|
Risk disclosures
|
 |
|
|
Realistic Disaster Scenario (RDS) analysis
The Group has a defined event risk appetite which determines the maximum net loss that the Group intends to retain for major catastrophe scenarios. Currently these are a maximum of £200 million for Syndicate 2001 and £145 million any one zone or £175 million for a multi-zonal loss for Amlin Bermuda Ltd.
The above losses are extreme events - with an occurrence probability in excess of 1 in 100 years estimated for the natural peril or elemental losses. The Group also adopts risk appetite maximum net limits for a number of other scenarios including aviation collision (£120 million), marine collision (£125 million) and North Sea rig loss (£75 million).
The risk appetite policy recognises that there may be circumstances in which the net event limit could be exceeded. Such circumstances include non renewal or delay in renewal of reinsurance protection, reinsurance security failure, or regulatory and legal requirements.
A detailed analysis of catastrophe exposures is carried out every quarter and measured against the risk appetite. The following assumptions and procedures are used in the process:
- The data used reflects the information supplied to the Group by insureds and ceding companies. This may prove to be inaccurate or could develop during the policy period;
- The exposures are modelled using a mixture of stochastic models and underwriter input to arrive at 'damage factors' - these factors are then applied to the assumed aggregate exposure to produce gross loss estimates. The damage factors may prove to be inadequate.
- The reinsurance programme as purchased is applied - a provision for reinsurer counterparty failure is included but may prove to be inadequate.
- Reinstatement premiums both payable and receivable are included.
Whilst the losses arising from the Atlantic hurricanes in 2004 and 2005 broadly support the RDS analysis there is no guarantee that the assumptions and techniques deployed in calculating these figures are accurate. Furthermore there could also be an unmodelled loss which exceeds these figures. The likelihood of such a catastrophe is considered
to be remote but the most severe scenarios modelled are simulated events and these simulations could prove to be unreliable.
|
|  Print |
|
Top