A container ship leaves Shanghai for Rotterdam. An aircraft crosses the Atlantic. A bank finances an offshore wind farm. A developer builds hundreds of homes along a coastline exposed to storms. An oil company operates an offshore platform. An operator places a satellite into orbit. Behind each of these operations lies a rarely visible reality: someone has to agree to carry the risk.

That role belongs largely to insurers and, behind them, reinsurers.

They do not issue building permits, grant operating licences or define trade routes. They do not formally decide whether a ship can sail, a power plant can operate or a building can be constructed. Yet their willingness to accept, price, limit or refuse a risk directly influences the economic feasibility of these activities.

An infrastructure project may be technically viable yet become difficult to finance if it cannot be insured. A ship without adequate coverage may lose access to certain commercial operations. A property exposed to a risk that has become prohibitively expensive to insure may suffer a deterioration in both its value and its financing capacity. An energy project may receive every administrative authorisation it requires and still become financially unviable if no one is willing to carry its construction, liability or operational risks.

Insurance is therefore one of the invisible infrastructures of the global economy.

And as natural catastrophes, concentrations of assets, geopolitical tensions and technological risks increase, one question becomes increasingly important: what happens when part of the world becomes too risky to insure?

I. Before modern finance, the voyage already had to be insured

The history of modern insurance is inseparable from the history of maritime trade.

In the late seventeenth century, Edward Lloyd’s coffee house in London became a meeting place for shipowners, captains, merchants and people with capital to deploy. Information circulated there about ships, cargoes, ports, trade routes and dangers at sea. Underwriters agreed to assume part of the financial risk of a voyage in exchange for a premium.

The earliest known reference to Lloyd’s Coffee House dates from 1688. The development of this activity accompanied the expansion of British maritime commerce and gradually helped structure a genuine insurance market. Lloyd’s itself recalls that the coffee house became a centre of maritime intelligence where shipowners could meet people willing to insure their vessels and cargoes. (lloyds.com)

The economic principle was already the one that governs much of contemporary insurance: transform uncertainty that could be catastrophic for an individual participant into a quantified risk distributed among multiple providers of capital.

Without this mutualisation, every merchant would need to be capable of absorbing the total loss of a ship and its cargo. With it, an exceptional loss becomes a probabilised cost incorporated into commercial activity.

Insurance did not simply accompany the globalisation of trade. It helped make it financially possible.

This history also has a much darker side. Lloyd’s today acknowledges the role its market played in insuring commerce connected to the transatlantic slave trade. The same mechanism that facilitated European maritime commerce could also provide financial protection to activities that were profoundly violent and inhumane. (lloyds.com)

This historical contradiction already reveals a fundamental characteristic of insurance: by nature, it is not a moral arbiter of risk. It is primarily a mechanism for selecting, pricing and transferring risk, subsequently constrained by law, regulation and the choices of market participants.

Three centuries later, that mechanism has expanded into almost every complex sector of the economy.

II. A global economy built on layers of protection

When an individual insures a car, the relationship appears straightforward: a policyholder, an insurer, a premium and a potential claim.

Large-scale risks work differently.

A refinery, an airport, an airline fleet, an energy network or a vast infrastructure project may represent exposures so large that a single insurer is unwilling — or unable — to carry them in full.

The risk is therefore fragmented.

Several insurers may participate in the same coverage. Those insurers themselves transfer part of their exposure to reinsurers. Reinsurers may in turn cede some of their own risk through retrocession. Certain catastrophic exposures can ultimately be transferred to capital markets through mechanisms such as catastrophe bonds.

The system therefore forms a multi-layered architecture:

economic asset → insurers → reinsurers → retrocession → capital markets.

This depth is essential. It allows enough capital to be mobilised to absorb events whose costs could vastly exceed the capacity of any individual participant.

This is where major global reinsurers such as Munich Re, Swiss Re and Hannover Re operate, alongside the Lloyd’s market.

Lloyd’s itself is not a conventional insurance company in the usual sense. It is a marketplace in which brokers present risks to underwriters operating through syndicates. In 2025, the Lloyd’s market recorded £57.9 billion in gross written premiums and held £49.8 billion in capital, reserves and subordinated debt. (lloyds.com)

This architecture produces a crucial consequence: global insurance capacity is not unlimited.

For something to be insured, it is not enough that the risk exists. Someone must be willing to commit capital against that risk, must be capable of assessing it, and the premium demanded must remain economically sustainable for the insured.

It is precisely at this point that insurance begins to exert considerable influence over the real economy.

III. The price of risk becomes the price of activity

An insurer does not merely ask whether a dangerous event might occur. It seeks to determine its probability, potential severity, the value exposed and its correlation with other risks already present in the insurer’s portfolio.

An isolated factory may be insurable.

