Empirical Analysis of US firms during 2013-2020

1. Using Capital IQ1, download data for at least 50 US firms for the period 2013-2020. Make sure that your sample includes at least 2 firms in the manufacturing sector. Present a table of variable definitions. Label this table: Table1: Variable Definitions. Make sure to include all the variables used in the regressions2. (450 words max)
[5 marks]

2. Present a table of summary statistics for all the variables used in the project (including the components of Tobin’s Q, although you don’t need to define each component in Table1). Make sure to include: mean, standard deviation, min, max, 25% percentile, 50% percentile, 75% percentile, number of firms, number of firm-year observations3. Check that your sample is balanced (e.g. the 50 firms have data throughout the period of study). Label this table: Table2: Summary Statistics4 and include a legend at the end of the table that defines each variable; e.g. Price denotes Day Close Price; Equity denotes Total Common Equity, etc.
[5 marks]

3. Firm size is measured as log(Total Assets). Performance is measured with Tobin’s Q (total assets plus market value of equity less book value of equity divided by total assets; where market value of equity equals price per share times the total number of shares outstanding)5. Choose the largest firm of your sample.

Using a difference in means test assess whether this firm performed better in the second term of Obama’s presidency (2013-2016) than in Trump’s presidency period (2017-2020). Show the results of your test (a direct output from STATA or SAS) and label this table: Table3: Differences in Means Test. Do you reject the null hypothesis? Say Yes or No and explain how did you get to this conclusion (50 words max).
 

4. Write 2 limitations of the test above (50 words max)
 
a A separate guide to Capital IQ can be obtained from the Project folder in Learn.

b An example of a variable definitions table is included at the end of the document.

c You can include the number of firms as a footnote since some students find challenging incorporating this figure in the table.

d Note that sometimes zeroes denote missing values. Make sure your summary statistics look sensible. Marks will be deducted for wrong construction of variables or careless calculations.

e Selecting the correct variables to construct Tobin’s Q is part of your mark. You may want to refer to academic papers to make the right choice and include references in Appendix 2.
 
5. You will assess whether the average manufacturing firm had a better performance than the average non-manufacturing firm during the Donald Trump era (this was one of his target sectors in his “America First” narrative). To do this, you will need to run a cross-sectional regression controlling for firm size. First, compute a time average for every variable for each firm in this time-period.

Variable Definitions

Then, run a regression that enables you to assess whether manufacturing firms are associated with better performance. Label this table: Table 4: Regression Results- Trump

6. Based on your regression results in Table 4, discuss whether manufacturing firms perform better than non-manufacturing firms. Make sure to discuss statistical significance and interpret beta. (50 words max).

7. Run a similar regression to that in (4) but for the Obama period and report your results in Table 5: Regression results- Obama. Based on the regression results you have so far, did Trump deliver on his promise to support manufacturing firms? Say Yes or No and explain your answer. (100 words max). Hint: Assume the OLS assumptions hold (although only for this answer).

8. Now you will run a cross-sectional regression using all the time-periods (i.e. use averages based on 2013-2020), industry dummies and firm size. Make sure the manufacturing dummy is not the base category. Is the OLS beta estimator for manufacturing the minimum variance estimator? Say Yes or No and justify your answer (150 words max)

9. Add the following variables to the regression:

a. a sensible explanatory variable of your choice (you may need to look at some academic papers to make a good choice).

b. A sensible dummy variable of your choice (you may need to look at some academic papers to make a good choice)

Include the table with results. Make sure a definition of the added variables and reference(s) to justify your choice are added at the end of the table. For example: Board Diversity: is gender diversity computed as the ratio of women directors divided by the number of directors (Adams and Ferreira, 2000). Label this table: Table 6: Multiple Regression Model

10. Discuss whether the variable chosen in 9a. above is statistically significant at the 5% level and interpret your result (50 words max).

11. Discuss whether the variable chosen in 9 b. above is statistically significant at the 5% level and interpret your result. (50 words max).
 

12. Assume your variable of interest is size. Give an example of an omitted variable that is not possible to include in the regression that could lead to a negative bias and include a brief explanation. (100 words max)

13. Researchers typically use returns rather than prices when running regressions. Explain whether this could be related to fulfilling the OLS assumptions. Feel free to include graphs or figures if this adds value to your explanation. (100 words max)

14. Using daily prices for 2020 for one of your manufacturing firms (you do not need to include this variable in the descriptive statistics or any of the questions above), what is the predicted price for January 1st 2021 based on the random walk model? Clearly show how you estimated this price and discuss whether this is a good prediction. (100 words max)

Originality will be rewarded, so make sure you do not share your answers. You should only refer to tests, concepts and methods covered i n this course.

Your tables need to be incorporated using images, directly obtained through Stata or SAS. These tables are not included in your word count. If the instructions only asks you to include a table, you do not need to provide any further discussion for that task.

You should use Times New Roman 11 point font on A4 pages with 2.5 margins from each side. An absolute maximum of 1500 words is expected but a good project could include less than that. Include word count in the right corner of your project.

Appendix 1 (not part of the word count) should include the names of the 50 firms you are using. References (max 250 words) should be included in Appendix 2. Only use papers published in journals ranked as 3, 4 and 4*. A list of journal rankings can be obtained from the Project folder in Learn.

Make sure that the names you give to your variables are consistent across tables. For example, if you refer to Log(Assets) as SIZE, then your regressions should show the coefficient for SIZE and not for Log(Assets).

You will need to submit the file of commands and data that replicates your Tables. For example, if you are using STATA you need the do file and its corresponding dta file. These files are not part of the word count.