Part I: Extract, transform and load the data (the ETL process)

The data for this case was extracted from company financial statements posted online from credible sources. The extraction of the data from the online sources was performed for you and the data has been loaded into the Excel file. That is, you can assume that the web scrapper accurately and completely extracted the information and loaded it into Excel. Most of the transformation work has been done for you as well. You will be required to do some transformation to analyze the data once it has been loaded into the appropriate analysis tool (e.g., you will need to compute the ratios involved in the DuPont Method).[1] You should load the data into an analytics tool for analysiswe will be using Tableau in this case.

Hint: when loading the data into Tableau, you need data from both the income statement and the balance sheet tabs. Make sure that you link the income and balance sheet data correctly by matching the data on both ticker symbol and year.

Required

When you are finished loading the data, you should answer these simple questions to make sure you loaded the data correctly (by testing for completeness and accuracy).

1. What are the combined total assets of all companies for all years?

2. How many different companies are listed in the dataset?

3. How many different companies are there in each industry?

Capital goods

Finance

Public utilities

Transportation

Consumer services

Technology

4. What are the total sales for each industry in 2017 (do not round your answers)?

Capital goods

Finance

Public utilities

Transportation

Consumer services

Technology

5. What company had the most sales over the three-year period and what was the total amount of those sales?

Analytics mindset

Implementing the DuPont Method

Part II: Apply appropriate data analytic techniques

You are now ready to analyze the data. Please answer the questions that are listed below. For each question:

Identify the type of the analysis to perform

Identify which data elements are needed to perform the analysis

Create the analysis using Tableau

Required

Questions about industries

1. Which industries for fiscal year 2019 are the highest and lowest performing for each of the following performance indicators? For measuring performance, use the median industry performance to control for the potentially large effects of outliers.

Item	Highest	Lowest
Return on equity
Profit margin ratio
Asset turnover ratio
Financial leverage ratio

2. Which industry has seen the greatest improvement in median ROE from 2017 to 2019? What are the best explanations, based on the ratios in the DuPont Method, for why the ROE has improved in that industry?

Questions about individual companies

3. What companies have the best ROE within each industry for 2019? Sort the data so you can see the companies listed from highest to lowest ROE. What observations do you make about differences in ROE for the different companies?

4. Companies that have negative profit margins but are increasing their asset turnover ratio are accelerating into a brick wall (i.e., they are getting better at losing money). What three companies in 2019 are the worst in that they have a negative profit margin and the highest asset turnover ratio? Give the name of the company.

Analytics mindset

Implementing the DuPont Method

Part III: Interpret and share the results with stakeholders

Now that you have completed your analysis, for each question, interpret your analysis and prepare a Tableau story including a visual dashboard or graphic with a concise written analysis of your interpretation of the analytic results. When considering your visualization, make sure to think about what type of visualization will provide the clearest and most compelling format for stakeholders to understand what you want to convey.

[1] Realize for many situations, extracting, transforming, and loading (ETL) the data can account for over 80 percent of the time in the entire data analysis process. This case simplifies this process so you can focus on developing other aspects of an analytics mindset. Also, the process does not always strictly follow the ETL format. Some transformation can happen before or after data is loaded into a tool for analysis.