Friday, 24 June 2011

Comparing inflation to term deposits

1 yr inf D DX
6% 3% 1 0.026549
4% 2% 1 0.021084
6% 4% 1 0.036852
5% 3% 1 0.029582
5% 3% 1 0.032537
5% 3% 1 0.027986
5% 3% 1 0.02591
4% 2% 1 0.023656
3% 3% 1 0.030281
5% 3% 1 0.031226
5% 6% 1 0.058018
4% 2% 1 0.018048
4% 2% 1 0.015833
6% 0% 1 -0.00249
7% 2% 1 0.01519
8% 5% 1 0.050532
5% 3% 1 0.025455
1993 6% 2% 1 0.019462
7% 0% 0 0
11% 2% 0 0
16% 7% 0 0
13% 8% 0 0
11% 8% 0 0
14% 7% 0 0
14% 10% 0 0
12% 8% 0 0
11% 3% 0 0
12% 9% 0 0

Regfession equation:       y=1.02x+0.0367

Assumptions:                    Classic Normal Linear Regression Model

Null hypothesis:                The population slope coefficient is zero

Alternate hypothesis:       The population slope coefficient is not zero

Test statistic:                      5.611895

P value:                               6.74E-06

Interpretation:                  The t-value is statistically significant. There is overwhelming statistical evidence that a relationship exist(positive or negative) between inflation and the ensuing one year term deposit rate.

Coefficient of det:           54.778%

Structural change in the model

Assumptions:                    Classic Normal Linear Regression Model.

Null hypothesis:                The population “Dummy x Inflation” coefficient is zero.

Alternate hypothesis:      The population “Dummy x Inflation” coefficient is not zero.

Test statistic:                     -1.01758

P value:                               0.319027

Interpretation:                  The coefficient is statistically insignificant. This inability to reject the null hypothesis renders us unable to conclude that there has been a structural change in the slope coefficient since the inflation target began in 1993. This seems strange since one would think that the pro-cyclical movement of the cash rate would force interest rates to be even closer related to inflation. In fact, the relationship loosened. The adjusted coefficient of determination was 46.9% before the inflation target and -0.85% (0) after.

No comments:

Post a Comment