Scribd
Upload
10 views8 pages

Problem Set 4

Uploaded by

manar.farhan.sayah
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
10 views8 pages

Problem Set 4

Uploaded by

manar.farhan.sayah
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 8

Manar Alsayah

Leen Bou Yahya

Exercise 1 In this exercise, we will build a regression model for predicting stock returns,
similar to what we have learned in lecture. The data file is apple.csv. This data tracks
the monthly performance of the stock of Apple. There are 264 months in the data. The
variables in the data are the following. ]

• Month: time variable


• AppleReturn: monthly % change (i.e., return or avkastning) in the value of Apple
stock
• MarketReturn: monthly % change (i.e., return or avkastning) in the value of the
whole stock market
• SP500Return: monthly % change (i.e., return or avkastning) in the value of the
S&P 500 Index (the 500 largest companies in the U.S.) bonds)

You can load the apple.csv data into Stata by using the command import delimited

(a) Estimate the SRM of AppleReturn (y variable) on MarketReturn (x variable).


(i) What is the equation of the sample regression line?

(ii) Interpret the estimated coefficients b0 and b1.

Intercept (b0) = 0.0152902

• This means when the market return is 0%, Apple stock is expected to have a
return of 1.53%

Slope (b1) = 1.345951

• This means that for every 1% increase in the market return, Apple stock’s return
is expected to increase by 1.35%
(b) Estimate the MRM regression of AppleReturn (y variable) on MarketReturn and
SP500Return (x variables).

(i) What is the equation of the sample regression line?

The equation of the sample regression line:

Apple^Return = 0.122065 + 3.017439 x MarketReturn – 1.764322 x SP500Return

(ii) Interpret the estimated coefficients b0, b1, and b2.

Intercept (b0) = 0.122065

• This means when both the market return and SP500 return are 0%, Apple’s stock
return is expected to be 1.22%

Market Return coefficient (b1) = 3.017439

• This means that for every 1% increase in the market return, Apple’s stock return
is expected to increase by 3.02%

SP 500 Return coefficient (b2) = - 1.764322

• Which means that for every 1% increase in the SP 500 return, Apple’s stock
return is expected to fall by 1.76%

(iii) Does the model have explanatory power?


No, because the R-squared is 0.2055, that means that the model has limited
explanatory power.

(iv) Does the model satisfy the MRM conditions? Hint: create the different plots and
figures we learned in lecture.

(c) Find the correlation of AppleReturn, MarketReturn, and SP500Return. Are there any
variables that are highly correlated? Which ones? Hint: use the command correlate.

The variables that are highly correlated are Market Return and SP 500 Return, as their
correlation coefficient is 0.9824, which is very close to 1.

(d) Looking at the results in parts (a)–(c), are there any signs of collinearity? Explain.
Yes, there are signs of collinearity. Market Return and SP 500 Return have a high
correlation of 0.9824, meaning they are very similar. This can make it hard to figure out
the individual effect of each variable on Apple Return, leading to less reliable results in
the regression.

(e) How would you suggest improving the MRM in part (b), or would you just leave it as
is?

To improve the MRM in part (b), i would suggest removing one of the correlated
variables (Market Return or SP 500 Return) to reduce collinearity.

Exercise 2 The supermarket company Rema1000 has a loyalty program through their
app Æ. In June 2022, the company changed the structure of the program. Now,
customers receive coupons with person- alized discounts in the Æ app.

Suppose that Rema1000 has asked you to investigate if the coupons lead to higher
spending. You are given a dataset of 1000 Æ app users in June 2022. The data has the
following variables: the amount of money spent by the user at Rema1000 (kroner),
whether the user used a coupon (Yes/No), whether the user is a student (Yes/No). See
the figure below for what the data looks like.

You want to use these data to estimate the effect of coupons on Æ users’ spending.
Note: There is no dataset for this question. You are asked to explain how you would do
the analysis if you had the data.

(a) You hypothesize that even if you compare two different users who both used a
coupon, a user who is a student will generally have lower spending than a user who is
not a student.
(i) Write the population regression model that would allow you to test this hypothesis.
Explain each variable in your model. Explain the meaning of each coefficient in your
model.

You might also like