One hundred factories concentrated in the same cyclone-exposed region present a different problem.

A single building may represent a reasonable risk.

Thousands of buildings insured simultaneously in a wildfire-prone region create an accumulation of exposure capable of generating enormous losses from a single event.

Contemporary insurance therefore relies heavily on modelling: weather data, building characteristics, cyclone trajectories, claims histories, seismic models, supply chains, asset geolocation, infrastructure vulnerability, liability scenarios and geopolitical exposure.

The result is ultimately expressed in financial terms.

Risk becomes a premium.

But as risk increases, several responses become possible: higher premiums, larger deductibles, lower coverage limits, additional exclusions, requirements for stronger prevention measures or a reduction in the capacity offered.

At the extreme, the insurer withdraws.

This is where the distinction between high risk and uninsurable risk becomes fundamental.

As long as a risk remains insurable at a price compatible with the profitability of a project, economic activity can continue. When coverage becomes prohibitively expensive or insufficient, the economics of the project may collapse before the physical risk ever materialises.

IV. Maritime: no globalisation without risk transfer

More than three centuries after Edward Lloyd’s coffee house, maritime transport remains one of the clearest examples of this dependence.

A ship concentrates several categories of risk: hull value, cargo, third-party liability, pollution, collision, damage to port infrastructure, injury to crew or passengers, business interruption and, in some cases, war risk.

Environmental liabilities add another layer.

The 2001 International Convention on Civil Liability for Bunker Oil Pollution Damage requires owners of ships over 1,000 gross tonnage to maintain insurance or another form of financial security covering their liability for certain bunker-oil pollution damage. This coverage is evidenced by a certificate issued by a State Party. (wwwcdn.imo.org)

Comparable mechanisms in other areas of maritime law also require certificates of insurance or financial security. (wwwcdn.imo.org)

Insurance here ceases to be merely voluntary protection against loss. It becomes part of the legal architecture that enables the vessel to operate.

Geopolitics can therefore be transmitted directly into the insurance market.

When a shipping route crosses a war zone or experiences a sharp increase in the risk of attack, war-risk premiums may rise, coverage terms may change and some insurers may reduce their exposure. The economic cost of insecurity thus appears in the price of transportation before a ship is actually hit.

The insurance market becomes a kind of seismograph of global geopolitics.

V. Aviation: insuring the machine, but above all its consequences

Aviation reproduces the same logic on a different scale.

The value of a modern aircraft can be considerable, but the risk does not end with the aircraft itself. It includes passengers, crews, baggage, cargo, ground infrastructure and liability towards third parties.

The 1999 Montreal Convention constitutes one of the major international frameworks governing air-carrier liability for damage involving passengers, baggage and cargo. Its liability limits are periodically reviewed by the International Civil Aviation Organization; the latest revision entered into force on 28 December 2024. (iata.org)

For an airline, insurance is therefore not merely an optional cost item.

It belongs to the financial and legal infrastructure of operations.

Aviation risk also has another distinctive characteristic: certain disasters can produce extremely large but statistically rare losses. Insurers must therefore have sufficient capacity to accept substantial exposures without immobilising a disproportionate amount of capital.

Reinsurance again plays a decisive role.

What enables an aircraft to take off is therefore not merely the combination of a manufacturer, pilot, airport, regulator and airline. Behind this visible chain exists another, financial chain whose function is to determine who will bear the loss if the improbable event occurs.

VI. Building also requires the ability to insure

The same logic extends to infrastructure.

Power plants, dams, motorways, tunnels, ports, industrial facilities, wind farms, digital networks and major property developments often require several successive categories of coverage.

During construction there are risks of physical damage, accidents, liability and delay. Once the asset becomes operational, other exposures emerge: breakdown, fire, business interruption, natural catastrophe and third-party liability.

More importantly, these assets are frequently financed with debt.

And lenders want to protect the value securing their financing.

This relationship creates an extremely powerful transmission mechanism:

physical risk → insurance → financing → investment.

An insurer does not need to formally prohibit a project to influence whether it happens. Sometimes it only needs to change its economics.

If coverage becomes more expensive, operating costs increase. If deductibles rise, more risk remains with the asset owner. If certain guarantees disappear, the lender may require more equity, increase the cost of credit or reconsider its commitment.

Insurance therefore operates as a filter between technical possibility and financial viability.

VII. Climate is gradually moving the frontier of insurability

This function becomes particularly visible with natural catastrophes.

According to Munich Re, natural disasters caused approximately $224 billion in global economic losses in 2025, of which around $108 billion was insured. Wildfires, floods and severe storms accounted for the majority of insured losses that year. (munichre.com)

The previous year had already demonstrated the structural scale of the phenomenon. Swiss Re estimated insured natural-catastrophe losses at $137 billion in 2024, against $318 billion in economic losses, leaving $181 billion uninsured. The institute also estimated that insured natural-catastrophe losses had been following a long-term real annual growth trend of between 5% and 7%. (swissre.com)

Global averages, however, conceal enormous disparities.

