Question

TABLE 16-4                                                                                                                                                                                  Given below are EXCEL outputs for various estimated autoregressive models for Coca-Cola's real operating revenues (in billions of dollars) from 1975 to 1998.  From the data, we also know that the real operating revenues for 1996, 1997, and 1998 are 11.7909, 11.7757 and, 11.5537, respectively.                                                                                                                                                                                                                         AR(1) Model:

	Coefficients	Standard Error	t Stat	P-value
Intercept	0.1802077	0.39797154	0.452815546	0.655325119
XLag1	1.011222533	0.049685158	20.35260757	2.64373E-15

AR(2) Model:

	Coefficients	Standard Error	t Stat	P-value
Intercept	0.30047473	0.4407641	0.681713257	0.503646149
X Lag 1	1.17322186	0.234737881	4.998008229	7.98541
X Lag 2	-0.183028189	0.250716669	-0.730020026	0.47428E-08

AR(3) Model:

	Coefficients	Standard Error	t Stat	P-value
Intercept	0.313043288	0.514437257	0.608515972	0.550890271
XLag1	1.173719587	0.246490594	4.761721601	0.000180926
XLag2	-0.069378567	0.373086508	-0.185958391	0.854678245
XLag3	-0.122123515	0.282031297	-0.433014053	0.670448392

Referring to Table 16-4 and using a 5% level of significance, what is the appropriate AR model for Coca-Cola's real operating revenue?

Question 16 options:

	AR(1)
	AR(2)
	linear
	AR(3)

Answer 1

P-value for XLag1 in AR(1) model is 2.64373E-15

Since, p-value is less than 0.05 significance level, we conclude that AR(1) model is significant.

P-value for XLag2 in AR(2) model is greater than 0.05 significance level, we conclude that AR(2) model is not significant. (Note that p-value of XLag2 in AR(2) model is incorrect. It should be a higher value for t-stat = -0.730020026)

P-value for XLag2 and XLag3 in AR(3) model is greater than 0.05 significance level, we conclude that AR(3) model is not significant.

The appropriate AR model for Coca-Cola's real operating revenue would be the significant model

AR(1)