Insurance

Insurance is a contract in which one party, the insured, pays a premium to another party, the insurer, in exchange for a promise to compensate defined losses or to pay defined benefits when specified events occur. Its basic function is risk transfer: an individual or business exchanges an uncertain and potentially large loss for a known, smaller cost. Insurers make this possible by pooling many policies, so that the losses of the few are paid from the premiums of the many. The idea rests on probability, expected-value and the law-of-large-numbers, and it is shaped throughout by problems of information-asymmetry, adverse-selection and moral-hazard. Insurance matters because it allows households and firms to take on economic activity they could not safely undertake alone, supports lending and investment, and cushions shocks such as illness, accidents, natural disasters and death. It is also a major financial sector in its own right, with large pools of invested premiums and a central role in pensions, health financing and public policy.

This entry covers the core principles of the insurance contract and the history that produced them, the mathematics of pooling and pricing, and the main product lines: life, health, property, casualty and liability. It then examines reinsurance, market structure and the variety of organizational forms, from stock companies and mutuals to Lloyd's and takaful. The entry also addresses social insurance, regulation and solvency supervision, and the macroeconomic role of insurers. Later sections turn to access and microinsurance in developing economies, the behavioral foundations of demand, catastrophe risk and climate change, digital technology, and the open questions that remain contested.

Core principles of the insurance contract

An insurance contract is a promise by an insurer to pay a defined sum or provide a defined service if a specified contingency occurs, in exchange for a premium paid by the policyholder. Economically, it moves a risk the insured already bears onto a party better able to absorb it, usually by pooling that risk with many others. This distinguishes insurance from gambling. A bettor creates exposure to a new risk in pursuit of gain, whereas an insured person seeks to shed exposure they already carry. Insurance also differs from saving. A deposit returns the depositor's own money with interest whether or not anything goes wrong, while a standard non-life premium paid by a customer who suffers no loss is not refunded, because it helps fund other people's losses. Some life products blur this line, so the distinction is one of degree rather than a sharp boundary.

Two conditions determine which risks can be insured. The first is fortuity: the loss must be uncertain from the insured's standpoint, either in whether it occurs or when. Life insurance covers a death that is certain eventually but unknown in timing, which is why it can be written at all. The second is insurable interest: the insured must stand to suffer a financial loss from the event. This requirement keeps contracts from functioning as wagers on other people's misfortune. In England, the Life Assurance Act 1774 made insurable interest a statutory condition for life policies, and comparable rules appear in most legal systems.

The principle of indemnity holds that a claim payment should restore the insured to the financial position held just before the loss, not leave them better off. It governs most property and liability contracts and is reinforced by two mechanisms. Subrogation allows the insurer, once it has paid, to take over the insured's rights against a third party who caused the loss, so the insured cannot recover twice. Contribution applies when several policies cover the same loss: the insurers share the payment, typically in proportion to their respective exposure. Not every contract follows strict indemnity. Valued policies that pay an agreed amount, and many life policies that pay a fixed sum, are exceptions.

Insurance contracts are also treated as contracts of utmost good faith (uberrimae fidei), a doctrine that grew out of marine insurance. Because the applicant usually knows far more about the risk than the insurer does, the applicant must disclose all material facts, whether or not asked, and the insurer must describe its terms honestly. The classic statement is Carter v Boehm (1766), in which Lord Mansfield held that a party with superior knowledge of the risk must disclose it. Non-disclosure or misrepresentation can allow an insurer to avoid a policy or deny a claim, although the tests for materiality and the available remedies differ across jurisdictions. This is a direct response to information asymmetry and the resulting adverse selection risk.

Actual coverage depends heavily on policy wording. A deductible (or excess) is the amount the insured pays before the insurer contributes, which removes small claims and reduces moral hazard. Co-insurance requires the insured to bear a fixed percentage of each loss above the deductible, keeping them partly exposed to every claim. Limits cap the insurer's payment per event or per period, and sub-limits cap particular perils within a policy. Exclusions remove specified causes, property or circumstances from cover, such as war or flood under many standard property forms. Conditions impose duties on the insured, such as prompt notice of loss, reasonable steps to protect property from further damage, and cooperation with investigators; failing to meet them can jeopardize a claim.

When wording is disputed, courts apply interpretive rules within contract law. The contra proferentem rule construes ambiguous language against the party that drafted it, normally the insurer. Exclusions are usually read narrowly, and the insurer generally bears the burden of showing that an exclusion applies. Some jurisdictions, including certain U.S. states, also recognize a reasonable-expectations doctrine that protects the insured's understanding of coverage even where the text is unclear. Courts generally read the policy as a whole and in light of its purpose, but the weight given to these principles varies considerably across legal systems.

Historical development

Antecedents

Risk-sharing in commercial form is very old. The best-documented early device is the bottomry contract: a merchant borrowed money against a ship or its cargo, and if the vessel was lost the debt was cancelled. The arrangement is usually tied to Mesopotamia and the Code of Hammurabi, which dates to about 1750 BCE. Britannica, however, places bottomry dealings among Babylonian merchants as early as 4000–3000 BCE, so the chronology is uncertain. The practice was known in India around 600 BCE and was well understood in ancient Greece by the 4th century BCE. Roman law also recognised comparable contracts. Lenders charged high interest because the interest also priced the risk of loss, which makes these loans an early, implicit form of risk transfer. Under bottomry contracts, the lender in effect absorbed the voyage's peril.

In the Middle Ages, guilds and similar associations offered members help after fires, illness and deaths, funded by members' regular contributions. Burial societies performed a similar function for funeral costs. These mutual arrangements were local and informal, but they pooled small contributions to cover individual misfortune.

Early modern expansion

Marine insurance became organised in London around the coffee house run by Edward Lloyd. The first known reference to the establishment appears in the London Gazette in 1688. Because it attracted shipowners, merchants and sea captains, it became the place where underwriters met to take on marine risks. Lloyd's List, which began as a shipping-news sheet, grew out of this trade. Lloyd's was reorganised in 1769 as a formal group of underwriters. Its history also includes a darker side: the historian Eric Williams noted that Lloyd's insured slave ships during the slave trade.

Property insurance followed the Great Fire of London. The fire began on 2 September 1666 in a bakery on Pudding Lane and burned for four days, destroying more than 13,000 houses. Nicholas Barbon, a physician and builder, founded the Fire Office, commonly dated to 1680, which sold annual fire policies and maintained its own firefighters. Insurers soon used fire marks, metal plaques that identified insured buildings, though the system left uninsured properties exposed. In 1833, ten fire insurers jointly created the London Fire Engine Establishment, which was funded by the industry and became London's first fire service.

Life insurance developed more slowly and depended on mortality data. The Society for Equitable Assurances on Lives and Survivorships was established by deed of trust in September 1762. It was organised on a mutual basis, so its surplus belonged to its policyholders. It pioneered premiums based on age and mortality rates, which laid the groundwork for scientific life insurance practice.

Nineteenth and twentieth centuries

Industrialisation expanded the demand for insurance. Factories, railways and urban housing concentrated large values in small areas, and the growth of wage labour made family income vulnerable to death and injury. In the United States, mutual life companies appeared in the 1840s, while fraternal societies became major providers of life cover for working-class people after the Civil War. Governments began to license and inspect insurers, and supervision became a permanent part of the industry.

Social insurance emerged as a state responsibility. In Germany, the Bismarck government introduced compulsory sickness insurance in 1883, accident insurance in 1884 and old-age and disability insurance in 1889. Similar schemes spread across industrialised countries in the following decades, notably in Britain before the First World War. These programmes, which are described in more detail in the social insurance section, shifted part of the risk from individuals and charities to statutory funds.

Post-war consolidation and quantitative pricing

After the Second World War, many countries expanded public insurance systems, and private markets increasingly operated across borders. Reinsurance and multinational groups grew, and the industry consolidated through mergers and acquisitions. Pricing also became more quantitative. Actuaries used larger data sets and, later, computer modelling to estimate claim frequencies, severities and mortality trends, replacing much of the older reliance on rules of thumb. This shift laid the foundation for the risk-based solvency regimes that followed in later decades.

Pooling, probability and the law of large numbers

Pooling and the variance of aggregate losses

Consider $N$ policies whose annual claim amounts $X_1,\dots,X_N$ are independent random variables with a common expected value $\mu$ and variance $\sigma^2$. The insurer's total claims $S=\sum_{i=1}^{N}X_i$ have expected value $E[S]=N\mu$. Because independence makes all covariances zero, the variance of the total is $\operatorname{Var}(S)=N\sigma^2$. The average loss per policy, $\bar X=S/N$, keeps the expected value $\mu$, but its variance falls to $\sigma^2/N$ and its standard deviation to $\sigma/\sqrt{N}$. Pooling therefore leaves the expected loss per policy unchanged while reducing its volatility in proportion to $1/\sqrt{N}$.

A simple case shows the scale. Suppose a policy pays 10,000 monetary units with probability 0.01 and nothing otherwise. The expected loss is 100. The variance is $0.01\times10^{8}-100^{2}=990{,}000$, so the standard deviation of one policy's loss is about 995, nearly ten times its mean. In a pool of 10,000 such policies, the average loss per policy still has mean 100, but its standard deviation drops to about 10. The coefficient of variation (standard deviation divided by mean) falls from roughly 10 to roughly 0.1.

The law of large numbers and normal approximation

The law of large numbers states that the sample average of independent, identically distributed observations converges to the expected value as the sample grows. This is the formal reason that realized claims per policy in a large book tend toward the figure on which premiums were set. The law does not describe the remaining fluctuation over a finite period. For that, insurers use the central limit theorem: for large $N$, the aggregate claims $S$ are approximately normally distributed with mean $N\mu$ and standard deviation $\sigma\sqrt{N}$.

If an insurer wants probability $\alpha$ that its reserve covers claims, it can set the reserve at approximately

where $z_\alpha$ is the standard normal quantile (about 2.33 for $\alpha=0.99$). In the example above with $N=10{,}000$, expected claims are 1,000,000. The 99% reserve is about $1{,}000{,}000+2.33\times99{,}500\approx1{,}232{,}000$, a margin of roughly 23% over expected claims. Because the margin scales with $\sqrt{N}$ rather than $N$, it becomes a smaller share of expected claims as the pool grows.

The normal approximation is unreliable when $N$ is small or when claims are rare and severe, because the loss distribution is then strongly skewed. In such cases actuaries model claim counts with a Poisson distribution and use compound models, often estimated by Monte Carlo simulation, to obtain tail quantiles.

Limits of pooling: correlated risks

The argument above depends on independence. If each policy has variance $\sigma^2$ and each pair of policies has correlation $\rho$, the variance of the average loss is

which approaches $\rho\sigma^2$ as $N$ grows. No matter how large the pool, a common component cannot be diversified away. With $\rho=0.05$ and $N=10{,}000$, the standard deviation of the average is about 0.22$\sigma$, compared with 0.01$\sigma$ under independence, which is more than twenty times larger.

