inference.bootstrap: Bootstrap inference for event study estimator

In nipfpmf/eventstudies: Event Study Analysis

Description

This function obtains a boostrapped confidence interval for estimates of magnitude over the event horizon.

Usage

inference.bootstrap(es.w,
                   to.plot = TRUE,
                   boot.run = 1000,
                   xlab = "Event time", 
		   ylab = "Cumulative returns of response series", 
		   main = "Event study plot")

Arguments

es.w	a zoo object indexed by event time: the “z.e” component of the list returned by the phys2eventtime function. The object should consist of more than one series.
boot.run	A ‘numeric’, controlling the number of simulations required for the bootstrap.
to.plot	a ‘logical’ indicating whether to generate an event study plot of the inference estimated. Defaults to ‘TRUE’.
xlab	the x-axis label of the generated plot. Used if “to.plot” is ‘TRUE’.
ylab	the y-axis label of the generated plot. Used if “to.plot” is ‘TRUE’.
main	main title of the plot. Used if “to.plot” is ‘TRUE’.

Value

A ‘matrix’ with 3 columns, the lower confidence interval (CI), the mean, and the upper CI which are the result of bootstrap inference.

Author(s)

Vikram Bahure, Vimal Balasubramaniam

See Also

boot phys2eventtime inference.wilcox inference.classic

Examples

data(StockPriceReturns)
data(SplitDates)

es.results <- phys2eventtime(z = StockPriceReturns,
                             events = SplitDates,
                             width = 5)
es.w <- window(es.results$z.e,
               start = -5,
               end = +5)

eventtime <- remap.cumsum(es.w, is.pc = FALSE, base = 0)
inference.bootstrap(es.w = eventtime,
                    to.plot = FALSE)

nipfpmf/eventstudies documentation built on June 7, 2020, 3:57 p.m.