Monte Carlo studies suggest that choosing the
lag order (p) of the unit root tests according to the formula: Int {12(T /100)1/
4} where T is the
number of observations (Harris, 1992). For example sample period chosen is 1985:01-2009:04
where T = 100 and thus, Int {12(100/100)1/4} = 12
Harris, R. I. D. (1992), “Testing for unit roots
using the augmented Dickey-Fuller test: Some issues relating to the size, power
and the lag structure of the test”, Economics
Letters, Vol. 38, No. 4, pp. 381-386.