Correlation arises from events that strike many policies at once. An earthquake damages many buildings in the same region. An pandemic raises death and illness claims across life and health books simultaneously. A systemic financial shock depresses the value of investment portfolios and raises the risk of default by reinsurance counterparties. Heavy-tailed severities create a related problem: even independent losses converge slowly to the normal benchmark, so the approximation understates the probability of extreme years. Insurers respond with catastrophe models, geographic limits, reinsurance, and additional capital.

A worked illustration of the pooled premium

Take the same risk, with an expected loss of 100 per policy, written in a pool of 10,000 policies. The gross premium is built from several components:

1. Pure premium: the expected loss, 100.
2. Expense loading: administration, commissions and claims handling, assumed at 20.
3. Capital charge: the insurer must hold a buffer above expected claims. For the pool, the 99% buffer is about $2.33\times99{,}500\approx231{,}000$, or about 23 per policy. A 6% annual cost of capital on this amount gives roughly 1.4 per policy per year. Because capital needs grow with $\sqrt{N}$, the per-policy charge falls as the pool grows.
4. Profit margin: 5.
5. Investment income offset: premiums are collected before claims are paid, so the float earns a return. Assume this offsets 3 per policy.

The resulting gross premium is $100+20+1.4+5-3=123.4$. The total loading of 23.4 is about 23% of the pure premium. The capital component is small here only because the pool is large; a single policy would require a much larger buffer per unit of exposure, and the normal approximation would be too crude to rely on.

Information problems: adverse selection and moral hazard

Adverse selection

Insurers price a policy against the average risk of a group, but applicants often know more about their own risk than the insurer does. This information asymmetry means the people most likely to claim are also the most inclined to buy, a pattern economists call adverse selection. If a pooled premium is set at average expected loss, low-risk applicants see it as overpriced and leave, which raises the average risk of those who remain and can push premiums higher still.

The classic analysis is Rothschild and Stiglitz's 1976 model of competitive insurance. In it, a single pooled contract is unstable, because a rival policy aimed only at low-risk buyers can screen them out of the pool. The equilibrium therefore has to separate risk types, with each group holding a different contract. This separation imposes a cost on low risks, who are only partially insured due to the self selection constraint. Their coverage is deliberately reduced so that high-risk buyers do not switch to their cheaper policy. The model also has a sharp limitation: a competitive equilibrium in pure strategies may not exist at all. Later work showed that considering firms which face capacity constraints, which might be due to limited capital, an equilibrium exists, where every consumer obtains his appropriate Rothschild–Stiglitz contract. When losses are continuous rather than all-or-nothing, the design problem changes too, and the choice between a deductible and co-insurance becomes central to offering partial cover to low-risk buyers.

Moral hazard

Moral hazard arises after a policy is bought. Because insurance shifts the cost of a loss onto the insurer, policyholders may take less care, drive more, or claim more readily. When the model is extended to include ex post moral hazard, both high and low risk individuals typically purchase less than full insurance, even in the absence of adverse selection. Insurers therefore usually keep some of the loss with the insured, using deductibles (a fixed amount paid before cover starts), co-payments or co-insurance (a percentage share of each claim), and no-claims bonuses that reward a claim-free record with lower premiums. Each device restores some of the policyholder's incentive to avoid losses, though at the cost of leaving them partly exposed.

Tools against information problems

Insurers try to reduce asymmetry before and during the contract. Underwriting questionnaires and medical examinations collect information at the point of application. Experience rating adjusts a firm's premium to its own claims history, and credit-based or behavioural data are used in some lines as proxies for risk. Telematics in motor insurance is a recent example of continuous monitoring. Compulsory participation is a different approach: when everyone in a class must buy cover, as with many motor third-party liability schemes or public pension systems, the insurer cannot lose low-risk members to the market, which weakens adverse selection.

Which information may be used: the genetic test debate

Genetic test results are the most contested case, because they can reveal risks that an applicant may not yet show. In the United States, the Genetic Information Nondiscrimination Act of 2008 prohibits discrimination against healthy individuals for employment decisions or health insurance purposes on the basis of genetic information alone. The United Kingdom has relied instead on a voluntary arrangement. The Association of British Insurers agreed not to ask people seeking insurance to reveal their DNA test results unless they were seeking life cover above £500,000 or critical illness cover above £300,000. Above those amounts, test results could only be used with prior approval of the Genetics and Insurance Committee, which has so far only approved the use of test results for Huntington's Disease for life insurance. The moratorium was later extended, and critics such as Sir John Sulston have called for statutory rules.

The debate turns on whether restricting information protects solidarity and genetic privacy or leaves insurers exposed to selection. Supporters of using test results argue that insurers need them to set fair premiums, while opponents see such use as discrimination against people with known predispositions. The thresholds show the practical compromise: low-value cover is shielded, while large policies, where selection and fraud concerns are greatest, remain open to scrutiny.

Actuarial science and premium setting

Actuarial work starts from two quantities for any block of risks: claim frequency (how often losses occur) and claim severity (how large each loss is). Counts of events are often modelled with a poisson-distribution, while severity is fitted with skewed distributions, because a small number of very large claims can dominate total cost. Combining the two, analytically or by simulation, gives the distribution of annual aggregate losses. The expected-value of that distribution anchors the price, and its variance and tail behaviour determine how much capital the insurer must hold against bad years.

Life and annuity pricing relies on a different input. Insurers use mortality-rate tables that give the probability of death at each age, usually adjusted for expected future improvement in longevity. Because an annuity pays for as long as the annuitant survives, small changes in projected life-expectancy can move the price materially. Sex-based differences show how these inputs interact with law. In life insurance and pensions, what matters is the potential life expectancy of the insured, which, for women, is statistically proven to be longer. Conversely, in motor insurance, young men are likely to make more expensive claims than women. Under the EU's 2004 Gender Directive, member states could allow such differences where they rested on relevant and accurate actuarial and statistical data. In the Test-Achats judgment, the Court of Justice held that this derogation was invalid, with effect from 21 December 2012 for new contracts. Insurers must price annuities and other products on a gender-neutral basis for new contracts from that date.

Components of a premium

A premium is normally built up in layers. The starting point is the pure premium, the expected loss per unit of exposure before any expense or profit is added. If 1,000 car-years produced $200,000 in losses, the pure premium is $200 per car-year. The pure premium is meant to cover losses and loss-related costs, while the loading is the part of the premium necessary to cover other expenses, particularly sales expenses, and to allow for a profit. Loadings typically include commissions and other acquisition expenses, premium taxes, general administrative expenses, contingency allowances and profit. A separate risk load sits on top of the pure premium: it is the surplus amount over the pure premium which is required by the insurer to achieve an expected return, meaning the capital that must be held against the policy. Investment income offsets part of the charge, since premiums are collected before many claims are paid and can be invested in the meantime; this offset is largest for long-tail life and liability business.

With proportional loadings, the gross premium $G$ for an expected loss $\mathrm{E}[L]$ is

where $e$ is the expense ratio and $\pi$ the profit and capital margin. If expected loss per policy is 600 units, $e = 0.25$ and $\pi = 0.05$, the gross premium is about 857 units, of which 600 pays claims and roughly 257 covers costs, capital and profit. Simplified formulas of this kind ignore investment income and differences in the timing of cash flows, which real ratemaking models include.

Reserving for unsettled and unreported claims

Because claims are settled over months or years, the premium earned in a year must be matched against losses that are not yet fully known. Insurers therefore hold case reserves for reported claims, incurred but not reported (IBNR) reserves for claims that have occurred but have not yet been reported, incurred but not enough reported (IBNER) reserves for underestimated case estimates, and reserves for claims-handling expenses. IBNR is inherently an estimate: since the insurer knows neither how many of these losses have occurred nor the severity of each, IBNR is necessarily an estimate. Adding it to reported losses gives the estimate of the total eventual liabilities the insurer will cover, known as ultimate losses.

Reserve estimates feed directly into reported results. Because a reserve is a liability on the balance sheet, a low estimate flatters current-year profit, while a later strengthening of reserves depresses it. Accuracy therefore matters to investors, regulators and management alike. The problem is most acute in long-tail lines such as workers' compensation and professional liability, where reporting and settlement delays are long.

Rating factors and regulatory limits

Rating factors are the observable characteristics used to differentiate premiums, such as age, driving record, vehicle, location and health status. Their use is constrained by anti-discrimination law as well as by statistics. The UK's Equality Act 2010 had contained an insurance exception for sex-based pricing, which the government repealed effective 21 December 2012 in response to the EU judgment. Location-based rating raises a related concern: a factor that predicts losses well may also act as a proxy for race or income. Regulators in many jurisdictions therefore require insurers to justify each factor with actuarial evidence while limiting the characteristics that may legally be used. The central tension is between predictive accuracy, which favours finer classification, and fairness, which favours pooling risks across groups that cannot be distinguished.

Life insurance and annuities

Life insurance pays a benefit on the death of the insured, or in some contracts on survival to a set date, in exchange for premiums. Products differ mainly in how long cover lasts and whether the contract builds a savings balance. Mortality estimates and life-expectancy projections drive pricing in every case, but the distribution of risk differs sharply between product types.

Term life covers a fixed number of years and pays only if death occurs within that window. Because it accumulates no value, it is the cheapest form of cover per unit of benefit, though premiums rise with age if the policy is renewed. Whole life covers the insured for life, typically with level premiums that over-fund the early years and create a cash surrender value. Endowment policies pay out either on death within the term or on survival to maturity, so they combine protection with a sizeable savings element. Universal life separates the death benefit from a flexible account credited at a rate the insurer declares, subject to a guaranteed floor. Unit-linked and variable products invest premiums in funds chosen by the policyholder, who therefore bears the investment risk rather than the insurer.

Annuities reverse the logic of life insurance. Instead of paying a benefit on death, the insurer pays an income for as long as the annuitant survives, so the product is essentially insurance against living too long. Pooling works in the opposite direction: those who die early help fund payments to survivors. This "longevity insurance" function is why annuity pricing depends so heavily on mortality improvement assumptions. The protection-versus-savings distinction shapes how products are priced and sold. Pure protection contracts are priced on mortality tables and sold on the basis of financial need, whereas savings-type contracts are priced around the credited interest rate and sold more like investments, which brings in conduct rules on disclosure and suitability.

Long-duration contracts create specific financial pressures. Policyholders may surrender a contract early, and the insurer then pays a surrender value that is usually reduced by charges, especially in the first years; these charges protect the remaining pool against the selection that early surrender can produce. Policy loans let holders borrow against accumulated cash value, and any outstanding balance reduces the eventual benefit. Because liabilities can run for decades, insurers try to match them with long-term bonds and other assets, and mismatches between the timing of cash outflows and asset maturities create liquidity and interest-rate risk. If many policyholders surrender at once, for example when rates rise and better products appear elsewhere, the insurer may have to sell assets at a loss, a dynamic that resembles a bank-run in a smaller form.

