In: Statistics and Probability
Estimate a multiple linear regression relationship with the U.K. stock returns as the dependent variable, and U.K. Corporate Bond yield (Interest rate), U.S. Stock Returns, and Japan Stock Returns as the independent variables using the monthly data covering the sample period 1980-2017 (Finding the determinants of U.K. stock returns).
Show the estimated regression relationship
Conduct a t-test for statistical significance of the individual slope coefficients at the 1% level of significance. Provide the interpretation of the significant slope estimates.
Conduct a test for the overall significance of the regression equation at the 1% level of significance. (Test for the significance of the regression relationship as a whole)
Present the R-Square (Coefficient of Determination) and its interpretation.
Solution:
First, we need to obtain the Excel print out using the data in Table GFD Final Monthly Returns.
1% level of significance in both t tests and the F test.
SUMMARY OUTPUT |
|||||
Regression Statistics |
|||||
Multiple R |
0.735992889 |
||||
R Square |
0.541685532 |
||||
Adjusted R Square |
0.538636877 |
||||
Standard Error |
3.523412257 |
||||
Observations |
455 |
||||
ANOVA |
|||||
df |
SS |
MS |
F |
Significance F |
|
Regression |
3 |
6617.396126 |
2205.798709 |
177.6801681 |
4.89572E-76 |
Residual |
451 |
5598.909705 |
12.41443393 |
||
Total |
454 |
12216.30583 |
Coefficients |
Standard Error |
t Stat |
P-value |
Lower 95% |
Upper 95% |
Lower 99.0% |
Upper 99.0% |
|
Intercept |
0.115186435 |
0.17010687 |
0.677141581 |
0.498663306 |
-0.219114035 |
0.449486905 |
-0.324841646 |
0.555214515 |
RSUS |
0.730332455 |
0.041983928 |
17.39552475 |
3.10678E-52 |
0.647824048 |
0.812840861 |
0.621729509 |
0.8389354 |
RSJA |
0.209729348 |
0.029769306 |
7.045154189 |
6.96789E-12 |
0.15122558 |
0.268233116 |
0.132722873 |
0.286735823 |
RUK |
0.113600031 |
0.030384391 |
3.738762846 |
0.000208723 |
0.053887475 |
0.173312588 |
0.03500247 |
0.192197593 |
The regression equation:
Regression equation dependent variables are linearly relationship with an output variables.
For every additonal RSUS, the Uk stock value increases by 0.730332455 units
Fore very additonal unit of RSJA value, the UK stock value increases by 0.209729348 units
For every additonal unit of RUK value, tha UK stock value increases by 0.113600031 units.
b) T-test for statistical significane for slope:
and
the degree of freedom: DFresidual= 451
The test statistic:
RSUS test statistic: 17.39552475 and P-value: 0.0000
The test statistic is significant and rejects H0. There is sufficient evidence to support that there is a significant relationship with output variable.
RSJA test statistic: 7.045154189 and P-value: 0.0000
The test statistic is significant and rejects H0. There is sufficient evidence to support that there is a significant relationship with output variable.
RUK test statistic: 3.738762846 and P-value: 0.0002
The test statistic is significant and rejects H0. There is sufficient evidence to support that there is a significant relationship with output variable.
c) Overall significane test:
Atleast one is not equal
df1= DFRegression=3
df2= DFresidual=451
The test statistic:
P-value: 0.000
The test statistic is significant and rejects H0. There is sufficient evidence to support that there is significant relationship between input and output variables.
R-squared value:
The proporiton of variation explained by regression equation is 0.541685532.