2.3 Catastrophe risk

2.3.1 Definition

64. Catastrophe risk covers risks associated with claims events that have yet to occur, and are risks associated with low frequency/high severity events, often arising from an aggregation of multiple claims originating from a single source. Catastrophe risk affects life and non-life business.

2.3.2 ICS methodology

65. Catastrophe risk is segmented at the risk/peril level. In the ICS, peril covers both naturally occurring perils (natural catastrophe) and man-made perils/scenarios (other catastrophe) and their consequences. It considers all losses arising as a consequence of events occurring at any point in time in the next 12 months and should take into account expected business volumes including expected new business to be written during the next 12 months. 
66. The segmentation is the following:

2.3.3 Calibration

2.3.3.1 Use of natural catastrophe models as part of the standard method

69. For the risk assessment of natural catastrophe, the ICS allows the use of stochastic catastrophe models (vendor or proprietary) to calculate the loss amounts resulting from natural catastrophe events.
70. During the development of the ICS, loss amounts related to natural catastrophe were requested for different confidence levels and different measures (VaR and Tail-VaR). Volunteer IAIGs were also asked to report qualitative information regarding the catastrophe model used as well as how the model was used.
71. Allowing the use of natural catastrophe models as part of the standard method is perceived as an appropriate approach leveraging on scientific risk assessment methodologies, embedded in such models, and aligning the risk assessment with generally recognised market practices.
72. To ensure the appropriateness of the models used, the ICS embeds safeguards about the quality of the model themselves but also about how those are used by IAIGs.

2.3.3.2 Man-made catastrophe scenarios

73. The man-made catastrophe scenarios have been defined, in the context of the standard method, to support the measure the 99.5% VaR over one year for each individual IAIG.
74. This involves a high degree of expert judgement and some simplifications acceptable for a standard method, focusing on achieving an appropriate level of comparability and accuracy while preserving a desirable level of simplicity and practicality. The Lloyds Realistic Disaster Scenarios  were used as a basis to derive the ones embedded in the ICS.
75. Also, one should note that the man-made catastrophe scenarios included for the purpose of the ICS have been selected based on their materiality. To do so, 2015 Field Testing requested loss amounts related to associated with natural catastrophes and the following man-made scenarios:

• a. a terrorist attack scenario;
• b. a liability catastrophe scenario;
• c. a pandemic scenario;
• d. a marine collision scenario;
• e. an aviation collision scenario; and
• f. a credit and surety scenario.

76. The obtained breakdown of total Catastrophe risk charges in 2015 Field Testing results helped refine and simplify the methodology, as the marine and aviation collision scenarios proved not to be material compared to other scenarios.

2.3.3.3 Diversification within Catastrophe risks

77. The contribution to the Catastrophe risk charge and ultimately to the ICS capital requirement from other Catastrophe components of the risk charge are considerably reduced by the effect of diversification.
78. For the purpose of calculating the Catastrophe risk charge, the other catastrophe scenarios are assumed to be mutually independent and independent of the natural catastrophe perils. Consequently, the total ICS Catastrophe risk charge is calculated as follows:

• $ICS_{cat}$ = $\sqrt{ICS_{NatCat}^2 + ICS_{Terror}^2 + ICS_{Pand}^2 + ICS_{credit}^2}$