Market conditions have shifted considerably in recent years. Swiss Re's sigma research estimated that total global life insurance premiums should reach USD 4.8 trillion by 2035, up from USD 3.1 trillion in 2024, driven by higher interest rates. A later edition raised the 2035 figure to USD 5.1 trillion, with the increase primarily driven by saving premium rises, and projected that primary global life insurance premium growth [would] slow to 1% this year in real terms, after a strong 6.1% gain in 2024. The rate environment has also changed the mix of products. Swiss Re expected that consumers in advanced markets will increasingly pivot from fixed annuities towards index-linked policies in the following two years, as central banks lower interest rates. In Europe, unit-linked sales are rising strongly, particularly in Italy and France.

The United States illustrates both the rate effect and its complications. LIMRA reported that total U.S. annuity sales increased 6% to $461.3 billion in 2025, and that indexed products — registered index-linked and fixed indexed annuities — represented 45% of total sales in 2025, up from just 24% market share a decade ago. Fixed-rate annuities, however, more than tripled between 2021 and 2025 as rising interest rates made guaranteed products far more competitive. LIMRA also noted that sales were elevated in the third quarter of 2025 as investors rushed to lock in rates before anticipated interest rate cuts. The two products sit at different points on the risk spectrum. In a fixed indexed annuity, the crediting rate depends on an index through tools such as caps, participation rates and spreads, and the specified floor, or the minimum interest rate, is typically 0%. Surrender periods on such contracts range between 3 and 16 years; typically about ten. Behind these guarantees, the insurer uses its own investments—mostly bonds and similar instruments—to generate the income that supports the guarantees.

Health insurance: private and public models

Forms of coverage

Private health cover takes several forms. Under a classic indemnity or fee-for-service design, the patient may choose any clinician and the insurer repays a share of the charges. Most modern plans instead work through contracted networks. Doctors and hospitals accept reduced rates, and the plan's cost-sharing rules favour in-network care The primary way in which managed care plans work is by establishing provider networks. A provider network serves plan members over a certain geographic area in which the health plan is available. The providers in these networks agree to offer their services at reduced costs. Your health plan pays more of the cost of your care if you see providers in the network.

The main managed-care variants differ mainly in how they treat referrals and out-of-network use. HMO plans require members to choose an in-network PCP to help manage their care and provide referrals. HMOs will not cover care from an out-of-network provider. PPOs will, but you'll likely save when you stay in network. Finally, HMOs generally have lower premiums and out-of-pocket costs. Hybrids also exist. Point of Service (POS) plans are a hybrid of HMOs and PPOs. Exclusive provider organisations, a further variant, combine features of both and tend to sit between HMO and PPO prices. Supplemental policies, which pay for gaps such as deductibles or co-payments left by a primary plan, are a separate layer on top of these structures.

Cost sharing and the RAND experiment

The most cited study of how patients respond to cost sharing is the RAND Health Insurance Experiment, which in the 1970s and early 1980s randomly assigned families to free care or to plans with differing coinsurance rates. Participants who paid for a share of their health care used fewer health services than a comparison group given free care. The reduction came from fewer contacts rather than cheaper care: reduced spending resulted entirely from less use of care; the costs of care were not affected. Critically for the moral-hazard interpretation, cost sharing cut the use of both highly effective and less effective services in roughly equal proportions. The overall picture was mixed. Cost sharing did not significantly affect the quality of care received by participants. The exceptions were concentrated among the poorest and sickest participants, for whom free care led to improvements in hypertension, dental health, vision, and selected serious symptoms.

The study's reach has limits. Its enrolees were adults under 65, so one should be cautious about drawing conclusions from the RAND HIE results for the Medicare population. Later research has focused on specific drugs. Evidence summarised by RAND-linked reviews finds that for each 10% increase in cost sharing, prescription drug spending decreases by 2% to 6%, depending on class of drug and condition of the patient. Findings on preventive care are not uniform either, and newer studies have reported conflicting findings, particularly regarding preventive care utilization.

Public, mixed and private financing

Health systems are usually classified by how they raise money. In tax-funded models, such as the United Kingdom's National Health Service, care is largely paid from general taxation. Social health insurance, associated with the German tradition, is financed by compulsory contributions into insurance funds. With a tax-financing system, the financial burden is shared among the whole population within a single scheme applying to all. Contributions to social insurance, by contrast, are often tied to earnings and capped, which limits risk-sharing and may reduce progressivity. However, risk sharing is limited in social health insurance, as contributions are often proportional to income. Private premiums spread risk among people in similar risk categories, while direct payments at the point of use involve no risk pooling.

Most countries combine these channels. Most health systems use a mix of these. Evidence on which mix works best remains contested. One review concludes that tax funded systems seem to be a more progressive health care financing mechanism than systems based on social insurance in low- and middle-income countries. That finding, however, is sensitive to the choice of ability to pay measures, and other analysts argue against financing health care through labour taxes in poorer countries. Each design has trade-offs. Tax funding achieves universality automatically and has the lowest administrative costs. Contribution-based systems earmark revenue for health, insulating it from general budget pressures, but introduce coverage gaps for informal workers.

Access, affordability and coverage

The central tension is between insurance coverage and the actual cost of care. Cost sharing that deters unnecessary use can also deter needed use, and the RAND results suggest the burden falls hardest on people who are poorest and sickest. Whether low-income groups respond more strongly to cost sharing is itself disputed, since one review found little evidence to support this contention. The goal of universal health coverage is usually pursued through public or mixed systems precisely because voluntary private markets leave some groups uninsured or underinsured, and because the affordability of premiums depends on how risk is pooled and who is compelled to participate.

Property, casualty and liability insurance

Non-life insurance is commonly split into property and casualty lines. Property insurance pays for physical loss to things the insured owns: homes, the contents of homes, commercial buildings, equipment and stock. Business interruption cover extends this to the income a firm loses while physical damage stops its operations. Casualty, or liability, insurance pays what the insured legally owes to third parties for bodily injury or property damage. It is the branch in which litigation and court awards drive costs most directly.

Property lines and settlement basis

How a claim is paid depends heavily on the settlement basis written into the policy. Under actual cash value (ACV) settlement, the insurer deducts depreciation for age and wear from the cost of replacement, so an old roof or appliance pays out less than a new one costs. Under replacement cost value (RCV) settlement, depreciation is not deducted, and the payout reflects what a new equivalent costs today, subject to the policy limit. RCV is usually more expensive in premium. In many homeowners contracts the structure is insured at replacement cost while personal contents default to ACV unless the policyholder buys an upgrade. Business interruption cover pays for lost profit and continuing fixed costs during a restoration period, and it usually responds only when the interruption follows an insured physical peril. Flood and earthquake are excluded from many standard property policies in some markets and must be bought separately, which is one reason the property market is closely tied to climate change and catastrophe exposure discussed elsewhere in this entry.

Motor insurance and compulsory third-party liability

Motor insurance is the most widely compelled form of non-life cover. Compulsion began in the early twentieth century: the United Kingdom's Road Traffic Act 1930 required owners and drivers to insure their liability for injury or death to third parties, Ireland followed in 1933, and the European Community required compulsory cover in all member states from 1973. The core product is third-party liability cover, which protects other road users rather than the insured vehicle. Designs differ. Australia runs state-based Compulsory Third Party schemes that cover personal injury only, while property damage and own-vehicle cover are bought separately. In France, where fault is shared, insurers divide responsibility according to a fault percentage. India's Motor Vehicles Act requires at least third-party cover, yet enforcement remains uneven. Where compulsion is weakly enforced, uninsured drivers can leave victims uncompensated, and some countries respond with backstops funded by levies. South Africa, for example, funds a Road Accident Fund partly through a fuel levy.

Liability lines and the cost of litigation

Liability insurance covers the insured's legal defence costs and damages arising from alleged negligence or wrongdoing. General liability covers third-party bodily injury and property damage arising from business premises and operations. Professional (errors and omissions) cover responds to alleged negligent advice or service. Product liability covers injuries caused by goods an insured makes or sells. Directors-and-officers cover protects individuals and, often, the company's reimbursement of defence costs and settlements. Liability claims are long-tailed: injuries may surface years after a policy expires, so reserves are uncertain and depend heavily on court outcomes.

Since the 2010s, the cost of liability claims has risen faster than general inflation, a phenomenon often called social inflation. The Swiss Re Institute has reported that US liability claims rose by 57 percent over a decade. Jury awards above $10 million, known as nuclear verdicts, have grown in both number and size. Estimates differ by compiler: Marsh reported total nuclear-verdict payouts of $14.5 billion in 2023, while other data sets report 2024 figures that rose sharply again. Drivers commonly cited include noneconomic damages, shifting jury attitudes and the growth of litigation funding. Insurers respond by tightening underwriting, adding excess and umbrella layers, adjusting reserves and placing more weight on litigation management and settlement strategy.

Specialty lines

Specialty lines cover risks that are large, technical or poorly suited to standard forms. Marine insurance includes hull cover for vessels, cargo cover and liability cover; shipowners' third-party liabilities are often placed with mutual protection and indemnity clubs. Aviation insurance covers aircraft hulls, passenger and third-party liability, and war risks. Workers' compensation provides statutory, largely no-fault benefits for employees injured at work, and in several systems it is administered through state funds or employer-funded arrangements. Environmental (pollution) liability is often separated from general liability because standard policies exclude gradual or intentional pollution. Cyber insurance, a comparatively young line, combines first-party costs such as data restoration, business interruption and extortion response with third-party liability for privacy breaches. Its main difficulty is accumulation: a single software vulnerability or cloud outage can produce many correlated claims, a concern linked to systemic risk and to broader cybersecurity practice. Disputes over war and state-backed attack exclusions are a further source of contention.

Reinsurance and the sharing of risk

Reinsurance is insurance bought by insurers. A primary insurer, called the cedent, passes part of its exposure to a reinsurer in return for part of the premium. It does this for three main reasons. The first is capacity: a company can write more business than its own capital would safely support. The second is capital relief, because regulators generally require less capital against risks that have been transferred. The third is protection against accumulation, where one event such as a hurricane damages thousands of policies at once. The cedent remains liable to its own policyholders, so the arrangement shifts the cost of certain losses rather than the legal obligation to pay. Under a reinsurance contract, an insurer is indemnified for losses occurring on its insurance policies and covered by the reinsurance contract.

Reinsurance contracts are usually grouped by how they divide losses and by what they cover. Proportional arrangements split premiums and losses by a fixed ratio. In a quota share, the reinsurer takes a set percentage of every policy, so a 40% quota share on a 1,000,000 loss means the reinsurer pays 400,000. A surplus treaty works similarly but cedes a share that depends on each risk's size relative to the cedent's retention. Proportional reinsurance arrangements allocate risk proportionally between the ceding insurance company and the reinsurer. When a so-called "quota share" arrangement is negotiated, the reinsurer and ceding company share premium generate and policy losses on a pre-agreed fixed percentage basis.

