# predh: Predicts factor and idiosyncratic log-volatilities h In factorstochvol: Bayesian Estimation of (Sparse) Latent Factor Stochastic Volatility Models

## Description

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

## Usage

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

## 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.

## Value

List of class `fsvpredh` containing two elements:

• idihArray containing the draws of the latent idiosyncratic log-volatilities.

• factorhArray containing the draws of the latent factor log-volatilities.

Other predictors: `predcond`, `predcor`, `predcov`, `predloglikWB`, `predloglik`, `predprecWB`
 ``` 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15``` ```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) ```