This page is part of an in-progress commentary by Roger
Frye
on High Altitude
Thinking: The International Informatics Summit.
Swiss Re: Risk is our business. The downside of others is our business.
Landscape: Nonhazard -- business, financial, political. VS hazard
Risk management is protecting the values of an enterprise.
Reinsurance is insurance for insurance companies. Are reinsurers in the clouds? Landscape showing houshold policyholders, with insurers in the alps and REs in the clouds.
Propagation: Damage -> loss to economy. Liability, Insurance, reinsurance
Founded 1863. Strengths: people and culture, skills and processes -- internal risk management.
Basic principal of Insurability. Must be able to access.
New dimensions: foreseeable, regional, reversible, ...
Typical example shows an extremely complex picture about GMO corn polluting regular corn. Several layers affected.
Claims ar not scenarios. Claim: something happened with a cost. Goal is adequate resources. Scenarios: nothing happened yet, how much? Goal is adequate RAC risk adjusted capital.
Look again at clouds. Ground is law of law numbers
Fundamental questions. How perceived, identified, accessed?
Property/Casualty. knowb perils, direct correlation, statistical data vs not.
The risk of change: tech, society, economics and finance, law.
Finding the right scenario. Not ONE for all insurers and REs. Have to customize.
Project Risk Perception and risk prediction. Develop early warning, ... Diagram shows shorcut.
Suppositions: earlier detection -> relevant info available -> change awareness.
outside world = risk landscape vs perceive in inside world. Data, recognition, information.
Findings: info does not have relevance, but we give relevance to it.
risk prception is vital to Swiss Re. An active process. Observation, not knowledge.
(Re-)cognition depends on: perception, knowledge, contest, method. So Swiss Re introduced SONAR.
Observation (all employees), collection (SONAR coordinators), processig and assessment, implementation, feedback.
Second part of project. 1st was early detection. 2nd new kinds of risks.
History 1980 NERIS by Frederick Vester. Now runs on Windows.
Insurance World Simulator starting at SFI with John Casti and continuing. Markets, Nature, Economy, 5 insurers, 5 REs. Cooperation of experts of the same profession, agreement on ONE insurance world. Useful, less guesswork.
NERIS vs IW-Sim. Easy vs blackbox, slow vs fast, high effort vs low, ..
The Risk&Value Simulator.
Simulators are not forecasting tools. Not product design tools. Not pricing tool.
Pros: easy, simple, checp, fast, adaptable, intriguing.
Cons: black box, unreproducible, maintenace, upgrade difficult.
Senior management quickly decides negatively. Junirs don't understand. Conclude: need experienced user, co-eveloped, no predictive use yet (head in the clouds), good for training, decision support tool, communication tool,. Economists want numbers, simulators generate scenarios.
Please make the network visible.
Question: validation: Answer: That is one of the problems because did not yet happen.
Question apply SONAR to terrorism? Answer: no works the other way around, not after event. Didn't predict the coincidence of events.
END