Non-proportional cover pays only when losses exceed a threshold. A reinsurer might agree to cover the ceding company for USD 50,000 any one risk (cover or limit) excess of USD 10,000 (the retention or deductible or priority). Excess-of-loss treaties can respond per risk or per event, while stop-loss treaties cover a whole portfolio. Facultative reinsurance differs from treaty cover in scope. Unlike treaty reinsurance, which covers multiple policies, facultative reinsurance is negotiated separately for each insurance policy. It is typically used for large or unusual exposures that the treaty does not capture. Reinsurers can also buy their own protection, a practice known as retrocession, which spreads concentrated catastrophe risk further through the system.

Reinsurance has also moved into capital markets. Catastrophe bonds transfer specified natural-catastrophe risk to investors, who lose some or all of their principal if a defined trigger is hit and otherwise receive a coupon. Risk spreads fell to levels last seen before Hurricane Ian, yet the market kept its momentum. All-in yields remained attractive compared to expected losses and returns from traditional fixed-income investments, according to Swiss Re's first-half 2026 analysis. Estimates of the outstanding market differ slightly by source. Artemis reported that the outstanding catastrophe bond market ended H1 2026 at a record size of almost $65.6 billion. Issuance has also grown, and a sustained boom in capital supply helps explain why these bonds have become a larger share of global property-catastrophe capacity.

Reinsurance prices move in a cycle. In a hard market, losses and depleted capital reduce capacity, so rates rise and terms tighten. In a soft market, quiet loss years and abundant capital push prices down. Recent renewals show the soft phase clearly. The Global Property Catastrophe Rate-On-Line Index has now fallen by 23% from the level it had reached at its most recent hard market peak in 2024. Howden's January 2026 renewal report put the average risk-adjusted fall at 14.7%, accelerating from the 8% fall recorded in 2025. Guy Carpenter attributed the trend to benign losses, coupled with abundant reinsurer capacity and growing risk appetite. Gallagher Re estimated that insured catastrophe losses in the first half of 2026 were about 28% below the 10-year average of US$64 billion.

For primary insurers, the cycle feeds directly into pricing. Reinsurance is a major cost component for property-catastrophe business, so softer reinsurance terms can ease pressure on primary premiums, while a hard market tends to raise them and can lead insurers to reduce exposure in high-risk areas. The effect is uneven: loss-affected programmes often still see increases even when the wider market is falling, as the broker data show.

Market structure and organizational forms

Ownership forms

Insurers differ first in who owns them, and that answer determines whose interests the company serves. A stock insurer is owned by shareholders, who elect the board and receive profits as dividends. Its policyholders are customers and have no say in management unless they also hold shares. A mutual insurer has no outside owners. Its policyholders are its members, who usually elect the board and may receive surplus as policy dividends, mainly on participating life contracts. Mutuals find it harder to raise equity because they must rely on retained earnings, debt, or borrowing from members. Many former mutuals have demutualized to gain access to capital markets, and mutual holding companies keep a mutual parent above a stock subsidiary. Mutuals remain significant in life insurance. In the United States in 2022, about 110 of 727 life insurers were mutual, yet their in-force life insurance exceeded half the amount held by stock companies.

Other forms are smaller but distinct. Reciprocal exchanges are groups of subscribers who insure one another through an attorney-in-fact who runs operations, a model found mainly in American property and auto lines. Cooperatives are owned by members such as farmers or local communities, often with links to agricultural or banking networks.

A captive insurer is an insurance company that is wholly owned and controlled by the businesses it insures. A single-parent (pure) captive covers its parent and affiliates. A group captive is jointly owned by several firms, often in the same industry or trade association. Rent-a-captive or sponsored arrangements let clients buy space in a vehicle owned by an outside sponsor. Companies use captives to retain predictable losses, reach reinsurance markets directly, and reduce dependence on the commercial market. Estimates of their prevalence vary widely: one industry source puts the share of Fortune 500 companies with captive subsidiaries near 90 percent, while a Utah regulatory publication cites about 40 percent of major U.S. corporations. Tax treatment depends on the domicile and on risk-distribution tests.

Lloyd's of London

Lloyd's is a marketplace rather than a single insurer. It is a corporate body governed by the Lloyd's Act 1871 and its successors. Its origins lie in Edward Lloyd's coffee house, where marine underwriting took shape in the late seventeenth century. Business is written by syndicates. Each syndicate is formed by one or more members, who provide capital and share in profits or losses in proportion to their stake. Managing agents run the syndicates, while brokers and coverholders bring business to them. Members were once mainly individuals known as Names, but corporate capital now predominates. At the end of 2025 there were 110 active syndicates.

Lloyd's governance centers on a layered capital structure often called the Chain of Security. The first link is the syndicate's own assets, including premiums held in trust for policyholders. The second is capital that members deposit at Lloyd's, sized to a Solvency Capital Requirement calibrated to a 99.5 percent confidence level over the full life of liabilities. The third is central assets, including the Central Fund, a pooled reserve that backs every syndicate. Because the security is shared, the market as a whole carries the ratings, and the market now falls under dual supervision by the UK's prudential and conduct regulators.

Intermediaries and distribution

Brokers usually act for the buyer, and agents act for the insurer, so the two channels create different incentives around commission and advice. Bancassurance sells insurance through bank branches and customer relationships. It offers insurers wide reach at low marginal cost and is especially prominent in parts of southern Europe and East Asia. In exchange, insurers often depend on a single banking partner, and the bank's bargaining power can shape pricing and commission levels. Direct digital channels, including websites, apps, and comparison platforms, cut out much of the intermediary cost and make prices easier to compare. This tends to intensify price competition, especially in motor and home insurance, and it pressures traditional agents to justify their fees through service or advice.

Concentration, entry barriers and the cycle

Market concentration varies by line and country. Large global groups dominate commercial and reinsurance business, while personal lines in many national markets are served by a few large composite insurers alongside numerous small or local mutuals. Entry is costly. New firms need substantial capital, licensing, and, for commercial clients, an acceptable financial strength rating. Distribution relationships, underwriting data, and claims-handling scale also protect incumbents, and trust matters especially for long-duration life products.

Pricing moves in cycles. In a soft market, capital is plentiful, competition drives rates down, and terms loosen. A hard market usually follows large catastrophic losses or a depletion of capital. Capacity contracts, premiums rise, deductibles increase, and exclusions tighten. Hard markets have also encouraged firms to form captives or group arrangements, as they seek to escape high commercial rates. Because reinsurance supply feeds directly into primary capacity, the reinsurance cycle is a key driver of these swings.

Social insurance and public programs

Social insurance refers to compulsory, state-organized programs that cover predictable life risks such as old age, unemployment, disability and work injury, and that are usually funded by payroll contributions tied to earnings. Unlike a private policy bought by an individual, membership is typically set by statute, contributions are often a fixed share of wages up to a ceiling, and benefits are defined by rules rather than by each person's own risk. Many systems also pool health care, but health insurance is treated elsewhere in this entry, so this section concentrates on pensions, unemployment, disability and work-injury protection.

The modern model is usually traced to Germany in the 1880s. Under Chancellor Otto von Bismarck, the Reichstag passed a health insurance law in 1883 that required workers to contribute to sickness funds, with employers paying one third of the contributions and workers two thirds. An accident insurance law followed in 1884, financed on a pay-as-you-go basis, with employers paying the full contributions. The old-age and invalidity law of 1889 offered benefits from age 70, a threshold later lowered to 65. Because average life expectancy was only about forty at the time, only a few pensioners benefited from it. The motives were partly political: the reforms were meant to ease industrial unrest and weaken socialism. Similar schemes spread quickly across Europe and later beyond it, and the German design became a reference point for welfare states in industrialized countries.

Two financing designs are central to the debate. In a pay-as-you-go system, current contributions finance current benefits, so each generation of workers supports the retirees of the day, and the implicit promise rests on future workers. In a fully funded system, contributions are accumulated in individual or collective accounts and invested, so benefits depend on past saving and investment returns. The first design is exposed mainly to demographic and wage trends; the second is exposed to market shocks, inflation and the cost of transition from one model to the other. Social insurance also contains a tension between actuarial fairness and redistribution. A strictly actuarial scheme would link each person's benefits closely to lifetime contributions, as a private insurer would, whereas a redistributive scheme gives proportionally more to low earners through flat components, minimum pensions or higher replacement rates. Most real systems combine elements of both, and the balance between them is a political choice as much as a technical one.

Demographic change is the main long-run pressure. Lower fertility and slower immigration reduce the growth of the workforce that pays into pay-as-you-go systems, while longer life expectancy extends the period over which pensions are paid. The United States shows how these pressures appear in practice. Unless Congress acts to shore up Social Security's finances, annual revenues will cover only about 83 percent of scheduled benefits once the combined trust funds are exhausted, which the 2026 Trustees Report places in 2034. The retirement-only fund is projected to run out earlier, in late 2032. Payroll tax revenue has not kept pace with annual expenses since 2009, so the reserves have been drawn down steadily. Depletion of a trust fund does not mean that payments stop altogether, since incoming taxes would continue to fund a large share of benefits, but it does mean that the scheduled benefits would have to be reduced unless legislation changes the rules.

Policy responses are well understood in principle, though they involve distributional trade-offs. Governments can raise contributions, adjust the retirement age, change benefit formulas (for example, indexing rules), broaden the tax base, or move partly toward funding. Each option shifts the burden between current workers, current retirees and future generations, and between higher and lower earners. Analysts generally agree that gradual adjustments made early are less disruptive than abrupt cuts made late, but reforms have historically tended to arrive only once a financial deadline is close. Public insurance schemes therefore remain sustainable or not largely because of decisions made by elected governments, not because of the arithmetic alone.

Regulation, solvency and consumer protection

Solvency regimes

Modern insurance supervision is largely built around capital adequacy. In the European Union, the Solvency II regime took effect in 2016 and rests on three layers: quantitative capital rules, governance and risk management, and public disclosure. Its core quantitative rule is the solvency capital requirement, which regulators set so that an insurer can absorb a severe loss over one year at a 99.5 percent confidence level. The insurer may compute this figure with a prescribed standard formula or with an internal model that the supervisor has approved. A lower minimum capital requirement acts as a floor, and a breach of that floor leads to immediate supervisory intervention.

Solvency II is a risk-based regime, so the capital charge depends on underwriting, market, credit, operational and counterparty exposures rather than on a fixed proportion of assets. Its Pillar 2 requires firms to run their own risk and solvency assessments, which supervisors then review.

