Psicle has a suite of gadgets that perform a range of statistical or domain-specific functions. They are coded in external mathematical and statistical languages (such as F#, Wolfram Mathematica, and R) and integrate seamlessly into the Psicle environment. The list of current gadgets includes:
Rapidly performs a range of scalable BODMAS calculations. This is to be embedded into PsiCode in Psicle 2.0
Generates undiscounted and discounted cash flows using input values, patterns, yield curves and bespoke user manipulations. This gadget is fundamental to Solvency II TP and IFRS 17 cashflow modelling in Psicle. This is to be embedded into PsiCode in Psicle 2.0
A tool to analyse data from numerical and categorical type random variables.
A sophisticated, interactive pricing and exposure management tool for primary and excess layer pricing, as well as portfolio level analysis.
Generalised Linear Model
This gadget fits GLMs to big data – used primarily in personal lines business for claims modelling.
Provides a series of options to quickly manipulate patterns into different formats, including the ability to Shift, Stretch, Scale and Accumulate patterns. It also includes a wide range of interpolation techniques. This is to be embedded into PsiCode in Psicle 2.0
Functionality to perform triangle reserving techniques. The gadget can be run “hands-free” using a series of reserving rules, or interactively using our graphic interface which allows users to interact with their data and make selections on exclusions and techniques for different classes of business. The reserving gadget has the standard range of reserving techniques (Chainladder, Bornhuetter-Ferguson, IELR). The included methodologies are being constantly expanded upon to include a wider range. We have recently included the Munich Chainladder.
Calculates the full Solvency II (EU) and SAM (South Africa) Standard Formula SCR in line with regulatory specifications and automatically populates quarterly reporting templates and Lloyd’s reports.
This is our native language and is due to be released in Psicle 2.0 and combines a range of the gadgets above into one united framework. It will also include many additional functions, such as the ability to simulate from a range of discrete, continuous and compound distributions and apply correlation structures.
Principal Component Analysis