gmm (version 1.1-1)

weightsAndrews2: Kernel weights

Description

Function to compute the kernel weights used to compute the HAC covariance matrix

Usage

weightsAndrews2(x, bw = bwAndrews2, kernel = c("Quadratic Spectral", 
                "Truncated", "Bartlett", "Parzen", "Tukey-Hanning"), approx = c("AR(1)", 
                "ARMA(1,1)"), prewhite = 1, ar.method = "ols", tol = 1e-7, verbose = FALSE)
bwAndrews2(x, kernel = c("Quadratic Spectral", 
          "Truncated", "Bartlett", "Parzen", "Tukey-Hanning"), approx = c("AR(1)", 
          "ARMA(1,1)"), prewhite = 1, ar.method = "ols")  
bwNeweyWest2(x, kernel = c("Bartlett", "Parzen", 
            "Quadratic Spectral", "Truncated", "Tukey-Hanning"), 
             prewhite = 1, ar.method = "ols",...)

Arguments

x

A $n\times q$ matrix of time series from which we want to compute the covariance matrix.

bw

The method to compute the bandwidth parameter. For now, bwAndrews2 is the only one possible. I leave the option there because I am planning to give more choices in futur versions of the package.

prewhite

logical or integer. Should the estimating functions be prewhitened? If TRUE or greater than 0 a VAR model of order as.integer(prewhite) is fitted via ar with method "ols" and demean = F

ar.method

character. The method argument passed to ar for prewhitening.

verbose

logical. Should the bandwidth parameter used be printed?

approx

a character specifying the approximation method if the bandwidth has to be chosen by bwAndrews2.

tol

numeric. Weights that exceed tol are used for computing the covariance matrix, all other weights are treated as 0.

kernel

The choice of kernel

...

It just allows to call either bwAndrews2 or bwNeweyWest without having unusued arguments.

Value

• weightsAndrews returns a vector of weights.

  bwAndrews returns the selected bandwidth parameter.

Details

weightsAndrews2, bwAndrews2 and bwNeweyWest2 are simply modified version of weightsAndrews, bwAndrews and bwNeweyWest from the package sandwich. The modifications have been made so that the argument x can be a matrix instead of an object of class lm or glm. The details on how is works can be found on the sandwich manual. kweights is the same as the one included in the package sandwich.

References

Zeileis A (2006), Object-oriented Computation of Sandwich Estimators. Journal of Statistical Software, 16(9), 1--16. URL http://www.jstatsoft.org/v16/i09/.

Andrews DWK (1991), Heteroskedasticity and Autocorrelation Consistent Covariance Matrix Estimation. Econometrica, 59, 817--858.

Newey WK & West KD (1987), A Simple, Positive Semi-Definite, Heteroskedasticity and Autocorrelation Consistent Covariance Matrix. Econometrica, 55, 703--708.

Newey WK & West KD (1994), Automatic Lag Selection in Covariance Matrix Estimation. Review of Economic Studies, 61, 631-653.

Examples

set.seed(345)
x <- arima.sim(n=200,list(ordre=c(1,0,1),ar=.9,ma=.4))
y <- 2*x + rnorm(200)
x = cbind(x,y)
w <- weightsAndrews2(x, bw = bwAndrews2, kernel = "Quadratic")
plot(w,type='l')
w2 <- weightsAndrews2(x, bw = bwNeweyWest2, kernel = "Bartlett")
plot(w2,type='l')