Table 8

ARDL estimation results with structural break

CONSSEASEFDILDSPCPIBREAK
Panel A: ARDL (1, 4, 4, 3, 4, 1, 1) long-run coefficient estimates – DV: LnTRC
14.419 (1.618)0.567** (4.08)−3.192** (−3.07)−0.252** (−2.46)0.895* (1.38)−0.293** (−4.72)−1.01 (−0.80)
Lag order0123
Panel B: ARDL (1, 4, 4, 3, 4, 1, 1) short-run coefficient estimates – DV: ∆LnTRC
∆SEA−0.260 (−0.38)−0.278 (−0.48)−0.122 (−0.21)−3.473 (−5.44) **
∆SE2.303 (4.64)**−1.262 (−1.94) **−1.831 (−2.74) **2.609 (4.20) **
∆FDI−0.305 (7.48) **−0.150 (−4.24) **−0.088 (−3.14) **
∆LDSP0.797 (5.28)**−0.001 (−0.01)−0.062 (−0.68)0.568 (6.26) **
∆CPI−0.162 (−6.19)**
BREAK0.272 (2.67)**
ECTt-1χ2SCχ2HETχ2Nχ2FFAdj. R2
Panel C: Diagnostic statistics tests
−0.83 (−9.85)**2.82 [0.11]0.26 [0.90]2.23 [0.33]0.89 [0.39]0.94

Note(s): The model is estimated by setting the maximum lag to 4, and the optimum lag length is suggested based on AIC. ∆ is the first difference operator. Asterisk (** and *) denote significance at 5% and 10% levels, respectively. The values in parentheses “(.)” in Panels A and B are the t-ratios, and values in parentheses “[.]” are the probability values of the LM test statistics. χ2SC, χ2HET, χ2N and χ2FF denote LM tests for serial correlation, heteroscedasticity, normality and functional form, respectively. DV refers to the dependent variable

Source(s): Author’s estimation from data

or Create an Account

Close Modal
Close Modal