In the United States, state regulators use a risk-based capital (RBC) formula developed by the National Association of Insurance Commissioners. It was introduced in the 1990s partly because older fixed-percentage capital rules ignored differences in the riskiness of assets and business lines. The RBC ratio compares an insurer's adjusted capital with a formula-derived requirement. Ratios falling to 200, 150, 100 and 70 percent of the authorized control level trigger escalating responses, from a corrective plan to direct regulatory control. Because these are threshold rather than target levels, comparing the ratios of different firms is of limited value, and weak insurers can sometimes pass the test for years.

Accounting for insurance contracts

Accounting rules shape how capital and profit appear to investors. Under IFRS 17, which took effect for annual periods beginning on 1 January 2023 and replaced the interim IFRS 4, insurers measure contract liabilities at current values. The liability combines the present value of probability-weighted future cash flows, an explicit risk adjustment for non-financial risk, and a contractual service margin that represents unearned profit. That margin is released into income as coverage is provided, which makes it harder for insurers to recognize profit at the point of sale. Under US GAAP, a separate long-duration reform applies to life and annuity contracts. Because liability discount rates now move with market interest rates, reported equity and earnings can be considerably more volatile than under the earlier practices.

Market conduct and policyholder protection

Beyond solvency, conduct regulation governs how insurers deal with customers. Typical rules cover disclosure of policy terms and costs, the suitability of sales, fair treatment in claims handling, and, in many US states and some other countries, prior approval of premium rates. These rules respond to the information asymmetry between sellers and buyers, in which policyholders may not understand complex exclusions until a claim is refused. When an insurer fails, guaranty or policyholder protection schemes pay covered claims up to statutory limits, funded by levies on the surviving companies. Those schemes reduce the damage to households but also create a moral hazard, since they can weaken the incentive for policyholders to scrutinize insurers' finances.

Supervisory coordination and systemic risk

Insurers operate across borders, so supervision is coordinated through bodies such as the International Association of Insurance Supervisors (IAIS) and the Financial Stability Board. Since 2013, the Financial Stability Board designated some large insurers as global systemically important insurers, and the IAIS later replaced that list-based approach with a holistic framework that assesses activities and group-wide risk. In the United States, the Financial Stability Oversight Council designated AIG and Prudential Financial as systemically important in 2013, and a federal court later overturned the designation of MetLife.

The central debate is whether insurers, unlike banks, pose systemic risk. Supporters of the traditional view argue that insurers mainly pool long-term risks and rarely face sudden runs. Critics point to non-traditional activities, such as large derivatives positions, short-term funding, and the sale of products whose surrender rights can prompt sudden withdrawals, as channels through which distress could spread to the wider financial market. This question remains contested, and it is a key reason regulators continue to refine capital and group supervision standards.

Macroeconomic and financial role

Insurers are among the largest institutional investors, and this follows from how their business works. Policyholders pay premiums up front, while claims and benefits are paid later. The gap between premiums received and claims paid is known as float, and insurers invest it until it is needed. European insurers held total assets of €9.5 trillion, according to EIOPA data for 2024. Their holdings are concentrated in government bonds, covered bonds, corporate bonds, equities and infrastructure, real estate or private equity. The maturity of the portfolio depends on the liabilities behind it. A motor insurer writing one-year contracts mainly buys short-dated bonds, while a life insurer or annuity writer buys longer assets to match obligations that can run for decades (up to 15 years for life insurance liabilities). Because the model is long-term, insurers can hold illiquid assets and avoid selling into market downturns, a point that matters for financial markets and liquidity.

Insurance also supports lending, which is the second channel through which it affects the real economy. Mortgage insurance is the clearest case: it transfers mortgage risk, and notably tail risk, from lenders to insurers. The borrower pays the premium, but the lender is the beneficiary, and the amount covered is usually capped as a share of the unpaid principal. In some markets lenders go further and make a policy a condition of loan approval. In India, for example, lenders often require borrowers to buy insurance before a loan is sanctioned, although this is not a general legal requirement. By reducing the losses that a lender can suffer from default, such cover allows banks to extend credit they might otherwise ration, which in turn supports household purchases and business investment.

The link is not free of problems. Regulators have noted that mortgage insurers can act as a second check on lending standards, but that cover can also weaken a lender's incentive to screen borrowers, which is a form of moral-hazard. Because default risk becomes concentrated in a small number of insurers, supervisors must ensure that the arrangement does not reduce the total capital held in the financial system. Credit-related insurance is generally bought voluntarily, but it still depends on sound underwriting by both sides.

Two ratios are standard tools for measuring how deeply insurance is embedded in an economy. Penetration is premium income relative to GDP, $\text{penetration} = \dfrac{\text{premiums}}{\text{GDP}}$, while density is premium per head of population, $\text{density} = \dfrac{\text{premiums}}{\text{population}}$ (see insurance penetration). The two measures move together with income but are not identical. The OECD reported an average penetration of 6.2% on average among OECD countries in 2024, up from 6% in 2023, and 5.4% on average among all reporting jurisdictions, and the level remained below that of a decade earlier. Other compilations put the global figure higher, at about 7.3% for 2024, which shows how sensitive the averages are to sample and definitions. Density varies even more than penetration. Average per capita spending on insurance in advanced markets was USD 4664 in 2019, and penetration in those markets was 9.6%. Country-level spreads are wide: it was 1.1% in Romania in 2024 and 33% in Luxembourg. Penetration is also shaped by the product mix. Because life insurance is closely tied to savings and pension arrangements, its level depends heavily on the design of those systems and on interest rates, and this trend is largely driven by life insurance penetration as non-life penetration has remained stable over the past decade.

Insurers also act as shock absorbers during downturns, though their capacity to do so is limited. Losses from a single event are spread over time, since liability claims and life benefits are paid across many years and reserves are built to meet them. They are also spread across space: reinsurance and international diversification allow a regional disaster or a local recession to be absorbed by capital held in many countries. Premium volumes do fall when economic activity contracts, because new business slows and some lines, such as motor cover, shrink with reduced driving. Insurers' own investment portfolios are exposed to falls in asset prices, and in 2008 some large insurers required public support, showing that they can also transmit financial stress. These issues are treated under systemic-risk and in the regulation section.

Insurance in developing economies and microinsurance

Why penetration stays low

Insurance remains thin in many low- and middle-income economies. Across Africa, premiums are roughly 2.7% of GDP, about half the global level of around 7%. Across the continent, insurance penetration remains at just 2.7% of GDP, which is less than half the global average of approximately 7%. Uganda is lower still, at below 1% of Gross Domestic Product (GDP), among the lowest in the region. In Nigeria, the largest market by population, the share of people with any cover is smaller still. Several obstacles recur in the literature: high premiums relative to irregular incomes, fragmented distribution, and weak trust in insurers. Financial literacy is a further constraint. In Mali, only 8% of adults understand basic insurance concepts despite high levels of mobile phone access. Cost, trust, literacy and the enforceability of contracts and claims through courts interact, so that no single fix tends to close the gap.

Microinsurance and index-based cover

Microinsurance refers to low-premium policies designed for low-income households, often in the informal economy. Its best-known design is index-based insurance, which pays out when a measurable variable such as rainfall or vegetation greenness crosses a threshold, instead of after an adjuster verifies an individual loss. Basing payouts on an index instead of verified losses largely overcomes the problems of moral hazard, adverse selection, high transaction costs and payout delays that made traditional loss-based crop insurance infeasible for smallholder farmers. The trade-off is basis risk: a farmer can suffer a loss that the index fails to register.

Programmes have reached considerable numbers. In Ethiopia and Senegal, the R4 Rural Resilience Initiative has scaled unsubsidized index insurance to over 20,000 poor smallholder farmers, and in Mongolia a livestock programme insures more than 15,000 nomadic herders. Rigorous evaluations of index-based livestock insurance in northern Kenya and southern Ethiopia have found welfare and productivity gains. Results elsewhere are weaker. Demand for livestock products has consistently remained low despite donor enthusiasm. A randomized trial of index insurance for cotton farmers in Burkina Faso found no impact of insurance on cotton, though it did detect spillovers into other investment. A 3ie evaluation of livestock insurance found no overall gains in investment or income, and it attributed much of this to limited awareness: many insured producers, particularly those with mandatory coverage, did not know they were insured or misunderstood what the product covered.

Community-based and subsidized health insurance

In Africa and Asia, many households rely on community-based health insurance (CBHI), in which members pool the risk of illness through a local or occupational scheme, often with subsidies from government or donors. The evidence on effectiveness is mixed. A systematic review found that compared to uninsured households, CBHI-insured households had improved utilization of healthcare services but only marginally improved financial protection while accessing care. The main reason is benefit design: most CBHIs only cover inpatient services, and so offer limited, if any, financial protection for outpatient care. They have however been shown to offer significant financial protection for inpatient care. Studies in Ethiopia reported that CBHI membership increased the utilization of health services and reduced the incidence of catastrophic health expenditure. Evidence from rural Burkina Faso suggested that CBHI protected household assets during 2004-2007.

The scale of the problem explains the interest. The World Health Organization's estimates indicate that 11 million people in Africa and 550 million people in Asia face catastrophic health expenditures annually. Out-of-pocket payments remain high in many systems: in Ethiopia they were about 31% of total health expenditure in 2017, against a recommended range of 15–20%. Equity is another concern, since research from Asia and sub-Saharan Africa suggests that community based health insurance has been less effective in securing equity than expected. Subsidies and co-payment rules therefore determine who actually benefits, which connects this topic to the adverse selection and moral hazard problems discussed elsewhere in this entry.

Mobile-phone distribution and embedded insurance

Mobile networks offer a cheaper route to customers than branch-based sales. Insurers partner with mobile network operators (MNOs) to use their technology platforms and agent networks to provide insurance to subscribers, most of whom have no cover. Scale has been demonstrated in several markets. Airtel Money and MicroEnsure launched affordable health insurance products that reached over 2 million people by 2018 in Ghana and Nigeria, and Kenya's M-Tiba platform enabled over 4.5 million individuals to access mobile-enabled health insurance. Embedded insurance goes further by attaching cover to a purchase that people already make, such as agricultural inputs or a loan. One Kenyan agritech example bundles a crop-insurance premium into an input purchase, and the premium is reported to be bundled into the transaction cost (typically between $5–10).

Reach does not guarantee take-up. In Uganda, observers note that the expected explosion in uptake has not occurred. "Access does not equal adoption." Impact evidence is also thin: current reporting often emphasizes products launched and premiums collected rather than active users, renewals and claims outcomes. Whether low-cost digital channels produce durable protection for low-income households therefore remains an open empirical question.

Behavioral and demand-side perspectives

Standard models predict that a risk-averse person should buy full cover whenever the premium is actuarially fair. Actual purchasing departs from that benchmark in systematic ways. Recent work treats risk, ambiguity and loss aversion as preference traits that help explain why observed insurance choices diverge from theory, and it uses survey, lab and field data to test those explanations. Behavioral economists use survey, lab, and experimental field data to study choices that do not match a perfectly rational and informed agent, and they treat decisions as rational within a wider frame that includes loss aversion, ambiguity, framing, reference points and emotions. The field's foundation is prospect theory, which places a reference point at the center of evaluation, so a premium paid every year for protection that may never pay out is felt as a recurring loss within behavioral economics. Prospect theory posits that individuals are more sensitive to losses than to equivalent gains relative to a reference point.

