Section 4 : Cats Essentials Flashcards
Define a catastrophe model and its primary purpose.
A catastrophe model is a statistical tool used to estimate potential losses from catastrophic events by combining historical data, hazard intensity, and vulnerability assessments.
List and briefly describe the five main components of a catastrophe model.
- Event Module – Simulates potential catastrophe events.
- Hazard Module – Determines event intensity at various locations.
- Exposure Module – Represents insured properties and assets.
- Vulnerability Module – Estimates damage given hazard intensity.
- Financial Analysis Module – Converts damage estimates into financial losses.
Name the primary perils that catastrophe models cover. (natural and non natural)
Natural perils: Earthquakes, floods, hurricanes, wildfires, convective storms.
Non-natural perils: Terrorism, pandemics.
Describe two key probability metrics used in catastrophe modelling.
Occurrence Exceedance Probability (OEP): Probability that the largest single event in a year will exceed a given loss amount.
Aggregate Exceedance Probability (AEP): Probability that total losses from all events in a year exceed a certain threshold.
How are catastrophe models used in setting insurance rates?
Models estimate expected losses per policy by simulating potential cat events, helping insurers determine risk-based premiums and assess reinsurance needs.
What are the step by step : Using Cat model in rate level indication
- Calculate the Expected Loss Cost
The core output from catastrophe models is the annual average loss (AAL) per exposure unit (e.g., per home or per $1,000 of insured value).
AALi=∑(pi×Li) -> sum from 1 to n
Where:
N = Number of simulated catastrophe events
pi= Probability of event ii occurring
Li = Estimated loss from event i
- Convert to Pure Premium (Loss Cost Per Policy)
Pure Premium= AAL / Number of Policies - Add expense loadings and profit margin
Final rate = (Pure Premium+Fixed Expenses+Variable Expenses+Profit Load) / (1- Commission - Contingency) - Adjust for Catastrophe Load and Reinsurance Costs
Final Rate + Reinsurance costs or CATL
What does a 1-in-100-year flood mean?
It means there is a 1% probability per year of an event of that magnitude occurring.
Why should insurers be cautious when using catastrophe model outputs?
- Uncertainty in hazard frequency and severity.
- Climate change effects not always well captured.
- Reliance on historical data may underestimate future risk.
- Secondary uncertainty (e.g., two buildings exposed to the same hazard may suffer different damage levels).
Explain the concept of secondary uncertainty in catastrophe modelling.
It refers to variability in damage outcomes for similar exposures due to differences in construction quality, age, and other local factors.
What methods can insurers use to test the reliability of catastrophe models?
- Comparing model outputs to historical event losses.
- Conducting sensitivity analyses on key assumptions.
- Benchmarking against multiple models from different vendors.
Why do catastrophe models need to be updated due to climate change?
Climate change alters event frequency, severity, and geographic distribution, making historical data less predictive of future risk.
How do reinsurers use catastrophe models?
They estimate maximum probable losses and determine appropriate reinsurance coverage and pricing to manage risk.
Why do regulators use catastrophe models?
- Ensure insurers hold enough capital to cover extreme events.
- Support solvency assessments.
- Improve risk management strategies.
Why is flood risk difficult to model?
- Limited historical flood data in some regions.
- Highly localized risk factors (e.g., elevation, drainage systems).
- Changing climate patterns affecting flood frequency and severity.
Differentiate PML and TVaR in catastrophe risk analysis.
PML: The largest likely single-event loss within a given confidence interval.
TVaR: The average loss beyond a threshold, capturing extreme tail risks.
How can catastrophe models lead to ethical concerns in insurance pricing?
- Higher premiums or denied coverage for high-risk communities.
- Bias in data or assumptions leading to unfair pricing.
- Lack of transparency in model methodologies.
What are Non-Modeled Catastrophes?
(Example: Hailstorms for personal auto coverage)
- Occur regularly and cause many claims.
Method to estimate losses:
- Use a catastrophe-to-non-catastrophe loss ratio over 10-30 years.
- Or, develop a pure premium or loss ratio for these events.
What are Modeled Catastrophes?
(Example: Hurricanes)
Occur rarely but cause high severity losses.
Estimation method:
- Use catastrophe models to estimate likelihood and damages (built by insurance pros and meteorologists).
- Calculate the expected annual loss by multiplying likelihood and damage.
- Add this to non-catastrophe losses for final expected losses in ratemaking.
State four applications of catastrophe modeling for insurance.
Any four of the following are acceptable:
* Ratemaking
* Underwriting and Risk Selection
* Loss Mitigation
* Catastrophe Reinsurance
* State and federal public policymakers use catastrophe models to address public policy issues.
* Capital adequacy (sensitivity) testing
* For reserving purposes
Provide a consideration in the selection of a risk load
Variability (Standard Deviation) and Uncertainty in the losses estimates
