PROPERTY CASTASTROPHE

Click on the headings below for more information.

• COBRA

Parametric reinsurance, such as Industry Loss Warranties (ILWs) and catastrophe bonds have increasingly become core elements of many reinsurance buying strategies. Given the nature of these products, they are unlikely to perform in exactly the same way as a traditional ultimate net loss (UNL) reinsurance program. Basis risk is a generic term for the difference in outcome and/or performance between a UNL program or layer and an equivalent parametric structure.

 

Basis risk has often been characterized as an unknowable quantity. However, through the development of the Carvill Operational Basis Risk Analysis (COBRA), ReAdvisory has shown that both good and bad basis risk can indeed be quantified and measured objectively. By using output from the catastrophe models, COBRA can identify situations where the performance of a parametric cover will diverge from an equivalent UNL layer. Using this analysis, COBRA can calculate both a probability of good and bad basis risk outcomes, and develop a severity curve for good and bad basis risk. Further, COBRA assigns a value to both good and bad basis risk so that clients can see what the actual dollar cost is for increasing or decreasing basis risk.

 

As part of COBRA, ReAdvisory performs a review of the modeled portfolio. By looking at how a client’s modeled losses are related to the industry’s modeled losses, any concentrations of risk are highlighted. These may be geographic concentrations, peril-specific concentrations or both. By highlighting any concentrations, these can be addressed specifically in the structuring of a parametric program which has the impact of reducing the extremes of bad basis risk severity.

 

• Catastrophe Modeling

Catastrophe modeling (‘cat modeling’) is the underpinning for virtually all property placements in the reinsurance market today. For a property catastrophe re/insurer accurately measuring the risk of an individual client or of the whole portfolio of business is the most important activity, since this impacts the prices charged and the capital held. Since modeling catastrophes is a non-trivial exercise, re/insurers typically do not develop their own models, preferring to use commercially available models.

 

Given that cat modeling is a very complicated proposition, developing a comprehensive model in-house at a re/insurer is impractical. Commercial catastrophe models have become a necessary part of the process to gain broad market acceptance to pricing differentials.

 

Through our strong links with each of the model vendors, our deep scientific understanding of the models, and our wide practical experience of using each of the models, ReAdvisory is able to support clients throughout all parts of a property catastrophe analysis. We have helped a wide variety of clients understand and effectively utilize cat models and cat model analysis by:

 

• Helping clients manage and verify exposure data
• Educating on the elements of a cat model analysis
• Developing client-specific metrics from cat model data which best fit a client’s business practice

 

On behalf of clients we have used cat models to support:

 

• Traditional UNL reinsurance placements
• Industry loss warranty (ILW) transactions
• Parameterized reinsurance cover, such as ‘storm-in-a-box’ concepts
• Portfolio management and optimization
• Rating agency submissions

 

• Portfolio Optimization

ReAdvisory has developed a number of portfolio optimization tools which are used to identify:

 

• Accounts which represent a ‘drag’ on the portfolio
• Accounts which indicate where new business could profitably be added
• Accumulations of risk

 

ReAdvisory’s portfolio optimization system uses a cyclical technique to sequentially determine which accounts or policies have the most detrimental effect on the portfolio as a whole. This technique allows the correlation between individual risks to be explicitly captured. The portfolio optimization system can perform analyses to identify the worst 10/20/100 risks in the portfolio or can select out risks until a minimum pure cat premium is attained. The analysis is based on three measures of risk:

 

• Return period loss
• Excess average annual loss (XSAAL)
• Tail conditional expectation (also known as Tail Value-at-Risk)

 

The methodology is also used to identify where new business could be added by determining which policies have the least detrimental effect on the portfolio.

Using state-of-the-art modeling and mapping solutions, ReAdvisory can identify where accumulations of risk occur in a client’s portfolio. ReAdvisory’s experienced specialists use this data to help clients mitigate the effects of accumulations in high-risk zones.

 

Contact
Dr. Steve Smith
sesmith@carvill.com