Andrew Shohet
Eco 43
Professor Schmidt
Introduction
This is my webpage.
My paper discusses the topic of comparison between fiscal policy and the
corporate dividends. Specifically, I am
answering the question of what is the relationship between the independent variable – the highest bracket of
income tax, and the dependant variable – the dividend yield, and how President
Bush’s plan to cut dividend taxes affected our economy. Prior, to my research, my hypothesis was that
there is an inverse relationship, with a negative slope between these two
variables. After my research, and
regressions, I learned that my hypothesis was wrong and that there is a direct
relationship between my two variables, with a positive slope.
Model
The
relationship between the highest bracket of income tax,
and the dividend yield is a very strong one.
The economic ideas that make me think that this relationship between
these two variables exist is that as the income tax bracket gets higher, the
dividend yield appears to increase as well, and vice versa. My regressions showed me that there was no
need for me to drop any of my variables, and that they were all important. The graph of the relationship between my two
variables looks like this with FSDXP being the dividend yield, and INCTAX being
the highest bracket of income tax:
Data
The
variables that I used for my regression all were macroeconomic data that I got
from the DRI database, with the exception of the highest bracket of income tax
rate. The variables that I used are: The
variables I am using in my paper are FSPCOM, P/E ratio (FSPXE), the Dividend
Yield (FSDXP), the unemployment rate (LHUR), and the highest tax bracket of
income tax (INCTAX). These variables are
measured as follows: the monthly DRI is S&P’s
common stock price index: Composite (1941-43=10), the P/E ratio is the price
divided by earnings ratio (%,NSA), the dividend yield
is S&P’s composite common stock: dividend yield
(% Per Annum), the unemployment rate is all workers unemployed, ages 16 years
and older, and the Income Tax Rate is the highest bracket of income tax
possible.
Results
The standard deviations of all of my variables are as
follows: FSPCOM: 0.0001, FSPEXE (P/E ratio): 0.004495,
INCTAX (Income Tax bracket): 0.001286,
LHUR (unemployment rate) : 0.014343 All of my standard errors are very small
which makes it even more so that all of my variables are important to my
equation. Before running my regression,
my equation was :
Because my t-Statistics were
all so small, I felt there was no need to drop any of my variables because they
are all so important to my equation.
Just to be sure of this, I ran a regression dropping FSPCOM. After doing an F-Test, I was assured of this
because my F-statistic was 6.684 which is greater than the critical value of 1.96 This made it so
that my equation remained the same.
Conclusion
From running my
regressions and carefully analyzing my data set, it was confirmed that all of
my variables are extremely important in my equation and there was no need for
me to drop any of these variables. What
was most important to me was that my original hypothesis that there was a
negative relationship between the dividend yield and the highest bracket of
income tax was wrong, and instead, there is a positive relationship between
these two variables. The parameter on
income tax is positive, not negative and this could explain for why I made such
an error in my original hypothesis. I
also learned that by cutting the dividend tax, President Bush had a tremendous
effect on our country’s economy because he is giving huge incentive for
investors to put their money in the stock market.
Regression
|
Dependent Variable: FSDXP |
||||
|
Method: Least Squares |
||||
|
Date: |
||||
|
Sample(adjusted): 1954:01 2002:12 |
||||
|
Included observations: 588 after
adjusting endpoints |
||||
|
Variable |
Coefficient |
Std.
Error |
t-Statistic |
Prob. |
|
FSPCOM |
-0.000262 |
0.000101 |
-2.585901 |
0.0100 |
|
FSPXE |
-0.129437 |
0.004495 |
-28.79840 |
0.0000 |
|
INCTAX |
0.004637 |
0.001286 |
3.604750 |
0.0003 |
|
LHUR |
0.167664 |
0.014343 |
11.68991 |
0.0000 |
|
C |
4.443096 |
0.165992 |
26.76696 |
0.0000 |
|
R-squared |
0.859226 |
Mean dependent var |
3.455969 |
|
|
Adjusted R-squared |
0.858260 |
S.D. dependent var |
1.129833 |
|
|
S.E. of regression |
0.425364 |
Akaike info criterion |
1.136722 |
|
|
Sum squared resid |
105.4847 |
Schwarz criterion |
1.173940 |
|
|
Log likelihood |
-329.1964 |
F-statistic |
889.5971 |
|
|
Durbin-Watson stat |
0.125982 |
Prob(F-statistic) |
0.000000 |
|