Probability weighting adds a second distortion, since people respond to small probabilities in ways that do not track their true size. A 2025 conference paper argues that this combination produces spikes in demand after salient disasters, alongside under-purchase of risks that are hard to perceive; it reports that only a minority of small businesses buy cyber cover despite a substantial attack probability. The paper links loss aversion to surges in insurance demand and to underinsurance of imperceptible risks, with cyber insurance as its example. Present bias affects long-duration products in particular. Because life cover pays out in a distant future, it is easy to postpone. A field experiment with a large life insurer and a pre-registered replication found that loss aversion and temporal orientation frames influence whether individuals aged 25–49 shop for life insurance. Framing a purchase as protection for the present is one way insurers have tried to counter the present bias that delays the decision.

Framing and incentives also shape take-up in measurable ways. In a Michigan field experiment, loss-framed incentives increased take-up by 4.7 percentage points, or 17 percent relative to the gain-framed redemption rate of 27 percent. The authors, however, argue that the mechanism is not mainly loss aversion. Loss framing promotes take-up by raising the perceived probability that the incentive will be provided as promised. The effect was concentrated among people who did not trust the partner organization at baseline. Take-up is higher for the loss frame than the gain frame incentive, but only among those who do not trust the partner organization at baseline. Defaults have comparably large effects. Opt-out enrollment in pensions and organ-donation systems, and automatic enrollment in workplace plans, consistently raise participation far above opt-in arrangements. Compulsory participation, as in statutory motor liability or public social insurance, removes the choice altogether, which is why researchers view it as a policy tool rather than a framing effect.

Experiments in low-income settings show how far take-up can fall when a product is unfamiliar. Rainfall index insurance pays out according to a publicly observable index like rainfall, temperature or a commodity price rather than the farmer's actual loss, which makes the contract harder to grasp. In Gujarat, take-up of index-based weather insurance was surprisingly low, despite its theoretical benefits. A randomized trial there found that financial education increased take-up from 8 to 16 percent, and a money-back guarantee, comparable to a price reduction of about 60 percent, increases demand by 7 percentage points. Another study found that payouts improve trust in the product, which suggests that experience shapes beliefs about whether a contract will perform. Mismatch between a buyer's expectations and a policy's actual triggers, exclusions or information gaps also matters. Basis risk, the gap between the index payout and a household's real loss, is one such gap, and it is a significant impediment to the take-up of the index insurance product. When buyers misread these terms, they may hold cover that pays out differently from what they expected, or they may drop cover they need. Regulators respond with disclosure rules, but the evidence suggests that understanding remains the binding constraint for complex products.

Cultural and religious perspectives also shape demand. Within Islamic finance, conventional insurance is contested because of three prohibited elements, riba (interest), gharar (excessive risk), and maysir (speculative nature). Takaful was developed as an alternative. Contributions to the risk pool are typically treated as tabarru’ (voluntary contributions) intended to assist co-participants, and losses are shared among members rather than transferred to a commercial insurer. The operator also plays a different role. In one account, the conventional insurer is the owner, while the takaful operator acts as an agent (wakala) on behalf of the participants. Investments are also restricted: riba is avoided by restricting takaful fund investments exclusively to Shariah-compliant assets. The permissibility of conventional insurance is disputed. One guide reports that the majority scholarly position considers conventional insurance impermissible, while other scholars accept conventional cover in cases of necessity, and critics question whether the tabarru structure truly removes uncertainty. These debates affect demand in Muslim-majority markets, where the choice between Islamic cooperative cover and conventional policies is often a religious as well as financial decision, as with religious attitudes more broadly.

Catastrophe risk, climate change and insurability

Catastrophe risk and its modeling

Catastrophe risk breaks the logic of pooling, because one event can strike many policyholders at once and their losses are strongly correlated. Before 1992, the industry's loss assessments rested almost entirely on claims history. Hurricane Andrew then showed that estimates built strictly from past losses can fall short of what a future storm will cost. The 1992 Andrew and 1994 Northridge earthquake led actuaries to rely on probabilistic simulation instead. Both events led actuaries to recognize that probabilistic computer simulation models would help estimate probable maximum losses for these severe events.

A modern catastrophe model is a computer program that simulates thousands of plausible catastrophic event scenarios. Scenario parameters draw on meteorological history, geology and geography. Policy terms such as deductibles and limits are then applied to each simulated loss, and the results are combined into a loss probability distribution. From that distribution, the probable maximum loss (PML) is read off at a chosen return period: PML is a point on the exceedance probability curve at a chosen return period, and a 1% annual exceedance probability is the 1-in-100 year loss. Insurers use these figures to test their exposure and to decide how much reinsurance to buy. Catastrophe models are not the same as climate models: they are calibrated largely on observed hazard records, which is one reason climate change poses a modeling problem.

Recent loss experience

Swiss Re's 2025 figures show how loss patterns are shifting. Wildfires, severe convective storms and floods together accounted for 92% of the USD 107 billion in global insured natural catastrophe losses, a record proportion. These are the so-called secondary perils, and the long-term trend of global insured natural catastrophe losses rising on average by 5–7% annually in real terms is intact. The January 2025 Los Angeles fires alone generated record-breaking combined insured losses of around USD 40 billion. The total still fell short of trend, largely because of the absence of a major US hurricane landfall.

Growth in insured losses is mostly explained by exposure. Exposure growth explains more than 80% of the upward trend in global weather-related insured losses since 1970. Swiss Re's chief economist attributes much of the long-term rise to valuable property being built in dangerous locations, combined with higher rebuilding costs. For North American wildfires and European severe convective storms, insured losses are rising roughly twice as fast as exposure alone would predict, which the institute links to hazard intensification and rising vulnerability.

The protection gap

More broadly, the protection gap describes the discrepancy between the potential financial loss for people or economies and the amount of insurance coverage to mitigate those losses. Estimates depend on method. Natural catastrophes caused an estimated USD 112 billion in global economic losses in the first half of 2026, of which only USD 44 billion was insured, according to Munich Re. Aon, by contrast, reported that the protection gap reached a record low of 38% globally for the first half of 2025. Differences in how events are classified and how losses are collected explain why firms publish different totals for the same period. Aon's and Swiss Re's figures differ because of different collection methods and criteria for classifying events.

The gap is widest in poorer economies, but it persists in rich ones too. In the US, the protection gap for the ten years from 2015 to 2024 was USD 571 billion, 43% of total economic losses from natural catastrophes. Causes include low awareness, unaffordable premiums and a belief that public aid will cover losses. Widening gaps could raise financial-stability risks by shifting more risk onto the banking sector, a concern that connects insurance to systemic risk. To narrow the gap, governments have created backstops: national flood programmes, and state-backed catastrophe pools such as Florida's Hurricane Catastrophe Fund, established after Andrew. Supervisors are also encouraged to enable innovations such as parametric insurance.

Climate change, pricing and availability

Climate change affects insurance mainly through hazard frequency and severity. Several categories of extreme weather have been becoming more frequent and intense. Because historical records describe a climate that is changing, insurers increasingly adjust hazard inputs to forward-looking conditions. That tends to raise expected losses and premiums in exposed areas. The effect is not confined to property: one industry estimate suggests that each additional 10 days above 35°C cuts annual labour productivity by about 0.3% across developed economies.

Availability is the more difficult problem. Rising catastrophe losses make it harder for insurers to provide affordable cover in high-risk areas, prompting them to reassess their risk-management strategies. Insurers may restrict or withdraw coverage in the most exposed zones, shifting costs to households and governments. Risk-based premiums can encourage adaptation such as elevation, stronger building codes and defensible space around homes, but they can also price out lower-income households. This tension, between market-based pricing and public affordability, is why regulators and governments remain central to the debate over insurability.

Technology and digital transformation

Telematics, connected devices and remote sensing

Motor insurance has been the line most changed by data collection. Telematics, meaning the gathering and analysis of trip-level data from phones, plug-in devices or connected cars, lets insurers price policies according to how much, how safely and when people drive. The first generation, pay-as-you-drive, charged by distance. Later pay-how-you-drive and manage-how-you-drive models added measures such as speed, acceleration and braking, and some added feedback intended to change behaviour. The approach is also meant to reduce adverse selection, since it gives insurers information that customers could not easily conceal. Its benefits are uneven. Location data can raise premiums for careful drivers who live in high-claim areas, a point noted in industry white papers, and in some jurisdictions regulation prevents insurers from surcharging drivers on the basis of such data.

Wearable devices play a similar role in health and life products, though mostly through rewards rather than direct price changes. Insurers offer discounts or premium credits linked to recorded activity, such as step counts or exercise. This turns the data into a behavioural incentive that can work against moral hazard, but it also raises the question of whether people with less capacity for exercise are effectively penalised. In property underwriting, aerial and satellite imagery allows insurers to assess roof condition, vegetation near buildings and flood exposure without sending an inspector, which speeds up quoting and renewal.

Machine learning in underwriting and claims

Machine learning is now used across the insurance cycle. Models trained on large datasets can rank risks more finely than traditional rating tables, flag claims that deserve a closer look, route simple claims to automated settlement, and detect patterns linked to organised fraud. These gains depend on data quality and on model validation, and a model can perform well on average while still producing errors for particular groups of policyholders.

The central concern is fairness. Complex models may use variables that act as proxies for protected characteristics, and their logic can be hard to explain to a customer or a regulator. In the United States, the National Association of Insurance Commissioners adopted a model bulletin in December 2023 that expects insurers to maintain a documented AI governance programme, test for unfair discrimination, audit their systems and oversee third-party vendors. The bulletin is not a statute but is adopted state by state. Sources differ on the exact count: one reports that it is in effect in 24 states and the District of Columbia as of the NAIC's spring 2026 meeting, while another puts the count at 25 states plus the District of Columbia as of August 31, 2026. In the European Union, the AI Act classifies AI systems used for risk assessment and pricing in life and health insurance as high-risk, which brings obligations on documentation, data quality and human oversight.

Parametric insurance and blockchain-based claims

Parametric insurance is the clearest case of technology changing the claims model itself. A policy pays out when an event of an agreed severity occurs, rather than on the damage actually sustained. Sensors, weather stations and satellite data verify the trigger, and the payment follows quickly. This model has been used mainly for catastrophe and agricultural risks in developing economies, for example through the Caribbean Catastrophe Risk Insurance Facility.

Its main weakness is basis risk: the index may not match a policyholder's real loss. A drought contract can pay out at one rainfall threshold while a farmer just above it still suffers serious losses, a gap that may leave economic impacts uncovered. Blockchain-based experiments aim to automate payment further, using smart contracts that execute when a verified external data feed, often called an oracle, reports that a trigger has been met. Pilots have focused on simple products such as flight-delay cover and crop insurance. They have shown that automatic settlement is technically feasible, but they have not yet replaced conventional claims systems at scale.

