Every business model in insurance draws its margin from a scarce resource: you own it and you have a moat, you lose it and the margin goes with it. Every underwriter knows what happens to margins when capacity floods the market. The same is now happening to information and technology.
I have seen this shift from several vantage points: from inside large P&C carriers where I worked on strategy and technology governance, including one of the largest mergers in the industry; from the MGA side (I co-founded a parametric MGA and built the US operations of a cyber MGA); and from the advisory side (I advise carriers and MGAs on strategy and operations).
Information
Information used to be the core scarcity in insurance. Brokers knew which markets would write which risks. Underwriters knew loss rates. Claims adjusters knew what a loss was worth. This knowledge was hard to acquire and harder to keep current. But this is changing rapidly. Now, risk data exists inside many systems companies use every day. For example, a cloud provider knows more about a company’s security configuration than most cyber application forms will ever capture. A payroll platform knows all about a company’s workforce in real time.
Technology
Technology used to be another scarcity. A carrier’s core system cost tens of millions of dollars and years to build. I was inside a carrier going through a modernization program and can say with some scar tissue: that investment was a real moat. Larger carriers could afford it. Smaller carriers couldn’t. They either lived with legacy systems, ran simpler platforms, or rented access through MGAs and service providers. Cloud platforms and modern APIs collapsed the infrastructure cost. Now AI is collapsing the development cost. An MGA can now stand up in a few months what used to take several years to build – without sitting through a single steering committee.
Capital
Capital died as a structural scarcity. The industry still experiences capital scarcity cyclically. Anyone who renewed a Florida cat program in January 2023 will agree: capacity vanished and reinsurance rates spiked. True, but “old” hard markets used to last several years and got resolved by forming new Bermuda companies: remember the class of 1992, 2001, and 2005? The post-Ian hard market in 2022 resolved in a few months. Capital repriced and quickly returned while the cat bond market set issuance records. When existing capital comes back that fast, the problem is not supply; it’s confidence.
These three scarcities organized the insurance industry for decades, and they started to erode. They haven’t completely vanished – try placing a complex casualty tower without a broker who knows the markets – but for standard risks, it’s well advanced. And the margins that were built based on those scarcities are going with them.
What stays scarce
Other scarcities, like origination, judgment, and regulatory permission weren’t clearly visible up to now, because the infrastructure layers (information, technology, and capital) were sitting on top of them, earning their own margins. A fourth scarcity – confident capital – determines whether those economics survive stress.
Origination
Origination means owning the moment a business first needs risk protection. For generations that moment belonged to the broker, locked in by two forces: the buyer couldn’t see the market clearly, and risk data had to be painfully collected by hand. Both anchors have come loose. The buyer’s ability to see the market without a broker translating it is growing fast, and more risk data exist digitally. For complex risks, what remains of the old distribution advantage is deep expertise and relationships cultivated over years. For standard risks, it’s mostly habit and inertia.
Origination can be owned by whoever sits where the risk data lives. Platform companies, such as payroll providers and cloud vendors, are quietly adding an insurance tab. I think the most interesting distribution deals in the future will be data partnerships, not broker acquisitions, at least for standard risks where origination and judgment can be separated. Broker acquisitions are still commanding double-digit EBITDA multiples because origination has real value today. But I would ask whether that origination is proprietary and locked in by expertise, or only rests on an information advantage with a half-life.
Judgment
Judgment is the second remaining scarcity. An AI model can read and compare policy forms in a minute and find the exclusion that guts the coverage. I have built exactly this. What the AI model can’t do is stand behind the answer. Our regulatory structure is built around a human who is ultimately accountable for the decision and who can be examined, held to a fiduciary standard, and sanctioned if needed. An underwriter decides what risk to write and at what price. A claims adjuster decides what is owed when the loss hits. When the policy wording is contested in court, someone with a license and professional liability on the line has to own the call. All three are concentrating – fewer people, higher stakes, hard cases only.
Regulatory permission
Regulatory permission is the third scarcity: licenses, admitted paper, filings across fifty states. This scarcity erodes slowly, on political, not technological, time. The market is already routing around it: the E&S market has grown from 7% of commercial premium in 2000 to over a quarter today. Fronting carriers rent admitted paper at scale to programs with origination and capital behind them. The industry has built an entire infrastructure to route around regulatory permission. This might be the most defensible scarcity precisely because it is the slowest to change.
Confident capital
The first three scarcities are structural, they are hard to build, slow to earn. Capital is different. It can be abundant, and for extended periods it is, but only while the market believes a risk can be adequately priced, diversified, and reinsured. When that confidence cracks, capacity turns scarce fast. Look at what happened in cyber after the ransomware spike of 2020-21. Capacity pulled back overnight, rates jumped 80% in a single quarter. The underlying risks were the same. What changed was confidence.
Capital is not scarce. Confident capital is.
The first three scarcities determine who captures value. Capital determines whether that value holds through a cycle.
Which scarcity do you own?
Every broker, MGA, and carrier should ask themselves which scarcity they actually own. In a benign market, most players look like they own several. However, under stress, the limiting factor reveals which ownership was real and which was borrowed.
The broker who serves complex risks with genuine placement advice owns judgment, and will be fine. The broker whose book is standard risks that is renewed annually on relationship and inertia owns nothing but habit.
An MGA with proprietary origination or genuine underwriting insight is the natural winner of this transition, the program that capital lines up behind. I say this as someone who co-founded an MGA and spent twelve months discovering that we had conviction but not enough proprietary origination. We could build the product, but we couldn’t find the buyer without going through someone who already owned that relationship. An MGA that is a distribution arrangement with delegated authority and rented paper owns none of the four scarcities.
The mid-tier generalist carrier is the most exposed to this transition: too small to be a capital utility, too undifferentiated to own origination, spending its transformation budget to arrive where the moat used to be. Small specialty carriers are a different story.
How scarcities combine
No scarcity works in isolation. Judgment without regulatory permission is an MGA, relying on its fronting carrier’s appetite and economics. Permission without judgment is a fronting carrier backing someone else’s underwriting decisions. Clear Blue has turned one scarcity into a business model by renting paper to MGAs that own another. Judgment combined with permission is a specialty carrier that underwrites its own paper – and that combination is why companies like Kinsale and RLI trade at premium multiples despite their size.
Chubb owns three of the four: judgment, permission, and capital. It doesn’t own origination directly – the customer moment sits with the broker. But the other three are strong enough that brokers bring the business to Chubb.
You don’t need all four if the ones you own make you the counterparty everyone else wants on the other side of the contract. Very few manage all four. Berkshire does, through GEICO’s direct origination, disciplined underwriting, the largest capital base in insurance, and licensing across every jurisdiction.
Own one, rent the rest
For everyone else, the honest question is which scarcity you would win if it had to stand alone.
Own that one completely. Rent the others.
For a carrier that has bundled all four for decades without excelling at any one, this means deciding which to own and which to let go.
For the investors rolling up this industry, the diligence question has changed. Brokerages priced on inertia carry more risk than the high multiples suggest. The best value creation play is taking an asset that genuinely owns one scarcity and stripping the cost out of everything else, because everything else is becoming commodity work with commodity margins.
I was inside this industry when the moats were still real. These moats have moved. Institutions that pick the scarcity they can genuinely own and have the discipline to rent everything else will thrive.
