Author: admin

Stay informed with free updates

Simply sign up to the European companies myFT Digest — delivered directly to your inbox.

There are many different techniques of bank supervision. Historically, the North American approach emphasises on-site inspections, while Europeans make use of regular supervisory reports and validation of internal models and controls. Comparatively little use, however, has been made of “dystopian collaborative fiction writing workshops”. Until now!

The European Banking Authority carries out a full-dress stress test every other year to establish the banking system’s resilience to an economic downturn scenario. In the off-years, the ECB often does its own exercise, somewhat less demanding in scope and focused on a specific area of immediate concern. In 2024, for example, it was a climate risk scenario. Next year, the theme will be geopolitical risk.

And it’s going to be a “reverse stress test”. Rather than being given a scenario to deal with . . . 

. . . each bank will be asked to identify the most relevant geopolitical risk events that could lead to at least a 300-basis point depletion in its Common Equity Tier 1 (CET1) capital. In addition to reporting on how the geopolitical risk scenario would affect their solvency positions, banks will also be asked to provide information about how it may affect their liquidity and funding conditions.

So it’s not quite “imagine the end of the world”, but rather “imagine something that would be quite bad, but not very bad, say about 300 basis points of capital ratio bad”. The ECB says that it will publish the main conclusions in summer of 2026.

Unfortunately for Alphaville we shouldn’t expect a compendium of horrific 28 Days Later fantasies. Banks will tend towards prosaic but nonetheless real possibilities, like “a big tariff war”, “escalation of the Ukraine conflict” or “Donald Trump might sue us”.

Hopefully there will be some wacky ones in there too though. The ECB plans to administer this reverse stress test to 110 banks, so surely there will be some of them that employ wannabe Tom Clancy types in the risk management department. It could kick off a new genre of fiction; in these stressful days, it might be nice to be able to pick up a low-stakes thriller where you know that the eventual outcome will only be three percentage points of Common Equity Tier One capital won or lost.

But, of course, in this case the published results are not actually the important thing. One of the great fallacies of scenario analysis is attempting to get the exact right result. (This is in many ways an original sin of the discipline; scenario analysis really caught on after Shell was able to prosper in the 1970s because its earlier exercise had included something like the 1973 oil shock. This cemented the use of scenarios in corporate planning, but left lots of people hoping they’d be able score a hole-in-one like that every time).

The real purpose of reverse stress testing (and all kinds of scenario analysis) is twofold. First, to instil flexibility and responsiveness in the management system and to choose strategies which are robust to a wide set of possibilities rather than super-optimised for current conditions. And second, to test reporting and information systems, to be sure that they are capable of representing the variety of possible outcomes.

It’s this second function that the ECB is probably most interested in. As well as writing their nightmare journal, each banks is going to have to show how it produces a 300bp hit to capital. Which is not necessarily a trivial task. All the banks which are significant enough to be directly supervised by the ECB are meant to have “risk data aggregation and risk reporting” (RDARR) systems that are capable of carrying out scenario analysis, but this is the first time they’ve all been asked to prove it.

Of course, delegating the task to the banks also conveniently means that the ECB itself gets out of having to design a “geopolitical risk” scenario, which is more or less by definition a piece of work that’s going to annoy somebody powerful. And it acts as a distributed brainstorming session across the industry, potentially identifying geopolitical problems that hadn’t reached Frankfurt.

Most importantly, it might give us all something fun to read over the summer. More supervisors should do this!