Conference – 1992 Inflation, Indicators and Monetary Policy

Table 3: Inflation Indicators: Sums of Coefficients Inline Equation
(Sample period: 1967Q4–1991Q4)
Constant 0.004 (2.1) 0.004 (2.1) 0.008 (3.2) 0.003 (1.8) 0.004 (2.0) 0.005 (2.5) −0.007 (1.8) 0.006 (2.2) 0.006 (2.8) 0.005 (2.1) 0.011 (3.5)
Lagged Δ p 0.815 (9.0;.00) 0.833 (9.9;.00) 0.335 (2.0;.04) 0.848 (9.6;.00) 0.79 (6.3;.00) 0.63 (5.3;.00) 0.565 (5.8;.00) 0.695 (5.4;.00) 0.779 (8.0;.00) 0.750 (7.1;.00) 0.737 (7.4;.00) 0.719 (7.2;.00) 0.625 (6.1;.00)
Output gap   0.273 (3.6;.01)                      
Wage rate     0.278 (3.4;.02)                    
Terms of trade       0.125 (2.6;.26)   0.189 (3.7;.02)              
Import price deflator         0.023 (0.4;.71) 0.176 (2.7;.06)              
Currency             0.525 (3.0;.00)            
Broad money               0.191 (2.0;.00)          
API credit                 0.097 (2.1;.00)        
Aggregate asset prices (nominal)                   −0.027 (0.7;.22)      
Stock prices (nominal)                     −0.021 (1.5;.09)    
House prices (nominal)                       0.059 (1.2;.34)  
Comm. property prices (nominal)                         −0.054 (2.0;.0.20)
Inline Equation 0.54 0.567 0.48 0.56 0.52 0.56 0.57 0.08 0.14 0.44 0.47 0.47 0.48
DW 2.02 1.94 2.07 2.02 2.03 2.03 2.03 1.89 2.01 2.13 2.09 2.05 2.18
SEE 0.007 0.007 0.007 0.007 0.007 0.007 0.007 0.005 0.005 0.007 0.007 0.007 0.007
Χ2 struct stability 84Q1–89Q4 4.3 11.2 7.7 5.9 24.9** 16.65** 15.78** 12.01* 8.0 36.9** 14.8**
Note: The sum of the lagged coefficients are shown. All standard errors are White (1980)-corrected for heteroskedasticity. Where six lags were insufficient to eliminate serial correlation (identified by Lagrange multiplier tests), Newey-West standard errors are reported. The numbers in parenthesis are: first, the t-valuefor the null that the sum of the coefficients is zero; second, the marginal significance level for Χ2 tests of the null that all coefficients are zero.
** Reject null hypothesis of structural stability at the 5 per cent significance level.
* Reject null hypothesis of structural stability at the 10 per cent significance level.
† This is the sample period (after allowing for lags) over which most of the models are estimated. Data limitations force exceptions to this which include broad money and AFI credit models, estimated from 1977Q4; stock price and aggregate asset price models, estimated from 1971Q2; house price and commercial property price models, estimated from 1969Q4; the wage model, estimated from 1970Q4 and the import price deflator model, estimated from 1967Q3.