factorstochvol (version 1.1.0)

predh: Predicts factor and idiosyncratic log-volatilities h

Description

predh simulates from the posterior predictive distribution of the latent log-variances h, both for factors as well as for idiosyncratic series.

Usage

predh(x, ahead = 1, each = 1)

Value

List of class fsvpredh containing two elements:

idih

Array containing the draws of the latent idiosyncratic log-volatilities.

factorh

Array containing the draws of the latent factor log-volatilities.

Arguments

x

Object of class 'fsvdraws', usually resulting from a call to fsvsample.

ahead

Vector of timepoints, indicating how many steps to predict ahead.

each

Single integer (or coercible to such) indicating how often should be drawn from the posterior predictive distribution for each draw that has been stored during MCMC sampling.

See Also

Other predictors: predcond(), predcor(), predcov(), predloglikWB(), predloglik(), predprecWB()

Examples

# \donttest{
set.seed(1)
sim <- fsvsim(series = 3, factors = 1) # simulate
res <- fsvsample(sim$y, factors = 1) # estimate

# Predict 1, 10, and 100 days ahead:
predobj <- predh(res, ahead = c(1, 10, 100))

# Trace plot of draws from posterior predictive factor log-variance
# (one, ten, and 100 days ahead):
plot.ts(predobj$factorh[1,,])

# Smoothed kernel density estimates of predicted volas:
plot(density(exp(predobj$factorh[1,,"1"]/2), adjust = 2))
lines(density(exp(predobj$factorh[1,,"10"]/2), adjust = 2), col = 2)
lines(density(exp(predobj$factorh[1,,"100"]/2), adjust = 2), col = 3)
# }