Definition:Law of large numbers
📊 Law of large numbers is the statistical principle at the foundation of insurance pricing and risk pooling: as the number of independent, similarly exposed units in a pool increases, the actual loss experience of the group converges toward the expected loss. For insurers, this means that writing a sufficiently large and diversified book of policies makes aggregate outcomes more predictable, enabling the business to set premiums with greater confidence and maintain stable loss ratios over time.
🔬 In practice, actuaries rely on the law of large numbers when building rating models and establishing loss reserves. They analyze historical data from thousands or millions of exposures to estimate expected claim frequency and severity for a given line of business. The principle works best when the risks are reasonably homogeneous and independent — conditions that hold well in personal auto or homeowners insurance but break down in catastrophe-exposed or highly correlated portfolios. Where the independence assumption fails, insurers turn to supplementary tools such as catastrophe modeling, reinsurance, and risk-based capital requirements to manage the residual volatility that the law of large numbers alone cannot eliminate.
🌐 Understanding this principle separates viable insurance ventures from speculative ones. A startup MGA entering a niche market, for instance, must recognize that a thin portfolio will produce volatile results regardless of how accurately individual risks are priced; only as volume grows will actual performance stabilize around expectations. The law of large numbers also informs regulatory frameworks: regulators require minimum capital and solvency margins precisely because smaller or concentrated books cannot rely on statistical convergence to absorb adverse deviations. In the insurtech era, access to richer data and broader distribution channels accelerates the path to scale, making the law's benefits attainable faster — but never eliminating the need for disciplined underwriting to ensure the underlying assumptions hold.
Related concepts