Date(s) - 14/06/2021
12:00 pm - 1:00 pm
NoCA Training: Capital Models and Management is being provided by Dr Matt Modisett, PhD, FIA, ASA, MMA.
The course is an overview of capital model creation, either to understand the standard formula or with an eye to creating an Internal Model or Partial Internal Model. The course will provide insight into what aspects should be considered from any implementation, with reference where to investigate specific topics further.
The course is independent of any programming knowledge or implementation. Accordingly, specific code is not provided. Illustrative examples will be drawn from Investments and Life products.
Matt has over 30 years of working experience in insurance modelling (Life, Capital, Investments) including 13 years in Solvency 2 modelling. He is also author of the book Solving Solvency: 100 tips for Managing Insurance Capital in a Shifting Regulatory environment. He is a Fellow in the UK and Hungary and an Associate in the USA.
The course will comprise of 3 lessons of 1 hour each. Registration closes 24 hours before the first session starts.
Course Level: Beginner
- 14th (Monday) June 2021, 12pm – 1pm
- 21th (Monday) June 2021, 12pm – 1pm, and
- 28th (Monday) June 2021, 12pm – 1pm
A NoCA Training Certificate will be provided on the completion of the course.
Some examples may also be given in R, if participants are interested, but R is not required for the course.
Course Sponsor: Summit Recruitment AG
- Overview Solvency II modelling
- Standard Formula vs Internal Model
- Selection of Risk Drivers
- Standalone Risk Factors: Loss models from historic data
- Statistical analysis
- Parameter error estimate, uncertainty measures
- Selection of distributions including non-statistical criteria.
- Basis risk
- PCA – Loss models for related risk drivers
- Correlation between risk drivers
- Confidence interval for correlation and the Fisher Transform
- Impact of correlation uncertainty
- Positive semi-definite (PSD)
- Aggregation – from several risk drivers to one capital figure
- Standard Formula
- Motivation from normal loss distributions
- Capital Aggregation Model (1): Without copulas
- Choice of Correlation Matrix aggregation vs scenario testing
- Both use same correlations (generally)
- Both are theoretically the same.
- Capital Aggregation Model (2): with Copulas
- Euler and Capital Allocation
- Stagging the Aggregation of an Internal Model
- Overlapping Data
- Proxy model
- Risk Margin
- Matching Adjustment
- Future of capital models (e.g. IFRS17, Climate, Social responsibility)
Cost of course: GBP 80.00.
Bookings are closed for this event.