CommonTrend (version 0.7-1)

optimlag: Search the lag order to maximize Johansen Statistics (1988)

Description

Search the lag order to maximize Johansen Statistics (1988) assuming there is only one common trend.

Usage

optimlag(data, k=10, ecdet= "none" )

Arguments

data

Data used to construct the cointegration system

k

maximum number of lags to search

ecdet

Character, the same argument used in ca.jo. In the package urca, ‘none’ is for no intercept in ECT (but there is constant in VECM, just outside ECT), ‘const’ is for constant term in ECT and ‘trend’ is for trend in ECT. See Details.

Value

Olag.value

The maximized Johansen statistics (trace)

Olag

The lag order that maximizes Johansen statistics

list.lags

A list of all the possible lag orders and their corresponding Johansen Statistics. Note that the value when lag order is one is NA since lag order must bigger than 2.

Details

Under the assumption that there is only one common trend, if a lag oder is chosen to maximize the Johansen statistics , then the estimated signle common trend will best fit the original series.

‘none’ is suggested to be chosen for ‘ecdet’, since currently this package only supports extrating common trend when intercept is included into VECM (but outside ECT).

References

Johansen, S. (1988), Statistical Analysis of Cointegration Vectors, Journal of Economic Dynamics and Control, 12, 231--254.

See Also

optimlag-class

Examples

data(benchmark)
x = seq(1, 6689, by = 23)  ## monthly data
global = data.frame(benchmark[x, 2:4])

optimlag(global, 15, "none")