What happens when risk meets capital directly?
By James W. Moore
Key Takeaways
- An insurance risk exchange could make placement faster and expose more risks to competing capital, while shifting agents’ value toward representing insureds.
- A bid is not financial security. Capital must be committed, held, and available for claims that may arrive years after the trade.
- Faster placement does not settle contested claims or guarantee coverage availability, especially when private capital withdraws.
- An exchange that succeeds could concentrate control over market access, rejected submissions, and the definitions of insurable risk.
- The exchange may emerge gradually through infrastructure built by the insurance industry itself, leaving familiar company names in place while the rules governing the market move elsewhere.
An Innocent Question Gets Out of Hand
I’ve been thinking about what happens when artificial intelligence allows an insurance company to make underwriting decisions almost instantaneously. Now imagine reinsurers evaluating their portfolios, changing their appetites, and making capacity decisions at the same speed.
A submission that once took days to evaluate might be analyzed in seconds. A reinsurance portfolio might be reassessed as soon as a meaningful piece of information changes. Underwriting decisions could move at the speed of electrons.
But the capital supporting those decisions still has to find its way through the traditional machinery of insurance: brokers, carriers, reinsurers, treaty arrangements, collateral, documentation, and the institutional relationships that connect one party’s risk to another party’s money.
That got me thinking about securities exchanges. They didn’t abolish investors, issuers, analysts, or the need to trust that transactions would settle. They changed the infrastructure through which buyers and sellers found one another.
And then I had an idea that got considerably out of hand.
What if every insurance risk could trade on one exchange?
Not just reinsurance or a marketplace comparing carrier quotes, but one global mechanism through which risks reach willing capital, with underwriting, claims, legal services, and other expertise brought in when they’re needed.
The disturbing thing about this idea isn’t how radical it sounds. It’s how many of the pieces already exist. The impossible-sounding part may be less about inventing technology than persuading people to organize insurance around it.
Meet Our First Customer
Imagine a taxi driver in Mumbai who needs motor insurance. He gives his agent the information required to describe the vehicle, driver, operating conditions, and coverage need. The agent checks it, adds relevant context, and submits a standardized risk package to the exchange.
Qualified capital providers, or the systems acting for them, evaluate the exposure. Some want a tiny slice of motor risk in India. Others want diversified participation across thousands of similar policies. Bids arrive, conditions are assembled, the legal coverage is issued, and the agent can show the driver what protection he’s purchasing.
There is no need for the driver to know whether the ultimate capital comes from an Indian insurer, a European reinsurance portfolio, or a fund in North America. Most insureds have never known the final destination of the risk they purchase protection against. Their concern is much simpler: Will the policy respond, and will someone pay the claim?
Even our invented transaction must obey real local rules. In India, motor third-party premiums are prescribed rather than freely negotiated by each insurer, so the exchange could not simply auction that part of the price. [1] It would need to respect the local framework while finding more efficient ways to assemble the capacity and services behind the policy.
Grant it that ability, and the exchange looks remarkably attractive. Competition for well-described risks becomes more direct. Available capacity can be assembled from several sources instead of depending on any single carrier’s appetite. An agent spends less time navigating carrier portals and more time understanding the insured, getting the submission right, explaining coverage, and advocating when something goes wrong.
That might be a particularly attractive future for independent agents. Knowing which carrier wants which account has value today because placement is difficult. If capacity becomes easier to find, the agent’s durable value shifts toward being the insured’s representative, not merely the navigator of insurer preferences.
Specialist underwriters don’t disappear. They can sell judgment where ordinary classification falls short. Engineers, adjusters, actuaries, attorneys, and claims administrators can serve the transaction without necessarily belonging to the same institution that supplied the capital. Existing carriers might participate as capital providers, underwriting specialists, administrators, or all three.
The exchange doesn’t need to remove expertise. It needs to make expertise easier to call upon.
We Have Been Building Pieces of It for Years
None of this requires AI to invent the idea of a risk marketplace.
Primary insurance is largely organized through local companies and legal jurisdictions, but capital has been crossing borders through reinsurance for generations. Treaty reinsurance, facultative placements, retrocession, and insurance-linked securities already connect portfolios of local insurance promises to capital around the world. In that sense, insurance has a global interoperability layer. It just doesn’t usually reach down to match capital with every individual policy.
