PlumBot: Bot: Creating new article from JSON

2026-03-16T11:29:53Z

Bot: Creating new article from JSON

New page

📐 '''Calibration standard''' refers to the quantitative benchmark or confidence level to which a [[Definition:Risk measure | risk measure]] or [[Definition:Capital model | capital model]] is set when determining the amount of capital an [[Definition:Insurance undertaking | insurance undertaking]] must hold. In the insurance regulatory context, this concept is most prominent under [[Definition:Solvency II | Solvency II]], where the [[Definition:Solvency capital requirement (SCR) | solvency capital requirement]] is calibrated to a 99.5% [[Definition:Value at risk (VaR) | Value at Risk]] over a one-year time horizon — meaning the insurer should hold enough capital to survive all but a 1-in-200-year adverse event. Other regimes adopt different standards: the [[Definition:Swiss Solvency Test (SST) | Swiss Solvency Test]] uses a 99% [[Definition:Tail value at risk (TVaR) | Tail VaR]] (expected shortfall), while [[Definition:Risk-based capital (RBC) | risk-based capital]] frameworks in the United States and under [[Definition:China Risk Oriented Solvency System (C-ROSS) | C-ROSS]] in China embed their own target confidence levels, sometimes implicitly through factor-based charges rather than explicit probabilistic statements.

⚙️ The calibration standard governs how stress scenarios, loss distributions, and correlation assumptions within a capital model are parameterized. When an insurer develops an [[Definition:Internal model | internal model]], regulators require it to demonstrate that the model's outputs are consistent with the prescribed calibration standard — for instance, that the modeled SCR corresponds to losses at the 99.5th percentile. This involves rigorous [[Definition:Model validation | validation]] processes, including backtesting against historical data, sensitivity analysis, and comparison with the [[Definition:Standard formula | standard formula]] results. The choice of risk measure matters as well: VaR captures the threshold loss at a given percentile but says nothing about the severity of losses beyond that point, whereas TVaR — favored by the SST and increasingly discussed by the [[Definition:International Association of Insurance Supervisors (IAIS) | IAIS]] in the context of the [[Definition:Insurance capital standard (ICS) | Insurance Capital Standard]] — accounts for the average of losses in the tail, providing a more conservative picture of extreme risk.

🎯 Differences in calibration standards across jurisdictions are far from academic — they directly affect how much capital insurers must raise, how competitively they can price products, and whether regulatory regimes are deemed "equivalent" for cross-border supervisory purposes. A regime calibrated to a lower confidence level may appear to require less capital, but it may also embed conservatism elsewhere (in reserving rules, for example) that compensates for the gap. Equivalence assessments between Solvency II and regimes like [[Definition:Bermuda Monetary Authority (BMA) | Bermuda's]] or Japan's hinge in part on whether the overall calibration delivers comparable [[Definition:Policyholder | policyholder]] protection, even when the technical specifications differ. For [[Definition:Reinsurance | reinsurers]] and global groups operating under multiple regimes, understanding each jurisdiction's calibration standard is essential for efficient [[Definition:Capital management | capital management]] and for explaining solvency positions to [[Definition:Rating agency | rating agencies]] that maintain their own proprietary capital benchmarks.

'''Related concepts:'''
{{Div col|colwidth=20em}}
* [[Definition:Solvency capital requirement (SCR)]]
* [[Definition:Value at risk (VaR)]]
* [[Definition:Tail value at risk (TVaR)]]
* [[Definition:Internal model]]
* [[Definition:Standard formula]]
* [[Definition:Insurance capital standard (ICS)]]
{{Div col end}}

Definition:Calibration standard - Revision history

PlumBot: Bot: Creating new article from JSON