| Library: | finance |
| See also: | kpss rvlm lo robwhittle roblm lobrob gph sclmtest neweywest |
| Macro: | kpssnum | |
| Description: |
Calculation of the KPSS statistics for I(0) against long-memory
alternatives. We consider the two tests, denoted by KPSS_mu and KPSS_t,
and respectively based on a regression on a constant mu, and
on a constant and a time trend t.
The quantlet returns the value of the estimated statistic for two
type of the tests, i.e., const or trend and
the critical values for a 95 percent confidence interval for I(0) (const, trend).
|
| Usage: | (kp, critvalue) = kpssnum(y,T) | |
| Input: | ||
| y | vector | |
| T | scalar, truncation lag for the autocorrelation consistent variance estimator | |
| Output: | ||
| kp | vector | |
| critvalue | vector | |
;Non-stationarity test on the series dmus58.dat.
library("times")
x = read("dmus58.dat") ; 25477 obs: log FX rates
x = x[1:1000]
kpssnum(x,0) ;
| Library: | finance |
| See also: | kpss rvlm lo robwhittle roblm lobrob gph sclmtest neweywest |