Lloyd’s offers a much longer-running experiment. It developed as a marketplace where separate participants supplied underwriting judgment and capital. Its modern structure still separates functions among brokers, syndicates, managing agents, capital providers, and a central market institution. In 2019, Lloyd’s proposed a digital risk exchange for relatively straightforward, high-volume risks, alongside a separate platform for complex placements and changes to claims, data, and access to capital. [2]
Accelerant calls its platform the Accelerant Risk Exchange, connecting specialist underwriting organizations with insurers, reinsurers, and institutional investors. It still depends on licensed insurers, but demonstrates that organizing and monitoring risk can be a business distinct from bearing it. [3]
No existing platform has built our imaginary universal market. What exists is evidence that some of its functions can operate separately.
A Bid Is Not a Balance Sheet
Before any claim arrives, there is a more basic problem. An investor’s willingness to bid is not the same as capital legally available to pay a claim. A pension fund could offer to participate in thousands of policies, but what assets are actually committed? Who holds them? In what currency? What happens if the investor wants its money back, the value of those assets falls, or several years of claims develop worse than expected?
Insurance companies do more than match risk with investors. They turn premiums, investments, reserves, surplus, reinsurance, and enforceable financial obligations into claims-paying security. Our exchange would need to do that too, whether through collateral, funded vehicles, licensed issuing companies, or some combination of them. The law would still have to recognize those assets and obligations across jurisdictions.
Lloyd’s offers a particularly concrete illustration. Its chain of security includes syndicate assets held in trust, capital supplied by members through Funds at Lloyd’s, and central assets available under defined conditions. It is not simply an electronic invitation for investors to bid. [6] Replacing that architecture with a common marketplace might be possible, but deleting it would not be.
Lloyd’s history also shows what can happen after the trade.
During the late twentieth century, asbestos, pollution, and other long-tail liabilities produced enormous claims problems for the Lloyd’s market, including liabilities originating decades earlier. Its response included restructuring old liabilities through Equitas, opening the market to corporate capital, and strengthening financial protections. [4] The disaster emerged not when risks were placed, but long after, as the claims became clear.
A marketplace that connects risk and capital is not automatically a marketplace that can live with the consequences after the transaction has been made.
Our Mumbai driver is about to discover why.
The Trade Clears. Then Someone Has an Accident.
Months after the policy binds, our driver hits a pedestrian. The driver says the pedestrian stepped into traffic. The pedestrian says the driver was distracted. Medical treatment begins. Liability is contested. The claim might settle quickly, or it could involve litigation and years of payments.
The exchange has a perfect record of the original bid. That’s helpful, but it does not answer who was at fault, what the policy covers, what damages are reasonable, whether a claim is fraudulent, or what should happen when the parties disagree.
The exchange can clear the trade. Someone still has to decide the claim.
On the exchange, placement might move at the speed of electrons. Claims can still move at the speed of courts. This is why the seemingly simple phrase “insurance risk” can hide a lot of work. The loss being insured often cannot be observed with the same clarity as the transaction that transferred it. Even the relevant facts may not be known until long afterward.
Some exposures are easier to settle through objectively measurable triggers. A parametric contract might pay when a specified storm measurement is reached. The exchange could handle that efficiently. But now the insured may suffer a loss without the trigger being met, or receive payment when the actual loss is smaller. Simplifying settlement moves uncertainty; it doesn’t make all of it vanish.
We can hire claims administrators and adjudicators, of course. We can establish policy language, appeals, independent review, reserves, and rules for advancing claims payments. But notice what we’re doing. We are rebuilding institutions around the promise, long after celebrating how easily we separated it from the capital.
There’s also a question of incentives. A conventional insurer has a reputation, renewal relationships, and a long-term business to protect. Anonymous capital that bought a fractional participation in one risk may not care whether a marginal claim is handled generously or litigated to exhaustion. The exchange can establish duties for claims management, but who writes those duties, who enforces them, and who resolves the inevitable conflict between the claimant and the investors?
