adf_lag_select: Identify Optimal Lag Order Selection for (Augmented)...

In sTSD: Simulate Time Series Diagnostics

Identify Optimal Lag Order Selection for (Augmented) Dickey-Fuller Tests

Description

adf_lag_select() runs a series of (Augmented) Dickey-Fuller tests and returns information that may (or may not) be useful in identify a potential lag order for unit root tests.

Usage

adf_lag_select(x, min_lag = 0, max_lag = NULL)

Arguments

x	a vector
min_lag	the minimum lag order to use. Defaults to 0.
max_lag	the maximum lag order to use. Defaults to Schwert's (1989) upper lag.

Details

This function removes missing values from the vector before calculating test statistics.

The lower bound lag order suggested by Schwert (1989) and the default suggested by Said and Dickey (1984) do not meaningfully separate from each other until the length of the series reaches 127. Under those conditions, if the note column returned by this function for a finite series does not identify the Said and Dickey (1984) default, but identifies the Schwert (1989) lower bound, interpret the latter as the former.

Value

adf_lag_select() returns a list of length 3. The first element in the list is a data frame of a series of (Augmented) Dickey-Fuller tests for no-drift, no-trend. The second is a data frame of a series of (Augmented) Dickey-Fuller tests for drift, no trend. The third is a data frame of a series of (Augmented) Dickey-Fuller tests for a drift and trend. Each data frame has the following columns communicating the following information.

1. The lag order

2. The (A)DF statistic for the lag order.

3. The Akaike information criterion for the model.

4. Schwartz' (Bayesian) criteron for the model.

5. The absolute value of the last lagged first difference in the model.

6. The "modified" Akaike information criterion for the model.

7. The "modified" Schwarz' (Bayesian) criterion for the model.

8. A note indicating if the lag was suggested by Schwert (1989) or Said and Dickey (1984)

Author(s)

Steven V. Miller

Examples

x <- head(tbills$tb3m, 500)
adf_lag_select(x)

sTSD documentation built on April 3, 2025, 7:37 p.m.