We’re going to calibrate both a Gompertz model and the stochastic CIR model to the HMD IT 1x10 table (2000-2009) table for a 65 years old man.
Here follows the survival function implied by the table.
Let’s fit a Gompertz model:
## theta = 0.1080508 mu0 = 0.01285239
Let’s calibrate the CIR model to the same survival function
## alpha = 1e-04 theta = 0.1105087 sigma = 0.01 mu0 = 0.01212986
Monte Carlo is used to estimate the residual life time. Finally we plot some simulated paths out of the calibrated CIR model along with the fitted Gompertz intesity.
## Expected residual life = 17.57362