The determinants of money demand in China

Figures & data

Table 1. Unit root tests

Variable	Test type	ADF	Results
(M/P)	(0, c, 1)	2.9962	Nonstationary
△(M/P)	(t, c, 1)	−3.5189*	Stationary
Y	(t, c, 1)	1.0641	Nonstationary
△Y	(0, c, 7)	−2.4573*	Stationary
I	(0, c, 1)	−0.0569	Nonstationary
△I	(0, c, 1)	−20.705***	Stationary
π^e	(t, c, 1)	−2.6904	Nonstationary
△π^e	(0, c, 1)	−7.9487***	Stationary
B	(t, c, 1)	−2.2375	Nonstationary
△B	(0, c, 3)	−3.1226**	Stationary
E^e	(0, c, 1)	−1.8592	Nonstationary
△E^e	(0, c, 8)	−3.9196**	Stationary
FIE	(t, c, 1)	−2.1979	Nonstationary
△FIE	(t, c, 7)	−33.026***	Stationary

Notes: ADF denotes Augmented Dickey–Fuller test. *, ** and *** indicate that the variable is statistically significant at the 10%, 5% and 1% levels, respectively. In the (t, c, l) of test type, t, c and l stand for Trend, Intercept and Lagged differences, respectively. Δ stands for the first-order difference.

Table 2. Johansen’s cointegration test for M/P, Y, I, π^e, B, E^e and FIE

Hypothesized no. of CE(s)	Eigenvalue	Likelihood rate	5% Critical value	1% Critical value
ρ = 0**	0.999708	365.6441	109.99	119.80
ρ ≤ 1**	0.991122	202.8790	82.49	90.45
ρ ≤ 2**	0.884291	108.3964	59.46	66.52
ρ ≤ 3**	0.748885	65.26298	39.89	45.58
ρ ≤ 4**	0.692419	37.62612	24.31	29.75
ρ ≤ 5*	0.401491	14.04582	12.53	16.31
ρ ≤ 6	0.172195	3.779543	3.84	6.51

Notes: The cointegration tests are conducted by the maximum likelihood procedure with a maximum lag of 0 and a test assumption of linear deterministic trend in the data. *(**) denotes rejection of the hypothesis at 5% (1%) significance level.

Table 3. Pairwise Granger causality tests

Null hypothesis	Obs.	F-statistic	Prob.
Y does not Granger cause M/P	16	5.20988	0.04709
M/P does not Granger cause Y		21.4247	0.00217
I does not Granger cause M/P	16	5.77095	0.03854
M/P does not Granger cause I		1.66426	0.29488
π^e does not Granger cause M/P	19	3.01951	0.08124
M/P does not Granger cause π^e		3.68546	0.05178
B does not Granger cause M/P	16	2.76304	0.14448
M/P does not Granger cause B		4.05096	0.07544
E^e does not Granger cause M/P	18	3.50514	0.05301
M/P does not Granger cause E^e		3.58036	0.05025
FIE does not Granger cause M/P	19	4.95379	0.02361
M/P does not Granger cause FIE		7.95531	0.00492

Table 4. The OLS estimation results of money demand function

Variable	Eq. 1	Eq. 2	Eq. 3	Eq. 4
Const.	1151.198	−3963.857	−2874.087	−4562.213
	(3.42***)	(3.29**)	(2.33**)	(2.11**)
Y	0.004061	−0.000862	−0.001509	−0.002212
	(37.64***)	(0.77)	(1.38)	(1.67)
I	−162.0770	622.1359	602.9495	786.6032
	(3.31***)	(3.37***)	(3.58***)	(3.07***)
π^e	−5027.313	−24749.34	−26446.94	−30952.09
	(2.65**)	(5.22***)	(5.99***)	(4.78**)
B		0.046556	0.048146	0.056783
		(4.32***)	(4.88***)	(4.32***)
E^e			−1.084193	25.97883
			(1.95**)	(0.91)
FIE				−26.32337
				(0.95)
R^2	0.99	0.99	0.99	0.99
F	1348.41	2210.45	2124.66	1757.87
D.W.	1.68	1.27	1.27	1.65

Notes: All equations were run using the OLS technique. The numbers in parentheses below the estimated coefficients indicate t-values. R² = coefficient of determination; F= F-statistic; D.W. = Durbin–Watson statistic; other symbols are defined in the text. *, ** and *** indicate statistical significance at the 10%, 5% and 1% levels, respectively.

Figure 1. Impulse response function.

Figure 2. Decomposition of variance graphic.