We conclude 3 cointegration vectors at 5% level of significance. The normalised
cointegration vectors are presented in table 6.
Table 6. Normalized cointegrating coefficients
|
lnPPI |
lnIPI |
InXREURO |
lnM1A |
TREND |
|
1.000000 |
0.000000 |
0.000000 |
0.100722 |
0.000237 |
|
(0.40240)a |
(0.00539) | |||
|
0.000000 |
1.000000 |
0.000000 |
0.432500 |
-0.003577 |
|
(0.12665) |
(0.00170) | |||
|
0.000000 |
0.000000 |
1.000000 |
-0.648281 |
0.008627 |
|
(0.12772) |
(0.00171) |
a standard errors in parentheses
The money slope coefficients are rather surprisingly negative for the industrial and
agricultural prices and positive for the exchange rate equation, not being statistically
significant in the agricultural price equation. The linear trend is significant in the industrial
prices and exchange rate equations, but not in the agricultural prices equation.
The money neutrality hypothesis expects the coefficients associated with the money supply
(lnM1A) to be close to one (i.e. the long run increase in the agricultural, industrial and
services prices to be unit proportional with the increase in the money supply). The lnM1A
coefficients with respect to the prices are 0.100, 0.432, -0.648, not supporting the money
neutrality hypothesis.
4.3. VECM model
Because the variables proved to be cointegrated, a Vector Error Correction Model is
appropriate to simultaneously depict the long and short run evolution of the system. The
residuals of the long run cointegrating equations are used to construct the VECM in table 7.
12