Group Project No3

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 4

The University of Oklahoma

Price College of Business


B-AD 5122 Quantitative Analysis I

Group Project No. 3

You must submit your answers to the project questions via Canvas. Each group submits only one
document. To do so, you must put the answers in one PDF file and submit it before the due date.

Before you begin, please list the name of all group members in the space provided below:

You must use R for this group project. Make use of the return dataset in wooldridge library in order to
address the following questions. Before you address those questions, however, you must make sure that
you call the wooldridge library as part of your code’s preamble by typing: library(wooldridge). Also, call
the tidyverse library by typing: library(tidyverse). Then, download the return data on your device using
the following command: data(return, package="wooldridge"). Lastly, put the downloaded data into a, so
called, tibble (a tidyverse tool) by typing: as_tibble(return). If these sound foreign to you, please review
the pre-recorded R demo in Module 5.
a. Estimate the parameters  (the intercept) and  (the slope) in the model below:
𝑅𝑂𝐸 = 𝛼 + 𝛽𝐷𝐾𝑅 + 𝜀
where ROE is the return on equity, DKR is the debt to capital ratio, and 𝜀 is an error term. Keep in
mind that none of these variables should be put into natural log. Just use them as they are without
putting them into log. Copy and paste the summary of your estimation results below.
Note: This part is worth 15 points.
CODE RESULT

Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 21.59352 1.48128 14.578 < 2e-16 ***
dkr -0.15059 0.04848 -3.107 0.00229 **
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

α = 21.59352
β = -0.15059
b. Report the 95% confidence interval estimation for the slope parameter: .
Note: This part is worth 5 points.
CODE RESULT

2.5 % 97.5 %
-0.24643435 -0.05475465

Finding: At 95% level of confidence, the sample evidence suggest that  is greater than -0.24643435
yet less than -0.05475465
Implication: A 1% increase in the the debt to capital ratio is associated with btw. -0.24643435%
and -0.05475465% increase in the return on equity.
c. Test for the following null hypothesis at 95% level of confidence (also referred to as 5% level of
significance): H0: =0.
Note: This part is worth 5 points.
CODE RESULT

Residual standard error: 9.595 on 140 degrees of freedom


Multiple R-squared: 0.06449, Adjusted R-squared: 0.05781
F-statistic: 9.651 on 1 and 140 DF, p-value: 0.002292

As the p-value is much less than 0.05, we reject the null hypothesis that β = 0.
d. Given your findings in parts a, b, and c, how does an increase in debt to capital ratio affect the return
on equity?
Note: This part is worth 10 points.
There is a negative linear relationship between DKR and ROE.
For each one unit increase in DKR, the expected change in ROE is equal to the estimated slope
coefficient (β) which is -0.15059 (decrease of 0.15059 units).
e. Illustrate your answer to part d using a scatterplot in which the debt to capital ratio (DKR) is put on
the horizontal axis and return on equity (ROE) is put on the vertical axis. Make sure that you include
the linear fit in your scatterplot.
Note: This part is worth 10 points; 5 points for the scatterplot and 5 points for the linear fit.
f. Let us now consider a different model. Estimate the parameters  (the intercept) and  (the slope) in
the model below:
𝑅𝑂𝐸 = 𝛼 + 𝛽𝐸𝑃𝑆 + 𝜀
where ROE is the return on equity, EPS is the earnings per share, and 𝜀 is an error term. Keep in mind
that none of these variables should be put into natural log. Just use them as they are without putting
them into log. Copy and paste the summary of your estimation results below.
Note: This part is worth 15 points.

CODE RESULT

Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 17.49038 0.71180 24.572 < 2e-16 ***
eps 0.11913 0.01653 7.205 3.28e-11 ***
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

>>> α = 17.49038
β = 0.11913
g. Report the 95% confidence interval estimation for the slope parameter: .
Note: This part is worth 5 points.

CODE RESULT

2.5 % 97.5 %
0.08643949 0.15181802

Finding: At 95% level of confidence, the sample evidence suggest that  is greater than 0.08643949
yet less than 0.15181802
Implication: A 1% increase in the the earnings per share is associated with btw. 0.08643949% and
0.15181802% increase in the return on equity.
h. Test for the following null hypothesis at 95% level of confidence (also referred to as 5% level of
significance): H0: =0
Note: This part is worth 5 points.

CODE RESULT
Residual standard error: 8.473 on 140 degrees of freedom
Multiple R-squared: 0.2705, Adjusted R-squared: 0.2653
F-statistic: 51.91 on 1 and 140 DF, p-value: 3.282e-11

As the p-value is much less than 0.05, we reject the null hypothesis that β = 0.
i. Given your findings in parts f, g, and h, how does an increase in earnings per share affect the return
on equity?
Note: This part is worth 10 points.

There is a positive linear relationship between EPS and ROE.


For each one unit increase in EPS, the expected change in ROE is equal to the estimated slope
coefficient (β) which is 0.11913 (increase of 0.11913 units).

j. Illustrate your answer to part i using a scatterplot in which the earnings per share (EPS) is put on the
horizontal axis and return on equity (ROE) is put on the vertical axis. Make sure that you include the
linear fit in your scatterplot.
Note: This part is worth 10 points; 5 points for the scatterplot and 5 points for the linear fit.

k. List at least one new thing that your group learned as part of the peer review process.
Note: This part is worth 10 points.
If needed, I will ask your team to share your RScript for this group project with me. So, make sure that
you keep a copy of the script that you use for answering the above questions.

You might also like