On the structure of the stochastic processes of mortgages in Spain (Q880894)

The mortgages process is described as a Poisson counting process \(N(t)\) with varying intensity \(\lambda(t)\) (which is also considered as a stochastic process). It is assumed that \(k\) different sample paths \(N_\omega(t)\), \(\omega=1,\dots,k\) of \(N(t)\) are observed. Point estimation and confidence intervals for \(\lambda\) are considered. Functional principal component analysis is proposed for continuous modelling of the intensity. Simulation results and application to mortgages in Spain (where the mortgages in 52 provinces are considered as different sample paths of the process) are presented.