Definition:Value at risk
📐 Value at risk (VaR) is a statistical measure widely used across the insurance industry to estimate the maximum potential loss on a portfolio of risks — whether investment assets, underwriting exposures, or an entire enterprise — over a specified time horizon and at a given confidence level. In insurance, VaR serves as a foundational building block for regulatory capital frameworks, internal models, and enterprise risk management programs: for example, the Solvency II Solvency Capital Requirement (SCR) is defined as the VaR of basic own funds at a 99.5% confidence level over a one-year horizon, meaning regulators expect insurers to hold enough capital to survive all but the worst one-in-200-year loss scenarios.
⚙️ Insurers calculate VaR using several methodologies depending on the risk type, data availability, and regulatory requirements. Parametric (variance-covariance) approaches assume a known distribution and work well for market risk on liquid investment portfolios. Monte Carlo simulation is more common for insurance-specific risks — particularly catastrophe risk and reserving risk — where loss distributions are skewed and fat-tailed. Historical simulation draws on past loss experience but can underestimate tail risk if the observation window lacks extreme events. Under Solvency II, insurers may use the standard formula or apply for approval of a full or partial internal model to calculate SCR; either way, VaR at 99.5% is the target metric. In contrast, China's C-ROSS framework and the RBC system used in the United States and parts of Asia employ factor-based approaches that implicitly embed VaR-like concepts but do not always express the capital requirement as a single VaR figure. On the investment side, insurers managing multi-billion-dollar bond, equity, and alternative investment portfolios rely on daily or weekly VaR reports to monitor market risk exposure and ensure compliance with board-approved risk appetite limits.
🎯 Despite its ubiquity, VaR has well-known limitations that insurance professionals must account for. It tells you the threshold of loss at a given confidence level but says nothing about the severity of losses beyond that threshold — a shortcoming that has led many insurers and regulators to supplement VaR with Tail VaR (TVaR), also known as Conditional Tail Expectation, which averages losses in the tail beyond the VaR point. The Swiss Solvency Test (SST), for instance, uses TVaR at 99% rather than VaR. VaR can also give a false sense of precision when applied to highly non-normal distributions typical of natural catastrophe or liability risks, where a single extreme event can dwarf the modeled estimate. Nonetheless, as a common language for quantifying and communicating risk, VaR remains indispensable — enabling insurers, reinsurers, regulators, and rating agencies to compare risk profiles across companies, lines of business, and geographies on a broadly consistent basis.
Related concepts: