The History of Insurance: How Sharing Risk Built Commerce

On the afternoon of September 2, 1666, a fire broke out in a bakery on Pudding Lane in London. By the time it burned itself out four days later, it had destroyed 13,200 houses, 87 churches, and most of the medieval City of London. The rebuilding cost was estimated at ten million pounds — roughly the entire annual tax revenue of the English Crown. Most of the destroyed property was uninsured; insurance, as an institution, barely existed in England at the time. The economic disruption was catastrophic and prolonged. Property owners who had accumulated wealth over generations found it reduced to rubble overnight with no mechanism for recovery beyond their own savings and whatever charity they could attract.

Six years after the fire, Nicholas Barbon opened what is generally recognized as the first property insurance office in London, selling fire insurance on houses. His innovation was not the concept of pooling risk — that had existed in various forms for millennia — but the institutional packaging of it as a standardized product that individual property owners could purchase for a known annual premium. Within a generation, Lloyd’s Coffee House had become the center of the marine insurance market, and the institutional infrastructure of modern insurance was taking shape. The Great Fire of London was a catastrophe. It was also the catalyst that made the English middle class understand, viscerally, what the absence of insurance meant — and drove the demand that created the industry.

Risk Pooling Before Insurance

The intellectual history of insurance begins not in seventeenth-century London but in the earliest recorded commercial societies, where the fundamental problem of large, random losses had already forced economic actors to develop risk-sharing mechanisms.

The oldest documented form of what we would recognize as insurance-like risk transfer is the bottomry loan of ancient Mediterranean commerce. A merchant or ship owner needing to finance a trading voyage would borrow money under an agreement that the loan would be repaid at a high interest rate if the ship returned safely, but would be cancelled if the ship was lost at sea. The lender was, in effect, selling insurance: accepting the risk of the voyage’s failure in exchange for an above-market return if it succeeded. The high interest rate embedded the insurance premium. Babylonian bottomry loans appear in the Code of Hammurabi, circa 1750 BCE. Greek and Roman merchants used them extensively throughout the Mediterranean trade.

This was a clever mechanism but a limited one. The same person providing the capital was also absorbing the risk, which meant the arrangement required lenders with both the capital and the risk tolerance to accept maritime exposure. It also meant that risk was not truly pooled — each lender bore the specific risk of the specific voyage they had financed, rather than holding a diversified portfolio of many voyages. The mathematical insight that would eventually make insurance efficient — that pooling many independent risks dramatically reduces the volatility of aggregate outcomes — was present in embryonic form but not yet formalized.

Chinese merchants in the first millennium CE developed a more sophisticated approximation of true risk pooling, distributing cargo across multiple ships so that a single shipwreck would not destroy any individual merchant’s entire inventory. This is not insurance in the strict sense — it is self-insurance through diversification — but it reflects the same underlying logic: that the statistical unpredictability of individual events can be tamed by aggregating many events, because their random variations tend to cancel out.

The Genoese Invention and the Mathematics of Risk

The transition from informal risk-sharing arrangements to formal insurance contracts is most clearly traceable to fourteenth-century Genoa and the broader northern Italian commercial world. The earliest surviving insurance contract dates from 1347, when Genoese merchants wrote what is recognizably a modern insurance policy: a specified premium, a specified coverage period, a specified peril, and a clear obligation to pay a defined sum on the occurrence of the covered loss.

What made this possible was not just commercial sophistication — it was the development of the legal and accounting infrastructure necessary to make enforceable promises about future contingent events. The Italian merchant cities of the thirteenth and fourteenth centuries were, in this respect, the most institutionally advanced commercial environments in the world. They had developed standardized notarial contracts, double-entry bookkeeping, correspondent banking networks, and commercial courts with genuine capacity to enforce complex financial agreements. Insurance contracts required all of these: a legal framework for defining and measuring losses, accounting systems for tracking premiums and reserves, and enforcement mechanisms that made an insurer’s promise credible to the buyer.

The mathematics followed the institutions, not the other way around. Jakob Bernoulli’s formal articulation of the law of large numbers — which explains why pooling many independent risks produces more predictable aggregate outcomes than bearing individual risks — came in 1713, three and a half centuries after the Genoese merchants were already operating insurance markets. Those merchants did not need the formal theorem to understand intuitively that an underwriter who wrote policies on a hundred voyages would face a more predictable outcome than one who bet everything on a single ship. The actuarial mathematics that now underlies insurance pricing formalized and extended an intuition that commercial practice had already validated.