In 2025, Munich Re estimated that Asia-Pacific suffered approximately $73 billion in natural-catastrophe losses, of which only $9 billion was insured. In Africa, losses reached around $3 billion, of which less than one fifth was covered. (munichre.com)

More broadly, the OECD observes that in 24 of its 38 member countries, less than half of natural-catastrophe losses since 2000 were insured. For floods alone, 32% of economic losses recorded across OECD countries between 2000 and 2024 were covered. (oecd.org)

This is the protection gap: the difference between economic losses and losses covered by insurance.

And that gap is far more than a technical indicator for the insurance industry.

When a house is destroyed but insured, part of the capital required for reconstruction has already been contractually organised before the disaster occurs. When it is uninsured, the loss must be borne by the owner, the creditor, the state or some combination of them.

At scale, insurability therefore becomes a question of economic resilience.

VIII. When insurance begins to redraw territory

The deepest consequence, however, may appear before the catastrophe.

When models indicate that an area is becoming structurally more exposed to wildfire, flooding or storms, that signal gradually enters insurance contracts.

Premiums rise.

Deductibles increase.

Terms tighten.

Certain types of coverage become more limited.

And, in some cases, insurers reduce their presence.

The OECD specifically notes that households and businesses in high-risk regions may face restrictions on coverage or premiums that become increasingly difficult to afford. It also warns that increases in the frequency or severity of certain natural hazards may lead insurers and reinsurers to increase pricing, making coverage less accessible. (oecd.org)

The mechanism can then propagate through an entire territory.

A home that is difficult to insure becomes more difficult to finance.

An asset that is difficult to finance becomes less liquid.

Lower liquidity can weigh on its value.

Falling values can weaken the collateral held by banks.

And if the phenomenon affects an entire region, what initially appeared to be an insurance problem becomes a property, banking, fiscal and ultimately political problem.

Yet the insurer has decided no zoning policy.

It has simply recalculated the risk.

This may be one of the most important manifestations of insurance power today: actuarial models can gradually influence economic geography without ever becoming official instruments of territorial planning.

IX. Reinsurance: the power behind the power

To understand how far this logic can extend, it is necessary to look behind the insurer.

When an insurance company covers thousands of buildings exposed to the same cyclone, it generally does not retain the entire exposure. It transfers part of it to a reinsurer.

The reinsurer is therefore the insurer of insurers.

This function gives it a distinctive view of global risk. Where a national insurer sees its own portfolio, a major reinsurer can aggregate exposures across numerous countries, sectors and categories of catastrophe.

It must therefore monitor not only individual risks but also correlations between them.

A hurricane can simultaneously destroy homes, interrupt businesses, damage infrastructure and generate thousands of motor claims. An earthquake can affect buildings, factories, electricity networks, ports and supply chains at the same time.

Reinsurance helps absorb these accumulations.

But its price is then transmitted downwards.

If reinsurance capacity becomes more expensive or scarcer, primary insurers may in turn raise prices, reduce the limits they offer or retain more risk themselves.

The frontier of insurability can therefore depend on actors whom the final policyholder may never even know exist.

X. When Wall Street insures the hurricane

The system no longer necessarily ends with reinsurers.

Part of catastrophic risk can be transferred to financial markets through insurance-linked securities, the best known of which are catastrophe bonds.

Under various structures, the principle is to make investors bear part of a predefined risk. In exchange for a return, their capital may be exposed to loss if the contractually specified catastrophic event occurs.

The economic significance is considerable.

Global capital markets have a depth far greater than the balance sheet of any individual insurer. Transferring part of the risk to investors therefore creates additional capacity that can complement conventional insurance and reinsurance.

The OECD identifies the development of catastrophe-bond markets as one of the instruments capable of contributing to financial protection against disasters, particularly in regions facing significant protection gaps. (oecd.org)

A Caribbean hurricane, an earthquake or a major storm can therefore indirectly become an exposure held in an investment portfolio thousands of kilometres away.

Physical risk is transformed into a financial asset.

The circle is complete.

XI. Insurance does not simply say “yes” or “no”

Portraying insurers as a private administration issuing permissions would nevertheless be misleading.

Their power is generally more subtle.

They rarely say: “this activity is prohibited.”

Instead, they say:

this risk costs this much.

Or: this coverage requires this deductible.

Or: this facility must have this protection.

Or: we will accept only this proportion of the risk.

Or, ultimately: this exposure no longer falls within our underwriting appetite.

