Skip to content

Using a Monte Carlo simulation, we analyzed an auto insurance pool to identify its risk characteristics through Gamma distribution and the Law of Large Numbers. This analysis also helped calculate the minimum premium required to achieve the APRA sufficiency level.

Notifications You must be signed in to change notification settings

quenstance/2018_ACTL2111_Auto_Insurance_Monte_Carlo_Simulation

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 

Repository files navigation

Overview

This project, ACTL2111 Probability & Mathematical Statistics (Individual Assignment 1), used a Monte Carlo simulation to model an auto insurance pool. The goal was to analyze the statistical properties of pooled risks and calculate a minimum premium to ensure financial solvency. The model was built to be easily adjusted for key parameters, demonstrating a flexible and reusable approach. For more details, refer to the project webpage.

How to Use

The simulation was built in Microsoft Excel. The spreadsheet allows users to adjust key parameters:

  • Number of drivers (n)
  • Individual probability of accident (p)
  • Average size of accident (beta)
  • Histogram bin minimum and bin interval
  • Probability of sufficiency

All calculations and results will automatically update based on the new parameters.

Methodology

The simulation ran for an accident year with 1,000 drivers. The number of claims and their costs were simulated using Excel's RAND() function, leveraging the property that the sum of independent exponential random variables follows a Gamma distribution. Key metrics, including mean and standard deviation of total claims cost and the average cost per policyholder, were calculated using AVERAGE, STDEV.S, and PERCENTILE functions to derive the minimum premium.

Limitations

  • Tool: The use of Excel's RAND() function is a key limitation, as its volatile nature makes results difficult to reproduce consistently. For larger-scale simulations, professional tools like Python or R are superior due to their performance and reproducibility.
  • Model Assumptions: The model uses static parameters and simplifies real-world conditions by assuming claims are independent and follow a perfect exponential distribution. While a reasonable academic assumption, real-world data may exhibit dependencies, which could affect the accuracy of the final premium calculation.

Author

Quenstance Lau

About

Using a Monte Carlo simulation, we analyzed an auto insurance pool to identify its risk characteristics through Gamma distribution and the Law of Large Numbers. This analysis also helped calculate the minimum premium required to achieve the APRA sufficiency level.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published