Our taxi driver doesn’t care that the risk was attractively priced or efficiently allocated. He wants the pedestrian’s claim handled fairly and the promise behind his policy honored.
We have not disproved the exchange. We have made its back end considerably more interesting.
What Happens When Nobody Bids?
Now imagine the next monsoon season brings flooding and loss deterioration in the driver’s region. Capital providers revise their assumptions. Bids for similar risks become more expensive, more restrictive, or disappear altogether.
An exchange is excellent at showing us where capital wants to go. It may be much less comforting when it shows us where capital refuses to go.
Today’s insurance market already has this problem. We use assigned-risk mechanisms, residual-market plans, state-created facilities, and other arrangements because availability of coverage is not always something private capital delivers on terms society finds acceptable. Texas property insurers and regulators certainly don’t need a thought experiment to explain the broader problem. In India, insurers have motor third-party business obligations, including specified growth obligations for some vehicle categories, reflecting a policy choice not to leave availability solely to voluntary appetite. [5] But if the individual insurer becomes secondary to the transaction, whom does the regulator obligate: the issuing company, the participating capital providers, or the exchange operator? Each answer changes the structure we thought we had simplified.
The exchange might bring new investors into a difficult market, improving availability. But it could also make attractive risks easier to extract from mixed portfolios, leaving specialized insurers and residual mechanisms with a more concentrated share of the difficult business.
There is also a timing problem. Capital might compete vigorously when loss experience is benign, then withdraw together when the environment deteriorates. The exchange could be most liquid when protection is easiest to provide and least liquid when protection is most urgently needed.
Those are not uniquely exchange-created problems. But a beautifully efficient system for matching today’s willingness to bear risk is not necessarily a system for sustaining tomorrow’s willingness.
It would be easy to stop here and declare the exchange a failure. But that would miss the other half of the experiment.
Suppose We Solve All That
Suppose the exchange develops credible claims institutions. It works with local regulators. It turns investor commitments into legally enforceable claims-paying security, with appropriate collateral, reserves, liquidity, and accountability. It establishes mechanisms for difficult risks. It develops appeals, liquidity facilities, contingency arrangements, and a reputation that people actually trust.
Suppose it becomes successful. Now the more interesting problems begin.
As more insureds and capital providers join, the value of being on the exchange grows. The operator sees an extraordinary volume of submissions, bids, pricing decisions, risk characteristics, claims outcomes, and behavioral information. It can observe patterns no individual carrier, reinsurer, broker, or regulator sees in the same way.
We began by imagining that the exchange would replace expensive intermediaries. But perhaps we have created the most powerful intermediary insurance has ever known.
Amazon’s third-party marketplace provides an uncomfortable analogy. A platform can enable transactions between independent parties while controlling the rules of participation, the data surrounding transactions, and the ways buyers find sellers. It may also compete with participants who depend on the platform.
An insurance exchange would have an even more consequential form of visibility. It could see not only what bound and at what price, but also the terms offered, the bids rejected, and the risks that never found a buyer. It could determine which capital providers see which submissions, which underwriting information must be supplied, and which participants qualify for access. Nothing about being a marketplace operator necessarily prevents it from also supplying capital, if the law and its governance allow it.
Central clearing in financial markets offers another version of the tradeoff. Shared infrastructure can reduce some bilateral exposures while creating a node whose governance and resilience matter to everyone who uses it. An insurance exchange might reduce fragmentation in transactions and increase dependency on the institution that processes them.
This is no longer simply a question about who bears the loss. It is about who controls the market where loss-bearing capacity is allocated.
Who Decides What Counts as a Risk?
A global exchange needs a common language. It has to identify a risk, classify it, attach data, define the terms on which it can be offered, and make submissions comparable enough for capital to bid.
For mature, well-understood exposures, standardization could be an extraordinary benefit. It might reduce duplicated data entry, ambiguity, and the informational advantage of whoever happens to know the right intermediary.
But standardization is not neutral.
What happens to a business that doesn’t fit the classification system? An unusual manufacturing process? A technology no one has insured before? A liability exposure whose trigger is still being argued in court?
