Financial Innovation

Table 1 Stationarity tests

From: The role of R&D and economic policy uncertainty in Sri Lanka’s economic growth

Perron (1989) test
Variable	t-statistic	TB1	Lags	Status
lnA	− 2.332	2001	0	I(1)
lnRD	− 1.861	2000	0	I(1)
lnPATENT	− 5.124	1994	1	I(0)
lnRD/L	− 1.731	2000	0	I(1)
lnRD/Y	− 2.603	2000	0	I(1)
lnRD/AL	− 2.053	2000	2	I(1)
lnPATENT/L	− 4.819	1994	1	I(0)
lnPATENT/Y	− 4.598	1994	2	I(0)
lnPATENT/AL	− 5.162	1994	1	I(0)

Variable	Narayan–Popp (2010) test
	M1						M2
	Test statistic	TB1	TB2	k	Status	Test statistic	TB1	TB2	k	Status
lnA	− 3.119	1996	2000	0	I(1)	− 1.709	1996	2001	0	I(1)
lnRD	− 5.031	1999	2009	0	I(0)	− 4.192	1999	2009	0	I(1)
lnPATENT	− 5.630	1993	1998	2	I(0)	− 4.040	1993	1996	2	I(1)
lnRD/L	− 4.314	1999	2009	0	I(0)	− 3.314	1999	2009	0	I(1)
lnRD/Y	− 5.708	1999	2009	0	I(0)	− 4.617	1999	2009	0	I(1)
lnRD/AL	− 2.970	2000	2009	0	I(1)	− 4.162	1999	2009	0	I(1)
lnPATENT/L	− 4.075	1993	2002	1	I(1)	− 4.564	1994	2002	2	I(1)
lnPATENT/Y	− 5.287	1993	1998	2	I(0)	− 3.775	1993	2002	2	I(1)
lnPATENT/AL	− 3.247	1994	1996	2	I(1)	− 1.664	1994	2002	2	I(1)

The table shows the results of the stationarity tests. We compared the Perron (1989) t-statistic, and Narayan and Popp (2010) M1 and M2 statistics to the critical values reported in their respective studies. We selected the lags in the Perron test automatically using the Schwarz information criterion (SIC). The lags in the Narayan and Popp (2010) test are chosen using Hall’s (1994) procedure. Perron’s (1989) test accommodates a single structural break, while Narayan and Popp’s (2010) test accommodates two structural breaks. Only the intercept term is included in the test regressions. k, TB1, and TB2 are the optimal lags, the first, and second structural break dates, respectively. Our sample is from 1980 to 2018

Back to article page