STADF.test: Supremum ADF tests with time transformation

View source: R/STADF.test.R

STADF.testR Documentation

Supremum ADF tests with time transformation

Description

See SADF.test. Tests with time transformation are the modified versions of the ordinary SADF and GSADF tests using Nadaraya-Watson residuals and reindexing procedure by Cavaliere-Taylor (2008).

Usage

STADF.test(
  y,
  trim = 0.01 + 1.8/sqrt(length(y)),
  const = FALSE,
  omega.est = TRUE,
  truncated = TRUE,
  is.reindex = TRUE,
  ksi.input = "auto",
  hc = 1,
  pc = 1,
  add.p.value = TRUE
)

GSTADF.test(
  y,
  trim = 0.01 + 1.8/sqrt(length(y)),
  const = FALSE,
  omega.est = TRUE,
  truncated = TRUE,
  is.reindex = TRUE,
  ksi.input = "auto",
  hc = 1,
  pc = 1,
  add.p.value = TRUE
)

Arguments

y

A time series of interest.

trim

A trimming parameter to determine the lower and upper bounds for a possible break point.

const

Whether the constant needs to be included.

omega.est

Whether the variance of Nadaraya-Watson residuals should be used.

truncated

Whether the truncation of Nadaraya-Watson residuals is needed.

is.reindex

Whether the Cavaliere and Taylor (2008) time transformation is needed.

ksi.input

The value of the truncation parameter. Can be either auto or the explicit numerical value. In the former case the numeric value is estimated.

hc

The scaling parameter for Nadaraya-Watson bandwidth.

pc

The scaling parameter for the estimated truncation parameter value.

add.p.value

Whether the p-value is to be returned. This argument is needed to suppress the calculation of p-values during the precalculation of tables needed for the p-values estimating.

Details

Refactored original code by Kurozumi et al.

Value

An object of type sadf. It's a list of:

  • y,

  • N: Number of observations,

  • trim,

  • const,

  • omega.est,

  • truncated,

  • is.reindex,

  • new.index: the vector of new indices,

  • ksi.input,

  • hc,

  • h.est,

  • u.hat,

  • pc,

  • w.sq,

  • t.values: vector of t-values,

  • the value of the corresponding test statistic,

  • u.hat.truncated: truncated residuals if truncation was asked for,

  • ksi, sigma: estimated values of the truncation parameter and resulting s.e. if ksi.input equals auto,

  • eta.hat: the values of reindexing function if reindexing was asked for,

  • p-value if it was asked for.

References

Cavaliere, Giuseppe, and A. M. Robert Taylor. “Time-Transformed Unit Root Tests for Models with Non-Stationary Volatility.” Journal of Time Series Analysis 29, no. 2 (March 2008): 300–330. https://doi.org/10.1111/j.1467-9892.2007.00557.x.

Kurozumi, Eiji, Anton Skrobotov, and Alexey Tsarev. “Time-Transformed Test for Bubbles under Non-Stationary Volatility.” Journal of Financial Econometrics, April 23, 2022. https://doi.org/10.1093/jjfinec/nbac004.


d9d6ka/RANEPA-R documentation built on May 4, 2024, 7:11 a.m.