Introduction
The insurance industry plays a key role in offsetting risks. While the theory of actuarial science provides a good basis for understanding individual risks, there is at this point very little scientific understanding of the systemic risks in the insurance industry. A step in this direction is to use agent-based modelling. Agent-based models (AMB) are computer simulations used to study the interactions between people, things, places, and time. The basic concept of ABM is that by describing simple rules of behavior for individual agents and then aggregating these rules, researchers can model complex system. This project aims to create a proof of principle that demonstrates how this type of modelling can be useful in providing a deeper understanding of phenomena such as the insurance cycle.
Explaining the science
Simulators for socio-economic systems have so far only delivered post-hoc explanations and analysis. The interdependence between a large number of actors is very computationally demanding, and models typically have numerous exogenous aspects, such as market data or parameterised data generators. Moreover, capturing the adaptive nature of the strategies used by market participants, sometimes based on historical data, is challenging. Finally, the incentives of these adaptive agents need to be accounted for so as to characterise agents’ behaviour and link the individual agents’ decisions to market macro behaviour.This project will take a significant step towards building agent-based models that prove to be predictive simulators for specialized insurance markets.
Project aims
- Improving our ability to simulate, experiment with, and therefore understand the dynamics of these markets, forecast future market environments, and ultimately help better decision making.
- Provide a better qualitative understanding of what drives the insurance cycle and what determines its duration and amplitude, as well as the impact and drivers of these dynamics
- Build an open-source tool that makes it possible for an insurance company to identify where we are in the cycle and optimize its planning based on that information by predicting how the cycle will unfold in the future.
- Develop heuristics which will enable better estimations of the current and future market environment.
Applications
The outcome of the project will benefit:
- The tools developed in this project will be of interest to a wide range of insurance practitioners.
- The benefits of advances in realistic simulation go beyond insurance markets, notable examples include house markets, online retail, and epidemiology.
Recent updates
To read more ABM for insurance markets, you can read the paper here: Exploring the Dynamics of the Specialty Insurance Market Using a Novel Discrete Event Simulation Framework: a Lloyd’s of London Case Study
Organisers
Prof. Carmine Ventre
Chair in Computational Finance at King's College LondonPriscila Lopez-Beltran
Research Project Manager, Finance and Economics ProgrammeTony Zemaitis
Programme Manager, Finance and EconomicsResearchers and collaborators
Dr Teresa Yu Bi
Research Associate King's College LondonContact info
For more information, email Priscila López-Beltrán at [email protected]