Data privacy, transparency and digital distribution

Telematics, wearable and health data are highly sensitive, so consent, purpose limitation and data minimisation are central to regulation. The EU's General Data Protection Regulation restricts decisions based solely on automated processing that have significant effects, and it gives individuals rights to information and human review. Regulators also expect insurers to explain why a premium changed or a claim was declined, which is difficult when the model is complex. Digital distribution, including comparison websites, app-based sales and insurance embedded in purchases such as cars or flights, widens access and lowers acquisition costs, but it also makes disclosure and sales practices a matter for conduct regulators, not only prudential ones.

Open questions and contested issues

The limits of insurability show most clearly where losses are hard to model, hard to bound, or hard to attribute. Pandemics, cyber attacks and armed conflict share a feature that strains the pooling logic: one event can hit many policyholders at once, so losses are correlated rather than independent, and historical data give a weak basis for estimating how often such events occur. The dispute is usually less about whether these risks exist than about where the edge of cover should sit, and that edge is typically drawn through exclusion wording and the courts' reading of it.

Cyber insurance offers a clear case. Concern about state-sponsored intrusions led Lloyd's of London to require, from 31 March 2023, that standalone cyber policies exclude losses from state-backed attacks, in addition to ordinary war exclusions. The clause targets attacks that significantly impair a state's ability to function or its security capabilities. Whether a given incident falls inside the exclusion depends on attribution, which the parties judge by reference to the reasonably available evidence, and which may rest on an official government statement. Lloyd's framed the concern as one of systemic risk: malicious code can spread across interconnected systems, so losses may not stay within a single insured or a single sector. Some market participants view the exclusion as a necessary limit on exposure that the market cannot otherwise price; others see it as leaving buyers uncertain about coverage precisely when a large incident occurs. The clause also raises questions for cyber security policy, because it signals which losses the private market will and will not absorb.

Pandemic losses produced a parallel debate. During COVID-19, many business interruption claims depended on whether a closure order amounted to "direct physical loss" of property. Early American rulings often favoured insurers, especially where policies contained explicit virus exclusions. Later decisions diverged. The Pennsylvania Supreme Court held that a physical alteration of property was a threshold requirement, so government-ordered closures did not trigger cover. The North Carolina Supreme Court, by contrast, ruled for policyholders in a policy without a virus exclusion, reasoning that a loss of the physical use for which property was insured could count. The episode illustrates a general problem: the wording of an exclusion, and the interpretive rules courts apply to ambiguous terms, can decide outcomes that the pricing model never anticipated. It also raises the broader question of whether a risk as widespread as an infectious disease outbreak can be insured privately at all, or whether it is better handled through public programmes, a point that links to public health policy.

Climate change poses the most immediate version of these questions. Insurers face rising and geographically concentrated losses from wildfire, hurricane and flood. In 2024, State Farm chose not to renew roughly 72,000 home policies in California, partly because of wildfire risk. The state's insurer of last resort, the FAIR Plan, grew from about 271,000 policies in 2022 to more than 684,000 by March 2026. Swiss Re reported that 57% of global natural catastrophe losses in 2024 were uninsured. Disputes centre on whether regulators should allow prices to reflect forward-looking risk, which would make cover expensive in exposed areas, or should limit rate increases to protect affordability, which can drive insurers out of those areas. Public backstops such as the US National Flood Insurance Program and state residual-market plans shift risk onto policyholders collectively or onto taxpayers, and critics argue that subsidised cover in hazardous locations weakens the incentive to avoid building there, a form of moral hazard at the level of places rather than individuals. Governments acting as insurers of last resort thus face a choice between accepting open-ended contingent liabilities and allowing parts of the housing market to become effectively uninsurable.

A second tension concerns pricing principles. Actuarially fair pricing charges each person according to their expected cost, which is efficient for risk signals but can exclude people who cannot pay or who live in high-risk places. Solidarity-based models, such as community rating in some health and social systems, set premiums across a whole population regardless of individual risk, which spreads costs and protects the vulnerable but weakens the link between behaviour and price. Community rating also invites adverse selection unless it is paired with compulsory participation or other measures. Whether these models should be preferred is a question of values as much as of mathematics, and it is contested in health insurance, pensions and disaster cover alike.

Finally, the long-term sustainability of public health and pension insurance is a fiscal as well as an actuarial question. Ageing populations and rising life expectancy lengthen the period over which benefits must be paid, while the ratio of contributors to beneficiaries falls. Governments can respond by raising contributions, adjusting retirement ages or benefits, drawing more on general taxation, or building funds, and each option redistributes costs across generations and income groups. These choices bear directly on universal health coverage and on whether social insurance can keep its promises under fiscal policy constraints. Across these issues, the unresolved question is less technical than political: who should bear the risks that private markets decline to price, and on what terms.

