An Organized Opacity
The fog is not an accident. It is an architecture. For the past forty-eight hours, two camps have clashed over French banks’ exposure to the Middle East with almost comical symmetry: some brandish €132 billion as proof of impending apocalypse, others dismiss it as a harmless accounting artifact. Both have arguments. Neither is asking the right question.
Because the issue is neither the magnitude of the figure nor its supposed benignity. It is the impossibility, for anyone — citizen, parliamentarian, analyst — to access a faithful picture of risk at the precise moment when such clarity becomes essential. And within this opacity, the danger is not even where the debate is looking for it.
What the numbers say
The European Banking Authority has published its Risk Dashboard Q4 2025. Direct exposure of EU banks to the Middle East stands at €132 billion, including €60.8 billion for France (46%). Relative to the sector’s €29.1 trillion in assets, that is less than 0.5%. CET1 stands at 16.3%, non-performing loans at 1.8%, and major French banks posted €35 billion in profits in 2025.
Taken in isolation, these indicators are reassuring. And those who decry sensationalism are right on one essential point: exposure is not risk. The European banking system could absorb significantly larger losses without jeopardizing its solvency.
But the net figure does not say much either.
Netting assumes a world that does not exist
The EBA publishes gross exposures. Banks publish net figures, after offsetting through collateral — sovereign securities, guarantees, derivatives. The method is legitimate, compliant with IFRS standards. The differences are considerable: for BNP Paribas, one moves from €55 billion gross to roughly €6 billion net.
The problem is not the method. It lies in its assumptions.
What is missing is not data, but assumptions. Netting assumes that markets function normally. That collateral remains liquid. That flows do not freeze. That the Strait of Hormuz does not close.
In times of stress, the value of collateral is no longer a market price; it is a conjecture.
Netting protects against credit risk. Not against liquidity risk.
And it is liquidity that kills — quickly, without warning.
The only figure that matters in a crisis — net exposure after stress on collateral — is published by no one.
The same illusion applies geographically. A sovereign rating measures repayment capacity in calm conditions. It does not model logistical disruption, cross-sanctions, or a regional energy shock. This does not mean banks ignore these risks. It means the public has no way to assess them.
The real scandal
The problem is not the level of exposure. It is the absence of synthesis.
The EBA publishes gross figures, with a lag.
Banks publish net figures, according to their own methods.
The gap can reach a factor of ten.
No one reconciles them.
The result is an organized blind spot: everyone publishes a fragment of reality, no one makes it intelligible.
In that space, perceptions are manufactured: banks choose their narrative, the regulator avoids taking a definitive stance, the media simplify. And public debate oscillates between panic and denial.
When François Villeroy de Galhau asserts that financial stability is “not at risk,” he is probably right. But an assertion unsupported by verifiable data no longer generates trust. Since 2008, solidity is no longer proclaimed. It is demonstrated.
What should exist
The supervisor faces a real dilemma: too much transparency can amplify a crisis. But this dilemma does not justify the current opacity. On the contrary, it calls for aggregated, standardized, intelligible transparency.
Others have understood this. The Federal Reserve publishes detailed stress test results. The Bank of England regularly presents extreme scenarios. These exercises are not real-time. They are readable, comparable, debatable.
Europe could do the same.
What is missing is not an additional flow of data, but a public synthesis: a dashboard reconciling gross, net, and stressed net exposures under explicit scenarios. A document a parliamentarian can read, a journalist can verify, a citizen can understand.
The problem is not active concealment. It is institutional: no body is mandated to produce this readability.
The risk that matters
While the debate focuses on €132 billion, the real risk is shifting elsewhere. The EBA highlights this: second-round effects are the primary threat. A persistently high oil price, weakened growth, compressed margins — these diffuse mechanisms are what undermine balance sheets.
The next banking stress, if it occurs, will not stem from a sovereign default in the Gulf. It will emerge from a fabric of European companies, gradually weakened, increasingly unable to meet their obligations.
French banks are solid. The data show it.
But as long as that solidity remains partially opaque, every published figure becomes a source of confusion.
The problem is not the risk.
It is that it remains, structurally, difficult to see.
Dear readers,
This blog is yours: I maintain it diligently, with both consistency and passion. Thousands of articles and analyses are available to you here, some dating all the way back to 1993!
What were once considered heterodox views on macroeconomics have, over time, become widely accepted and recognized. Regardless, my positions have always been sincere.
As you can imagine — whether you’re discovering this site for the first time or have been reading me for years — the energy and time I dedicate to my research are substantial. This work will remain volunteer-based, and freely accessible to all.
I’ve made this payment platform available, and I encourage you to support my efforts through one-time or recurring donations.
A heartfelt thank you to all those who choose to support my work.