Lapse, surrender and recovery curves estimation are a key components in the pricing and valuation of Insurance business.
Dynamic web dashboard can be used in day to day actuarial practice to reduce manual tasks and allow actuaries to focus on value added analysis.
Benefits of Dynamic dashboards
Regulatory requirements, IFRS17, increased market pressure and internal needs require Actuarial team to increase segmentation, periodicity and granularity of assumption update
Need for automation
The volume of information to be treated and the number of portfolio subsets to be studied makes effective automation an important brick in the closing process of insurance companies.
Data visualisation
Complexity and volume of information make effective data visualisation a must have to detect and manage trends in portfolios.
Flexible, immediate and distributed
The ability to query “on-demand” the estimation on various portfolio subsets provides increased reactivity for informed decision-making and proactive management.
Methodology
The Kaplan-Meier estimator - is the reference for univariate survival function estimation. More information can be found on the wiki page or in any statistical resource.
\[\hat{S}(t) = \prod_{t_i \lt t} \frac{n_i - d_i}{n_i}\]with \(d_i\) the number of “failures” at time \(t\) and \(n_i\) the exposed population just before time \(t\).
Tip:
The Kaplan-Meier estimator intuition is generally easy to grasp on a simple practical example. One of them can be found here: KaplanMeier_Refresh.xlsx
Lifelines:
The lifelines package provides a direct Kaplan-Meier estimator within the python environment, along with many other functionalities: Lifelines package