est.CKLS.MCMC: MCMC estimation for the CKLS model

In alejandralopezperez/estsde: Parametric Estimation of SDE

MCMC estimation for the CKLS model

Description

Parametric estimation for the CKLS model using Markov Chain Monte Carlo and involving data augmentation, as proposed in Elerian et al. (2001) and Eraker (2001). The parametric form of the CKLS model used here is given by

dX_t = (α - κ X_t)dt + σ X_t^γ dW_t.

Usage

est.CKLS.MCMC(X, Delta = deltat(X), par = NULL, niter = 4000,
  burn_in = 1000)

Arguments

X	a numeric vector, the sample path of the SDE.
Delta	a single numeric, the time step between two consecutive observations.
par	a numeric vector with dimension four indicating initial values of the parameters. Defaults to NULL, fits a linear model using generalized least squares with AR1 correlation and a power variance heteroscedasticity structure.
niter	an integer, number of iterations.
burn_in	an integer indicating the number of initial iterations to be discarded.

Value

A list containing a matrix with the estimated coefficients and the associated standard errors.

References

Elerian, O., Chib, S., and Shephard, N. (2001). Likelihood inference for discretely observed nonlinear diffusions. Econometrica, 69(4):959–993.

Eraker, B. (2001). MCMC analysis of diffusion models with application to finance. Journal of Business & Economic Statistics, 19(2):177–191.

Examples

set.seed(987)
x <- rCKLS(480, 1/12, 0.09, 0.08, 0.9, 1.2, 1.5)
est.CKLS.MCMC(x)

alejandralopezperez/estsde documentation built on Sept. 4, 2022, 4:48 a.m.