That distinction is fundamental.

Insurance does not exercise legal sovereignty. It exercises capital discipline.

And that discipline can also produce positive effects.

An insurer may require fire-suppression systems, stronger safety procedures, building reinforcement, improved flood protection or additional preventive measures before accepting an exposure.

Munich Re has emphasised that adaptation and preventive measures — including more resilient buildings and infrastructure — can help limit losses and maintain reasonable premiums in exposed regions. (munichre.com)

The OECD likewise considers adaptation essential to limiting future increases in climate-related damage and preserving the availability of affordable insurance. (oecd.org)

The insurer therefore does not merely observe risk.

By putting a price on it, it creates an economic incentive to reduce it.

XII. But who insures the uninsurable?

This mechanism nevertheless reaches its limits when risks become too large, too correlated or too difficult to model.

A global pandemic, a natural catastrophe striking a megacity, a systemic cyberattack or certain consequences of a major conflict can generate losses exceeding the capacity that private markets are reasonably willing to mobilise.

The state then reappears.

Public insurance programmes, sovereign guarantees, public-private mechanisms, catastrophe funds and arrangements of last resort can preserve forms of protection where the market alone can no longer provide them at a socially acceptable price.

The OECD notes, in particular, that a growing number of countries have introduced or expanded public-private arrangements to broaden access to affordable flood coverage. (oecd.org)

This intervention reveals an essential political boundary.

If the actuarial price of a risk becomes incompatible with the socially acceptable price of covering it, two principles collide.

The first says: risk should be paid for by whoever bears it.

The second responds: some activities, territories or assets must continue to receive collective protection.

At that point, the question is no longer merely how to insure the risk.

It becomes: who should pay to preserve insurability?

XIII. The world therefore has a second map

We generally view the global economy through its political borders, trade routes, financial centres, natural resources and infrastructure.

Insurers see another map.

A map of concentrations of value.

A map of cyclone paths.

A map of seismic faults.

A map of ports, platforms, factories and networks.

A map of wildfires.

A map of legal liabilities.

A map of conflicts.

A map of technological risks.

A map in which every activity carries a probability of loss, every asset an exposure and every territory an implicit price of risk.

This map is not fixed.

It evolves with the climate, urbanisation, technology, regulation, conflict and the growing concentration of wealth in particular regions.

And changes to it eventually appear in the real economy.

Conclusion — The invisible permit

Insurers do not govern the world.

They possess neither the political legitimacy nor the legal powers to do so.

But a considerable part of the modern world cannot function unless someone agrees to carry its risks.

Building infrastructure, financing property, transporting cargo, flying an aircraft or operating an industrial facility requires not only authorisations, capital and technical expertise, but also an answer to a much quieter question:

who will pay if everything goes wrong?

From Edward Lloyd’s coffee house to contemporary climate models, insurance has organised precisely that answer.

As long as risk can be measured, distributed and sold at an acceptable price, economic activity can proceed.

When that price rises, it slows.

When conditions tighten, it adapts.

When capital ultimately refuses to carry the risk, a boundary appears.

It is marked on no official map, and no government necessarily drew it.

Yet behind ports, aircraft, power plants, buildings and global trade routes lies an invisible permit: insurability.

And in a world where some physical risks are increasing faster than our capacity to absorb them, the decisive question may no longer simply be what we are capable of building.

It may be what someone will still agree to insure.

Main Sources

  • Lloyd’s — Our History; Coffee and Commerce 1652–1811; 2025 annual results and institutional documentation on the Lloyd’s market. (lloyds.com)
  • Munich Re — Natural disaster figures 2025, NatCatSERVICE, January 2026: global economic and insured losses, regional distribution and developments in catastrophe risk. (munichre.com)
  • Swiss Re Institute — sigma 1/2025: Natural catastrophes: insured losses on trend to USD 145 billion in 2025: 2024 natural-catastrophe data and the global protection gap. (swissre.com)
  • OECD — Financial Protection Against Catastrophic Risks: Floods, Fires and Other Major Risks, March 2026: protection gaps, insurance affordability and public-private arrangements. (oecd.org)
  • OECD — Protection Gaps in Insurance for Natural Hazards and Retirement Savings in Asia, July 2025: protection-gap mechanisms and the economic and financial consequences of uninsured losses. (oecd.org)
  • OECD — Enhancing the Insurance Sector’s Contribution to Climate Adaptation: adaptation, prevention and the preservation of insurability as climate risks evolve. (oecd.org)
  • International Maritime Organization — documentation concerning compulsory insurance and financial-security requirements, including those under the 2001 Bunkers Convention. (wwwcdn.imo.org)
  • IATA / Montreal Convention 1999 — international framework for air-carrier liability and the revision of liability limits effective December 2024. (iata.org)