Instructions

1.The assignment can be done in groups of one to three students. All team mem- bers are expected to contribute approximately equally to a group assignment.  A group can eliminate an underperforming member who then will need to do  the assignment individually or join another group.  Similarly,  one can quit

an underperforming group to do the assignment individually or join another

group. All group members will get the same mark for the assignment.

2.The maximum score is 25 points.

3.The presentation of your write-up is important.

4.All numerical analysis (including data cleaning etc.), all  tables  and  gures

need to be done using SAS or SAS Visual Analytics (however, you may use  Excel or Word etc. to make tables for regressions as the standard SAS output for regressions is not very nice).

5.Please retain your SAS code and make sure that it is user-friendly (use com- ments where necessary). Using your submitted code, one should be able to  produce all your results, tables,  and gures.

6.Please retain a copy of the problem set that is submitted.

7.Only one member of a group submits 3 les (alternatively, they can be com- bined into a single le):

ˆ `Assignment Cover Sheet', which must be signed (electronic signature is okay) and dated

ˆ    the report (in doc, docx, or pdf format) for Tasks  1 and 2; the report should be properly formatted and be similar to a business report; font: 12 pt Times New Roman; maximum number of pages: 10 (no penalty for exceeding this limit); at the end of the report (in the appendix) include your SAS code

ˆ   a presentation slides (in pdf, ppt, or pptx format) for Task 3.

8.Lecturer can refuse to accept assignments, which do not have a signed ac- knowledgment of the University's policy on plagiarism.

9.Any suspected plagiarism will be severely punished. This includes any student  that submits copied work or any student that allows their work to be copied.

10. You must acknowledge any external material you use in your answers, e.g.,

material from websites, textbooks, academic journals and newspaper articles.

11.All queries (including deadline extensions) for this project should be directed to Lecturer.

Agenda

Assume that you are a compensation consultant working at a leading consulting rm. Your  client  is  Encompass  Health  Corporation  one  of the  United  States'  largest providers of post-acute healthcare services. They need your help to determine the compensation of their CEO in year 2021.

You have been provided with 2 data sets. The dataset `salaries_2023_S2' con- tains the following variables:

ˆ   GVKEY Company ID Number

1    OLS regressions (9  points)

This task needs to be done using SAS onDemand for Academics. Discuss brie y your sample, including the number of observations, outliers. Provide the descriptive statistics of the sample. How you choose to do this is entirely at your discretion. However, it is recommended that you consider using both summary statistic and graphical methods (this task should include at least one properly formatted table, one pie chart, one histogram, and one scatter plot) while also noting any peculiarities within the data  set. You should put more emphasis on TDC1. Tables and gures should be included in the report rather than appendix.

Find the determinants of the total compensation (TDC1). Estimate 3 di erent OLS regressions (in order to ensure the robustness of results) with year and industry xed e ects and several independent variables.  To ensure that regression residuals `behave well', you may need to scale or transform one or more variables. For ex- ample, to use a natural logarithm value of the variable instead of its raw value. Provide a properly formatted table with the regression results in the report (not in the appendix; however, you may put additional tables in the appendix if needed). Discuss the determinants of the total compensation: what variables are statistically signi cant; which variables increase and which variables decrease total compensa- tion;   any   insights from the coe cient estimates of year and industry    xed e ects and so on.

Predict the total compensation in 2021 (this value has been deleted in the dataset `salaries') using the results from the 3 regressions for Encompass Health Corporation (GVKEY = 012589). Are the predictions similar across the 3 models?

2    Time series analysis (8 points)

This task needs to be done using SAS onDemand for Academics. Generate a time series values for total compensation (TDC1); that is, annual averages for each year. You may use SAS procedure MEANS with the following option:

BY year;

You may  need to sort the data prior to running SAS procedure MEANS. Plot the obtained time series. Use ARMA type models to predict its values for the next 2 periods. Motivate and discuss ARMA orders used in the analysis. Plot the  predicted values (a scatterplot against the actual values; then time series plots of the actual and predicted values in the same gure). Do actual and predicted time series tend to move in the same direction over time?

Given the time series predictions for years 2022 and 2023, what would be the predicted total compensation of the CEO of Encompass Health Corporation, ob- tained in Task  1 (assume that the total compensation of the CEO of Encompass Health Corporation evolves similarly as an average CEO total compensation)?

3    In ation forecasting (8  points)

This task needs to be done using SAS Viya for Learners platform. Assume that you  are an economist working at the Reserve Bank of Australia (RBA), and you have  been tasked to forecast quarterly in ation for the next 4 quarters (i.e, Sep-2023,Dec-  2023,Mar-2024, and Jun-2024) using autoregressive moving average (ARMA) type models. Historical in ation data can be downloaded from RBA website:https://     www.rba.gov.au/statistics/tables/xls/g01hist.xls?v=2023-10-04-10-19-06. The variable of interest is Consumer price index; Year-ended change (in per cent) (GCPIAGYP).

You should generate the predictions using the in ation data:

ˆ   from Dec- 1959 to Jun-2023 and then for shorter period,i.e.

ˆ   from Jun-2005 to Jun-2023,

and then compare the results. In the analysis, you may set `Binning interval' (under Options) to `Automatic' (this will make the time series plots look nicer).

Create presentation slides (e.g., using MS PowerPoint) that include:

ˆ   time series plots with the forecast

ˆ   the models SAS Visual Analytics used

ˆ   predicted in ation for the next 4 quarters

ˆ   90% forecast con dence for the rst quarter forecast (how do you interpret those values?).