A Python library implementing a bisection method to find the optimal convergence speed parameter α for the Smith & Wilson algorithm used in insurance rate extrapolation.
Simple bisection method that finds the optimal parameter α for the Smith & Wilson algorithm.
This tool is used by actuaries and financial analysts in the insurance sector to calibrate the convergence speed parameter α for the Smith & Wilson algorithm, ensuring accurate interpolation and extrapolation of interest rate curves. It automates the root-finding process to match long-term yield behavior, facilitating risk assessment and data security in financial modeling.
To ensure consistency with EIOPA guidelines, the lower bound for α should be set to at least 0.05. Users should provide accurate market data inputs and set appropriate precision and iteration limits to avoid non-convergence. The repository duplicates Smith & Wilson functions for completeness; users should verify compatibility if integrating with other implementations.
Clone the repository from GitHub
Ensure Python environment is set up with numpy installed
Import the necessary classes (SWCalibrate, SWExtrapolate, Galfa, BisectionAlpha) from the repository modulesGalfa(M_Obs, r_Obs, ufr, 0.15, Tau)
Calculates an example α value using the Galfa function with given market maturities, yields, ultimate forward rate, and tolerance.
BisectionAlpha(0.05, 0.5, M_Obs, r_Obs, ufr, Tau, Precision, 1000)
Runs the bisection method to find the optimal α between 0.05 and 0.5 with specified market data, ultimate forward rate, tolerance, numeric precision, and iteration limit.