25.10 Advanced Scenario Planning & ERM

Model Expansion

• Expend into several LoBs

• Internal consistency requires coordination

• Need to model dependencies

• Responses will be politically charged, as limited u/w capacity must be allocated across the company

  • Better to do this during planning than in the heat of a market crisis

Stochastic Scenario Planning

Generate stochastically simulated scenarios that have responses based on a set of rules (This has been done with success in asset management)

1. Need to define goals that we want to achieve

   (e.g. profit or premium level etc)

2. Need to have downside constraints

   (e.g. capital loss of x% with y% probability)

3. Need to determine the essential environmental variables that define the state of the world

   (e.g. size of market, price level, current market share, etc)

4. Define action rules that respond to environmental variables

   (e.g. hold price, increase u/w capacity according to price adequacy levels)

   • Express as mathematical formulas that can be programmed into the simulation

25.10.1 Agent-Based Modeling

Agents can be: competitors, customers, regulators

Program each agent to respond to the state of the market

• e.g. Chase market share, focus on technical price, customer may reduce coverage if prices are high, etc

Creates a complex system with emergent properties that describe the behavior of agents

• There emergent properties are often difficult to ascertain without running the simulations