with group debt and strip out the profit as deductible interest. But banking is the sector where this lever is most powerful, because lending is the core business and the volume and variety of intra-group financial flows is vast, which makes the flows extremely hard for a tax authority to price one by one. A subsidiary in a high-tax country borrows from a related entity in a low-tax one; the interest is deductible locally and lightly taxed where it lands.
The scale shows up in the research. A 2023 academic study found that, for banks, tax havens account for more than 20% of all cross-border debt financing. A CEPII gravity-model analysis, which compares where a bank's activity actually is against where business volume alone would predict it should be, found one major UK bank with more than 50 subsidiaries in the Cayman Islands, two in Panama and five in the Bahamas, a presence the study judged out of proportion to its real business there; on the study's measure, British and German banks showed the most haven-weighted footprints.
A more technical lever specific to finance is the hybrid instrument: a financial product treated as debt in one country, so the payment is a deductible interest expense, and as equity in another, so the receipt is an exempt dividend. The same payment generates a deduction in country A and tax-free income in country B, a genuine double non-taxation. Banks, whose business is designing financial instruments, are unusually well placed to build these. The OECD's BEPS Action 2 targets them, but implementation is uneven.
The bigger story: finance as the plumbing of avoidance
This is the part that makes the finance sector different. Beyond its own tax position, banks and financial intermediaries are the infrastructure of avoidance for every other sector. They operate the offshore accounts, structure the special-purpose vehicles, design the hybrid instruments and run the intra-group lending that the tech, pharma, extractive and retail structures all depend on. Offshore financial centres exist in the first place because international banks accept deposits from non-residents with minimal disclosure, channel funds through special-purpose vehicles, and use favourable treaties to return profits clean. CEPII estimated that global banks intermediate around EUR 550 billion in offshore deposits, equivalent to about 5% of their home countries' GDP, structurally connected to tax minimisation.
There is some evidence the disclosure rules bite. Public country-by-country reporting, required for EU banks from 2015, has produced a measurable effect: one study found a decline of up to 33% in the number of subsidiaries in tax and regulatory havens after mandatory public reporting came in. But voluntary disclosure still conceals most of the mechanism, which is why the enabling role is so hard to measure from the outside.
Why a local lender is not in this game
A building society or a small regional lender borrows and lends in one country, at real market rates, and deducts real interest. It has no captive group entity in a low-tax country to route interest to, no offshore subsidiaries booking phantom profit, and no business designing hybrid instruments across borders. The signature moves here need a multinational banking group with entities on both sides of every internal loan. They are not available to a domestic lender, and they are certainly not available to the small businesses that bank with one.
Sources
- 01Transparency International EU, *Murky Havens and Phantom Profits: the tax affairs of EU and UK banks* (2024, 2015-2019 data) (39 banks; 31 used havens; 29 ghost operations)
- 02EU Tax Observatory, *Have European banks left tax havens?* (Sept 2021) (EUR 20bn/yr = 14% of profits; HSBC 58%)
- 03CEPII, *Banks Defy Gravity in Tax Havens* (WP 2017-16) (Barclays 50+ Cayman subsidiaries; British and German banks most aggressive)
- 04OECD, *Neutralising the Effects of Hybrid Mismatch Arrangements*, BEPS Action 2