Today, some of those risks eventually find their way into coverage because an underwriter in a specialty market is willing to listen, write unfamiliar language, limit the exposure, and see what happens. The industry’s next insurance product often begins as an awkward exception to the last one.
Our exchange could provide a specialist lane for these risks, and probably would. But the incentives of a high-volume, highly automated marketplace run toward things that can be described, compared, and processed consistently. The risks that need imagination rather than classification may remain at the edges.
This could be the exchange operator’s greatest source of power, and it would be easy to overlook. Not the power to set a premium, but the power to determine what must be described before a risk is allowed to reach the market at all.
The exchange’s taxonomy becomes a kind of underwriting constitution. Risks it describes well can reach capital efficiently. Risks it describes poorly may become expensive, invisible, or dependent on special permission.
We’ve spent decades treating insurance’s many markets, companies, jurisdictions, and specialist practices as a source of expense. Some of that duplication surely is expensive. But some of the fragmentation gives people other places to go, other definitions of the risk to try, and other decision-makers to persuade.
Insurance understands diversification of financial risk. We might also want to think about diversification of control.
What If the Industry Builds It Itself?
There is one last possibility that might be less dramatic, and for that reason more plausible, than a startup announcing it has replaced the insurance industry.
What if the industry builds the exchange a piece at a time?
A group of reinsurers creates shared capital-access infrastructure. Core-platform vendors allow direct submission and capacity matching through their systems. Carriers adopt common data standards because doing so reduces expense. Regulatory systems become easier to connect. Specialty markets establish common electronic placement and claims procedures. Each improvement is sensible on its own.
No one announces the end of the insurance company. The companies are still there. Policies carry familiar names. Agents still see customers. Claims adjusters still investigate losses.
But over time, the definitions, access rules, data, and transaction mechanisms governing a growing portion of insurance migrate to a smaller number of shared systems.
The change might look less like a revolution than a software upgrade that never quite ends.
Shared infrastructure can make markets more accessible, auditable, and efficient. Independent governance and interoperable alternatives can preserve some of the checks that a single dominant operator would weaken. On the other hand, nominally independent insurers may have little practical freedom if they all depend on the same market vocabulary and gateway to capital.
Our Mumbai driver might never notice. His policy still carries a familiar company’s name. His agent still answers the phone. But the rules that determine whether the next policy can be offered, what information counts, and which capital may bid could increasingly live somewhere neither of them can see.
I began by wondering what happens when underwriting and reinsurance decisions move faster than the machinery used to put capital behind them. The surprising thing is how much of the problem turns out not to be matching risk with capital. It is deciding what counts as a risk, what counts as a loss, who must pay, and who gets to make those decisions.
If nobody had invented insurance companies, would we build them?
And would we recognize the moment the market became more important than the companies trading on it?
Sources
- Insurance Regulatory and Development Authority of India (IRDAI), Motor Insurance: Policyholder Information. Explains prescribed third-party premiums and insurer-rated own-damage cover.
- Lloyd’s, Lloyd’s Reveals Plans to Become Most Advanced Insurance Marketplace in the World, September 30, 2019; and Lloyd’s Ready to Deliver First Phase of Transformation, February 26, 2020.
- Accelerant, Form 10-Q for the quarter ended June 30, 2026, describing its Accelerant Risk Exchange and relationships with underwriting members and risk-capital partners; Investor Relations overview.
- Lloyd’s, Sir David Rowland: Lloyd’s Historical Profile; HM Revenue & Customs, Lloyd’s Market Developments: 1992 to 2002. On asbestos/pollution losses, corporate capital, and restructuring.
- IRDAI, Master Circular on Rural, Social Sector and Motor Third Party Obligations, 2025, No. IRDAI/NL/CIR/RSS/85/07/2025 (July 25, 2025). Regulator-issued PDF hosted by an independent legal-document archive; IRDAI document record. Applies to 2025–26 and 2026–27.
- Lloyd’s, Capital Structure: Chain of Security, including syndicate assets, Funds at Lloyd’s, and central assets.
AI Disclaimer: This content was created with assistance from artificial intelligence technology. While content is based on factual information from the source material, readers should verify all details directly with the respective sources before making business decisions.