References

1. The history of insurancewesternfinancialgroup.ca
2. Insurancebritannica.com
3. Lloyd%27s of Londonen.wikipedia.org
4. History of Insurancestudy.com
5. The Great Fire of London and the birth of the insurance industry - Insurance Postpostonline.co.uk
6. The Birth of Property Insurance in Englandalwaysausten.com
7. The Equitable Life Assurance Societyen.wikipedia.org
8. A hurried history of life insurance!activequote.com
9. Life Insurance in the United States through World War Ieh.net
10. Equilibrium in competitive insurance markets with ex ante adverse selection and ex post moral hazard - ScienceDirectsciencedirect.com
11. Competitive insurance markets under adverse selection and capacity constraints - ScienceDirectsciencedirect.com
12. When the Sh*t hits the Ban - UK Human Rights Blogukhumanrightsblog.com
13. Insurance industry agrees to genetic test moratoriumprogress.org.uk
14. Genetics and insurance: Should the moratorium continue?progress.org.uk
15. Equilibrium in a Competitive Insurance Market Under ...nber.org
16. Adverse selection and equilibrium in liability insurance marketslink.springer.com
17. core.ac.ukcore.ac.uk
18. Lecture 19: Adverse selection in insurance marketshealy.econ.ohio-state.edu
19. Lecture 18 - Information, Adverse Selection, and Insurance Marketsdspace.mit.edu
20. UK insurance moratorium on use of genetic data extended until 2014progress.org.uk
21. Test-Achatspinsentmasons.com
22. Test-Achats and sex-based actuarial factors: UK insurance and occupational pensions implications, Equality Act 2010 amendments, new contract definition, and post-Brexit outlook - Legal Guidance - LexisNexis UKlexisnexis.com
23. Pure Premium Method Calculatormetricgate.com
24. Rate Making: How Insurance Premiums Are Setthismatter.com
25. Benchmark premium determination system and methodimage-ppubs.uspto.gov
26. Claims reserving and financial impactfiveable.me
27. Incurred but not reporteden.wikipedia.org
28. Principle of equal treatment of men and women outside the employment marketeur-lex.europa.eu
29. UK Government Publishes Response to Test-Achats Ruling on Gender Neutrality in Insuranceinsurereinsure.com
30. Introduction: Incurred But Not Reported (IBNR Reserves) • The Actuarial Clubtheactuarialclub.com
31. sigma 5/2024: Global economic and insurance market outlook 2025-26swissre.com
32. No 2/2025 World insurance in 2025: a riskier, more fragmented world orderlegismap.com.br
33. Life insurance drives global premium growth as interest rates remain higher for longer, says Swiss Re Instituteswissre.com
34. LIMRA: U.S. Retail Annuity Sales Top $460 Billion in 2025, Marking Fourth Year of Record Saleslimra.com
35. Annuity Sales by Product Typeicfs.com
36. What Is a Fixed Index Annuity?annuityexpertadvice.com
37. Equity-indexed annuityen.wikipedia.org
38. Why Fixed Index Annuities Are More Attractive Than Ever - Annuity.organnuity.org
39. What is Managed Care?cigna.com
40. HMO vs PPO: Whatʼs the Difference Between HMO and PPO Plans?humana.com
41. The Health Insurance Experiment: A Classic RAND Study Speaks to the Current Health Care Reform Debaterand.org
42. 40 Years of the RAND Health Insurance Experimentrand.org
43. A little bit more about the RAND health insurance experimenttheincidentaleconomist.com
44. RAND Health Insurance Experimenten.wikipedia.org
45. Quantifying Changes to Healthcare Utilization After a Reduction in Cost-sharing Among Deductible Plan Enrolleesarxiv.org
46. Equity in healthcare financing: A review of evidence - ScienceDirectsciencedirect.com
47. How Healthcare is Financed: Revenue Sources Explainedeconomicsbydesign.com
48. Is tax funding of health care more likely to be regressive than systems based on social insurance in low- and middle-income countries?scielo.org.co
49. Financingeconomicsbydesign.com
50. Effects of Cost Sharing on Use of Medical Services and Healthrand.org
51. Health Insurance and the Demand for Medical Care: Evidence from a Randomized Experimentrand.org
52. The Case Against Labor-Tax-Financed Social Health Insurance For Low- And Low-Middle-Income Countrieshealthaffairs.org
53. HMO vs PPO vs POS vs EPO: What's the difference?healthinsurance.org
54. Managed Care Planstaylorbenefitsinsurance.com
55. Actual Cash Value vs. Replacement Costusnews.com
56. Replacement Cost vs. Actual Cash Valueprogressive.com
57. Actual Cash Value vs Replacement Costallstate.com
58. Vehicle insuranceen.wikipedia.org
59. How do compulsory third party and third party car insurance differ?suncorp.com.au
60. Around the World on Four Wheels: How Auto Insurance Works Across the Globethezebra.com
61. Motor Vehicle Insurance (India)en.wikipedia.org
62. Around the world look at CTP - CTP in nine countries - greenslips.com.augreenslips.com.au
63. The Impact of Nuclear Verdicts on Casualty Claimsclaimsjournal.com
64. Nuclear Verdicts Driving Up P&C Liabilitysimplesolve.com
65. Inside the verdict: what is driving the rise in nuclear and thermonuclear awards?sedgwick.com
66. Nuclear Verdicts: An Update on Trends, Causes, and Solutions - ILRinstituteforlegalreform.com
67. The Reinsurance Contractreinsurance.org
68. Excess Insurance Policies and Reinsurance Treatieslalitigationlawyers.com
69. Introduction to Resinsurance ©London School of Insurance 2016 INTRODUCTION TOelearning.londonschoolofinsurance.com
70. The 8 Types of Reinsuranceaccelerant.ai
71. Catastrophe bond issuance reaches record $17.6 billion in first half By Investing.comm.za.investing.com
72. Catastrophe bond market records that were broken in H1 2026 - Artemis.bmartemis.bm
73. Global and US property cat rates down 16%, APAC 19% after July renewals in 2026: Guy Carpenter - Artemis.bmartemis.bm
74. 2026 renewal sees sharpest decline in risk-adjusted global property rates since 2014: Howden - Reinsurance Newsreinsurancene.ws
75. Guy Carpenter U.S. Property Catastrophe Rate-On-Line Indexartemis.bm
76. Catastrophe reinsurance prices post historic annual fallinsurancebusinessmag.com
77. Introduction to Lloyd's of Londonbuckhill.co.uk
78. Capital structure - Lloyd'slloyds.com
79. 5.0 How Lloyd’s Worksaepronet.org
80. What Is Lloyd's of London? A Complete Guide for 2026genasystech.com
81. Alpha Insurance Analystsaianalysts.com
82. graphsearch.epfl.chgraphsearch.epfl.ch
83. Captive insuranceen.wikipedia.org
84. What Is Captive Insurance?captive.com
85. What Are Captive Insurance Companies And How Do They Operate?agentsync.io
86. Captives & Risk Retention Groupsinsurance.utah.gov
87. What Is a Mutual Insurance Company?3blmedia.com
88. What is a stock insurance company? Definition & Meaningachievable.me
89. What Is a Stock Insurance Company?harbor-ins.com
90. What Is a Mutual Life Insurance Company?nerdwallet.com
91. Stock vs. mutual insurers: Who should and who does charge more? - ScienceDirectsciencedirect.com
92. Bismarck's social legislation 1881 to 1890bismarck-biografie.de
93. Bismarck's Social Welfare Legislationfiveable.me
94. Social Security Report Projects Trust Fund Shortfall in 2034aarp.org
95. 2026 Social Security Trustees Report, Explainedbipartisanpolicy.org
96. The Earliest History of the Evolution of Social States in the ...sociostudies.org
97. The Standard Formula: Encyclopaedia of Prudential Solvencyskadden.com
98. Solvency II: The Complete Guide to the EU Prudential Regimefinengyadvisory.fr
99. Solvency II Capital Requirements Explainedsolvencyiisolutions.com
100. Solvency II definition - Risk.netrisk.net
101. Risk-Based Capital Preamble History of Risk-Based Capital by the NAICcontent.naic.org
102. Risk Based Capital (RBC) for an Illinois Based Insurance Company May 11, 2018math.illinois.edu
103. What Is Risk-Based Capital: A Primer for Captivescaptive.com
104. RBC Risk-Based Capital Forecasting and Instructions health 2022in.gov
105. IFRS 17 Insurance Contracts Darrel Scott November 2022, Viennacfrr.worldbank.org
106. IFRS 17 Insurance Contracts Overviewscribd.com
107. HSBC Holdings plc Report on Transition to IFRS 17 ‘Insurance Contracts‘hsbc.com
108. Investing from insurers’ perspective Executive summaryinsuranceeurope.eu
109. Basel Committee on Banking Supervision Joint Forum Mortgage insurance:bis.org
110. Insurance penetration: Global Insurance Market Trends 2025oecd.org
111. No 4 /2020 extra World insurance: Regional review 2019, and outlookswissre.com
112. insurance all about floatfool.com
113. Float: How Insurance Companies Can Leverage Buffett's Secret to Wealtheinvestingforbeginners.com
114. Insure to Ensure Your Loan?vinodkothari.com
115. Indian Insurance Market - Policy Holder - IRDAIpolicyholder.gov.in
116. India's Insurance Gap Widens: Penetration Stagnant at Half Global Average, Density Trails Far Behind!whalesbook.com
117. How A Smartphone Got Ghana's Everyday Earners To (Finally) Trust Insuranceweetracker.com
118. Why technology alone cannot fix Uganda’s low insurance penetration - Africa Aheadafahpublishing.com
119. The Evolution of Mobile Money and Insurance Partnerships in West Africawaica.net
120. Livestock Insurance - an overviewsciencedirect.com
121. Scaling up index insurance for smallholder farmers: Recent evidence and insightsccafs.cgiar.org
122. The Spillover Impact of Index Insurance on Agricultural ...nber.org
123. Agricultural Insurance Evidence Program3ieimpact.org
124. Impact of community-based health insurance in low- and middle-income countries: A systematic review and meta-analysisjournals.plos.org
125. Providing Financial Protection and Funding Health Service ...resyst.lshtm.ac.uk
126. The impact of community-based health insurance on health service utilization and financial risk protection in Ethiopia - PMCpmc.ncbi.nlm.nih.gov
127. (PDF) Does Community‐Based Health Insurance Protect Household Assets? Evidence from Rural Africaresearchgate.net
128. The path to universal health coverage in five African and Asian countries: examining the association between insurance status and health-care use - The Lancet Global Healththelancet.com
129. Community based health insurance in developing countries - PMCpmc.ncbi.nlm.nih.gov
130. MOBILE PHONES AND MICROINSURANCEilo.org
131. The Insurance Penetration Paradox: Why Coverage Matters More Than Premiumsafricanpact.org
132. Digital opportunities for the distribution of index‐based microinsurance: Evidence from a discrete choice experiment in Mali - Kirchner - 2024 - Journal of Agricultural Economics - Wiley Online Libraryonlinelibrary.wiley.com
133. The Environmental Impacts of Microfinance: Index-Based Livestock Insurancebarrett.dyson.cornell.edu
134. AI, mobile and cyber risks: The new horizons for African insurance - The African Financial Summit Platformafis.africa
135. Embedded Insurance for Fintechs and Banks: How to Ship Coverage That Actually Works at Scalecuracel.co
136. How the demand for insurance became behavioral - ScienceDirectsciencedirect.com
137. Loss aversion or lack of trust: Why does loss framing work to encourage preventive health behaviors? - ScienceDirectsciencedirect.com
138. Highlights in Business, Economics and Management WTED 2025 Volume 58 (2025)hbem.org
139. Life insurance, loss aversion, and temporal orientation: a field experiment and replication with young adultslink.springer.com
140. Journal of Behavioral and Experimental Economics 104 (2023) 102022deanyang-econ.github.io
141. How Does Risk Management Influence Production ...newyorkfed.org
142. Marketing Weather-Indexed Agricultural Insurance to ...theigc.org
143. Evidence from Rainfall Insurance in India Sarthak Gaurav, ...povertyactionlab.org
144. Copyright © 2013, 2014 by Shawn Cole, Xavier Giné, and James Vickeryhbs.edu
145. Selling Formal Insurance to the Informally Insured Principal Investigatorsbasis.ucdavis.edu
146. "Takaful: The Legal Architecture of Islamic Insurance" by Jo Chitlik - Canopy Forumcanopyforum.org
147. Introduction to Takaful: Islamic Insurance - Islamonweben.islamonweb.net
148. What is Takaful? — Complete Guide to Islamic Insuranceislamicfinancecalculator.com
149. Offering rainfall insurance to informal insurance groups: Evidence from a field experiment in Ethiopia - ScienceDirectsciencedirect.com
150. Takaful (Islamic Insurance)islamicbankers.center
151. NAIC Catastrophe Modeling Primer March 2025content.naic.org
152. NAIC Catastrophe Modeling Primer December 2024 Exposure Draftcontent.naic.org
153. Insurance Topicscontent.naic.org
154. What is Catastrophe Modeling? Cat Models and AALwherobots.com
155. Wildfires, storms, floods contribute to record 92% of global insured losses in 2025, says Swiss Re Instituteswissre.com
156. 2025: A year of hail storms, fire and high waterswissre.com
157. Wildfires, storms, floods contribute to record 92% of global insured losses in 2025, says Swiss Re Institutepreventionweb.net
158. Swiss Re Says Wildfires, Storms, Floods Account for 92% of Global Insured Lossesclaimsjournal.com
159. Natural Catastrophe Insured Losses Hit $107 Billion in 2025 as Wildfire and Storm Risks Accelerate - Risk & Insurance : Risk & Insuranceriskandinsurance.com
160. The climate adaptation imperative: tackling the protection gap for climate insuranceschroders.com
161. Infographic: Insured and Economic Losses from Natural Catastrophes (H1 2026)beinsure.com
162. Climate events have cost $162b in 2025. Insurance covered mostweforum.org
163. Facts + Statistics: Global catastrophes - Triple-I®iii.org
164. IAIS report examines financial stability implications from natural catastrophe insurance protection gaps - International Association of Insurance Supervisorsiais.org
165. FSI Insights on policy implementation No 62 Uncertain waters: caniais.org
166. Global natural catastrophe losses in 2025swissre.com
167. What is Probable Maximum Loss (PML)? EP Curveswherobots.com
168. Auto insurance telematics explained: How driving data improves pricing, claims, and customer engagement - Arityarity.com
169. Prescriptive Optimization for Adaptive Auto-insurance Pricing with Telematics Dataarxiv.org
170. Understanding telematics: How your driving behavior influences insurance rates - Insurance Newsinsurancenewsnet.com
171. The Implications and Scope of the NAIC Model Bulletin on the Use of AI by Insurershklaw.com
172. NAIC Model Bulletin on the Use of AI by Insurers, Explained — Meilynxmeilynx.com
173. NAIC AI Model Bulletin: A Guide for AI Vendors - Insurancetryclean.ai
174. Disaster Risk Finance Instruments: Parametric Insurance for Disaster Responsefinancialprotectionforum.org
175. How parametric insurance is building climate resilienceweforum.org
176. Telematics Combined Actuarial Neural Networks for Cross-Sectional and Longitudinal Claim Count Dataarxiv.org
177. Synthetic Dataset Generation of Driver Telematicsarxiv.org
178. Enhancing Claim Classification with Feature Extraction from Anomaly-Detection-Derived Routine and Peculiarity Profilesarxiv.org
179. Parametric insuranceen.wikipedia.org
180. Lloyd's requirements for state-backed cyber-attack exclusionsdwfgroup.com
181. Cyber War Exclusions Explained: State-Backed Attacks, Attribution and Systemic Cyber Riskinsurer724.com
182. Lloyd's cyber war exclusioncliffordchance.com
183. Pennsylvania Supreme Court finds no coverage for COVID-19 business interruption losseskennedyslaw.com
184. North Carolina Supreme Court: Government-Ordered Business Closures During COVID-19 Lockdowns Constitute “Direct Physical loss” Under Insurance Policy Lacking Virus Exclusionitpaystobecovered.com
185. State of the Law for Business Interruption Insurance Coverage for COVID-19 Claimsbtlaw.com
186. COVID-19 Business Interruption Suits: An Overview Of Decisions To Dategoodwinlaw.com
187. Viewpoint: How Climate Change Is Creating Uninsurable Areas Across Europeinsurancejournal.com
188. The Insurance Industry Just Built A Shared Language For One Of Climate Change's Biggest Hidden Risksforbes.com
189. Climate Risks Present a Significant Threat to the U.S. Insurance and Housing Marketsheinrich.senate.gov
190. 2024 home insurance real estate crisisbloomberg.com