SVARIV: SvAR-IV.

In martinbaumgaertner/varexternal: Vector autoregression with external instruments

SvAR-IV.

Description

Implements standard and weak-IV robust SVAR-IV inference.

Usage

SVARIV(
  ydata,
  z,
  p,
  confidence,
  NWlags,
  norm,
  scale,
  horizons,
  ci_type = c("msw"),
  print_wald = T,
  instrument_name
)

Arguments

ydata	Endogenous variables from the VAR model
z	External instrumental variable
p	Number of lags in the VAR model
confidence	Value for the standard and weak-IV robust confidence set
NWlags	Newey-West lags (set it to 0 to compute heteroskedasticity robust std errors)
norm	Variable used for normalization
scale	Scale of the shock
horizons	Number of horizons for the Impulse Response Functions (does not include the impact or horizon 0)
ci_type	confidence intervals to include (choose from "msw" (Montiel, Stock and Watson, 2020), "delta" or "plugin"), just triggers msw waldtest. specify ci by prett.irf command
print_wald	Number of horizons for the Impulse Response Functions (does not include the impact or horizon 0)
instrument_name	Number of horizons for the Impulse Response Functions (does not include the impact or horizon 0)

Value

irfs: list containing all irf data

waldstat: contain msw waldstat

Examples

p           = 24    #Number of lags in the VAR model
NWlags      = 0;  # Newey-West lags(if it is neccessary to account for time series autocorrelation)
norm        = 1; # Variable used for normalization
scale       = 1; # Scale of the shock
horizons    = 20; #Number of horizons for the Impulse Response Functions(IRFs)
confidence=c(0.6,0.9,0.95);
data(oil)
colnames(oil)<-c("a","b","c","d","year","month")
ydata<-oil[,1:3]
z<-oil[,4]
VAR<-SVARIV(ydata,z,p,confidence,NWlags,norm,scale,horizons,instrument_name="test")

martinbaumgaertner/varexternal documentation built on April 27, 2022, 1:31 a.m.