Table 5.
Cointegrating long-run estimations using the Autoregressive Distributive Lag (ARDL) model.
| Panel A: Variables | Model 1 (1, 0, 3, 1) | Model 2 (2, 2, 1, 1, 3) |
|---|---|---|
| LY | 0.86*** (0.09) | – |
| LRP | −0.26** (0.10) | 0.34* (0.17) |
| LR | −0.015*** (0.004) | – |
| LX | – | 1.01*** (0.23) |
| LFCG | – | −0.96*** (0.29) |
| LI | – | −0.28 (0.22) |
| Trend | 0.02*** (0.01) | 0.02*** (0.01) |
| C |
−1.10 (1.10) |
14.70 (3.72) |
|
Panel B: Cointegration test through the F-Bound test | ||
| Statistics values | 9.36 | 10.54 |
|
Critical values |
I(0) I(1) 5%: 3.16 4.20 1%: 4.43 5.82 |
I(0) I(1) 5%: 4.04 4.51 1%: 5.60 7.17 |
| Diagnostic tests | Observations = 39 R2 = 0.98 DW = 2.08 SER = 0.02 Normality = X2 0.83 (p. 0.66) LM = 0.29 (p. 0.86) Hetero. = 12.40 (p. 0.25) ARCH = 0.09 (p. 0.95) Ramsey-reset = 0.30 (p. 0.58) |
Observations = 38 R2 = 0.99 DW = 1.94 SER = 0.01 Normality = X2 0.49 (p. 0.78) LM = 3.42 (p. 0.33) Hetero. = 13.99 (p. 0.52) ARCH = 5.52 (p. 0.14) Ramsey-reset = 0.40 (p. 0.53) |
Notes: Models run with automated lag selection process in the statistical software (Eviews 10), P = probability.