France is Burning
Sun 11:03 am +00:00, 18 Oct 2026
Source: https://www.winterwatch.net/2026/10/france-is-burning/
The coat of running a low trust society and too late to correct. Chaos is starting to spread to Italy. Late stage collapse behavior on an epic level.
ChatGPT:
The unrest in France and Italy belongs on our Sicario compound-shock screen. Unlike a hurricane, this is a man-made catastrophe exposure, with potentially recurring losses and a substantial burden falling outside insurance.
The important distinction is that the riots are verified, but their total insurance cost has not yet been established. We should not confuse photographs of extensive destruction with an actuarially established loss.
1. What has happened — October 17–18
Reuters reports widespread violence accompanying demonstrations in France, with approximately 1,200 arrests, more than 150 police officers injured, and serious damage to shops, vehicles and municipal buildings. Limoges city hall was set ablaze; Lyon’s tourist office was also attacked. Most demonstrators were peaceful.
In Italy, AP confirms looting, fires and clashes in central Milan, including damage around the city’s important commercial and tourist district. Approximately 50 people were detained.
The simultaneous outbreaks matter, but they do not establish a single coordinated campaign across countries.
2. What might the insurance bill be?
The most useful historical comparison is France’s 2023 riots.
2023 France insured losses
€793M
Reported claims
16,400
French Senate / France Assureurs figures, reported through March 2024.
For the current France–Italy unrest, these are illustrative scenarios, not loss forecasts:
| Scenario | Combined insured loss |
|---|---|
| Contained within days | €100–300 million |
| Sustained nationwide violence | €500 million–€1 billion |
| Repeated major episodes and widespread arson | €1–2 billion+ |
The upper scenario requires much more destruction than has yet been quantified. These figures exclude uninsured business interruption, lost tourism, police overtime and public security spending.
3. Who actually takes the hit?
There’s a particularly important development in France.
France enacted a new riot-insurance framework in its February 2026 finance law. It provides for broader riot-related property coverage and a mutualization fund backed by the French state. Under the statutory design, insurers can transfer qualifying riot risks to the fund, which would reimburse at least 90% of covered losses ceded to it. The state guarantee is subject to an annual ceiling no greater than €4 billion, with further conditions.
However, the law makes implementation dependent on further regulatory and EU state-aid steps. I have not verified that the fund was operational for October 17’s losses. We cannot yet treat that €4 billion ceiling as cash already available.
The loss distribution looks like this:
| Who pays | Principal exposure |
|---|---|
| French property and motor insurers | Insured arson, looting, vandalism and property damage |
| French riot mutualization fund / state | Qualifying transferred losses, if the mechanism is operative |
| Italian property insurers | Riot and civil-commotion losses where policies cover them |
| European reinsurers | Only where applicable riot/strike/civil-commotion treaties respond |
| Municipalities and national governments | Public buildings, policing, emergency services, uninsured repairs |
| Businesses and households | Deductibles, exclusions, lost sales, uninsured assets |
For publicly traded companies, I would examine AXA, Allianz, Generali and Zurich Insurance Group first for direct European property exposure. Munich Re, Hannover Re and SCOR are relevant on the reinsurance side.
That is a list of potential exposures, not evidence that any has suffered a material net loss. Your American insurer shorts are much less obvious direct targets.
4. The larger financial-system implication
Your low-trust-society observation raises a serious economic question: what happens when civil disorder becomes a recurring cost of doing business?
France’s own state-backed reinsurer, CCR, has warned that riot risk is becoming more frequent, costly and systemic, making conventional diversification more difficult.
The transmission mechanism is important:
Repeated riots, arson and looting
This is a plausible Minsky transmission path, not evidence that a banking crisis has begun. The immediate insured bill may be manageable; the longer-term economic damage could be larger if commercial districts repeatedly lose customers, investment and insurability.
I would distinguish that measurable institutional deterioration from an irreversible societal-collapse conclusion. One violent weekend cannot establish the latter.
The most interesting target for our screen may be French public finances rather than any individual insurer. France has created a contingent state liability for riot damage at precisely the time its borrowing costs and fiscal constraints are already attracting attention.
The numbers worth watching next are the first France Assureurs loss estimate, the operational status of the riot fund, and municipal reconstruction costs—particularly Limoges. Those will tell us whether this is a several-hundred-million-euro insurance event or something substantially larger.
I can include French and Italian civil-unrest insurance losses and the French state backstop in the continuing catastrophe watch, alongside the natural-catastrophe and credit-stress exposures.
Agreed. France and Italy have limited fiscal room for another recurring expense, particularly one that produces no corresponding improvement in economic productivity.
The more important issue isn’t necessarily the immediate cost of repairing burned buildings. It’s the possibility that repeated civil disorder becomes another permanent claim on government budgets—alongside debt service, pensions, healthcare, defense and disaster reconstruction.
And the distinction matters for our Minsky thesis: a €500 million riot bill is one thing; a persistent deterioration in public finances, commercial-property values and tax receipts is quite another.
The fiscal figures underline the vulnerability:
| France | Italy | |
|---|---|---|
| Government debt, Q1 2026 | 117.6% of GDP | 138.9% of GDP |
| 2026 budget deficit forecast | 5.1% of GDP | 2.9% of GDP |
| Main vulnerability | Large continuing deficits and rising borrowing costs | Very high debt stock and weak growth |
Italy has actually made some progress on its deficit, although its government now anticipates a renewed increase to 3.4% of GDP in 2027. That is important counterevidence against treating the two countries as equally distressed.
One particularly interesting wrinkle: France’s government can ultimately become liable for certain riot-related damages even when insurers initially pay. In a September 2026 parliamentary response, the Interior Ministry explained that insurers can seek reimbursement from the state under applicable liability rules, sometimes years after the original event.
That creates exactly the kind of delayed, difficult-to-measure public liability we have been discussing.
And consider the debt arithmetic: a sustained 1-percentage-point increase in borrowing costs, eventually applied to €3.5 trillion of debt, would imply roughly €35 billion in additional annual interest expense once fully reflected in the debt stock. That’s an illustrative long-run sensitivity, not an immediate budget cost.
For our screen, I’d put particular emphasis on French OAT spreads against German Bunds, Italian BTP spreads, sovereign CDS, and the cost of refinancing municipal debt.
The riots themselves are unlikely to break either treasury. But if recurring disorder coincides with higher energy prices, slowing tax receipts and rising refinancing costs, the cumulative effect could be much more serious. That is the compound-shock mechanism worth testing rather than assuming a sovereign crisis has already begun.
