diagnostic.fGarch: Diagnostic information derived from the estimation

In yzhao7322/CurVol: Curve Data Volatility Analysis

Description

diagnostic.fGarch function provides the estimation parameters that can be used as the inputs for a diagnostic purpose.

Usage

diagnostic.fGarch(params, basis, yd, p = 0, q = 1, xd = NULL)

Arguments

params	List of model paramters: d, As, Bs (and Gs).
basis	The M-dimensional basis functions for the projection.
yd	A grid_point x N matrix drawn from N functional curves.
p	The order of the depedence on past squared observations. If it is missing, p=0.
q	The order of the depedence on past volatilities. If it is missing, q=1.
xd	A grid_point x N matrix drawn from N covariate X curves. The default value is NULL.

Value

List of parameters:

eps: a grid_point x N matrix containing fitted residuals.

sigma2: a grid_point x (N+1) matrix that the first N columns are the fitted conditional variance, the N+1 column is the predicted conditional variance.

yfit: a grid_point x N matrix drawn from N fitted intra-day price curves.

kernel_op: estimated kernel coefficient operators.

References

Rice, G., Wirjanto, T., Zhao, Y. (2020) Forecasting Value at Risk via intra-day return curves. International Journal of Forecasting. <doi:10.1016/j.ijforecast.2019.10.006>.

See Also

est.fArch est.fGarch est.fGarchx gof.fgarch

Examples

## Not run: 
# generate discrete evaluations of the FGARCH process and smooth them into a functional data object.
yd = dgp.fgarch(grid_point=50, N=200, "garch")
yd = yd$garch_mat
fdy = fda::Data2fd(argvals=seq(0,1,len=50),y=yd,fda::create.bspline.basis(nbasis=32))

# extract data-driven basis functions through the truncated FPCA method.
basis_est = basis.est(yd, M=2, "tfpca")$basis

# fit the curve data with an FARCH(1) model.
arch_est = est.fArch(fdy, basis_est)

# get parameters for diagnostic checking.
diag_arch  = diagnostic.fGarch(arch_est, basis_est, yd)

## End(Not run)

yzhao7322/CurVol documentation built on Sept. 5, 2021, 8:41 p.m.