The crucial implication is that insurance is not primarily a mathematical achievement. It is an institutional achievement. The invention of insurance was the invention of a contractual and organizational framework for making credible promises about future uncertain events — and that framework required legal, accounting, and enforcement infrastructure that took centuries to develop. Societies that lacked that infrastructure could not have functioning insurance markets regardless of how well they understood the underlying probability theory.

How Insurance Enabled Trade at Scale

The economic significance of insurance lies not primarily in what it does for individual policyholders — providing recovery funds when losses occur — but in what it does for economic behavior before losses occur. Insurance enables economic actors to make investments and commitments they would not make if they had to bear the full variance of possible outcomes themselves.

This enabling function is clearest in maritime trade, where insurance had its earliest commercial development. A merchant who had to self-insure a cargo of silk from Venice to London needed either to be wealthy enough to absorb the total loss of the cargo or to send a cargo small enough that its loss would not be catastrophic. Insurance, by transferring the catastrophic loss risk to an underwriter, allowed merchants to ship larger cargoes than their personal wealth could support as self-insurance. This was not just convenient — it was transformative. It expanded the effective scale of commercial operations by allowing merchants to leverage their capital against covered risks rather than having to reserve capital against worst-case scenarios.

The same enabling logic applied, and still applies, to every sector where insurance spread. Fire insurance enabled property owners to invest in more valuable structures and equipment than they could afford to lose. Life insurance enabled families to commit to long-term financial obligations — mortgages, children’s education — without being destroyed by the death of the primary earner. Crop insurance enabled farmers to plant more intensive and productive (but more variable) crops than they would risk without coverage. In each case, the insurance contract redistributed risk from parties less able to bear it to parties — the insurers — specifically organized to absorb it at the lowest possible cost through diversification. And in each case, the redistribution enabled more productive economic activity than would otherwise have occurred.

This is why the development of insurance markets correlates so strongly with economic development: not because wealthy societies happen to buy insurance, but because insurance is a precondition for certain forms of wealth creation. The expansion of seaborne trade in the early modern period required marine insurance. The industrial revolution’s investment in fixed capital — factories, machinery, infrastructure — required fire and casualty insurance. The growth of consumer credit and mortgage finance required life insurance. The causal arrows run in both directions, but the enabling direction is real and underappreciated.

The Problem of Moral Hazard and Adverse Selection

The history of insurance is also the history of the two problems that have plagued insurance markets since their inception and that actuarial science has never entirely solved: moral hazard and adverse selection.

Moral hazard — the tendency of insured parties to take more risk than they would uninsured, because the cost of bad outcomes falls partly on the insurer — is as old as insurance itself. Medieval marine insurers understood that merchants who insured their cargoes might pack them more carelessly, take routes their ships were not designed for, or, in extreme cases, arrange “accidents” that were more profitable than successful deliveries. The deductible, the co-insurance requirement, and the condition that the insured have an “insurable interest” in the covered property — all standard features of modern policies — are direct responses to moral hazard concerns that underwriters identified in the earliest days of the market.

Adverse selection — the tendency for those at higher risk to seek insurance more actively than those at lower risk, generating an insured pool that is systematically riskier than the general population — is the more structurally dangerous problem. An insurer who prices premiums based on the average risk of the general population will find that their actual customers are disproportionately drawn from the riskier segment. If the insurer responds by raising prices, the lower-risk customers drop out, the insured pool becomes riskier still, and prices must rise again — a spiral that can unwind an insurance market entirely.

The history of health insurance in the United States is the most thoroughly documented case study in adverse selection in insurance history, but the mechanism is not peculiar to health insurance. It appears wherever buyers have better information about their own risk than insurers can access, and it has generated continuous institutional innovation: underwriting practices, medical examinations, exclusion clauses, mandatory participation requirements, community rating rules, and public insurance programs that short-circuit adverse selection by making participation compulsory.

The fact that modern insurance markets continue to function despite these inherent problems is itself a testament to the institutional sophistication they have accumulated. They function not because the problems have been solved — they haven’t been — but because the institutional frameworks for managing them have become sufficiently robust to contain them within workable bounds. The Great Fire of London burned the city because the institutional framework for sharing risk had not yet been built. The institutional framework that was built in its wake has, with many failures and reforms along the way, been supporting economic activity for three and a half centuries. That is the real history of insurance: not a product, but an infrastructure — and one of the most consequential infrastructures that commercial civilization has ever constructed.