Part 1

Return on Capital for Four Different Sectors

See Case 16 that starts on page 576 in the Aczel text. I have simulated those data and increased the available sample sizes for each sector. I have attached SPSS, NCSS and EXCEL files that contain the simulated data. Answer the following questions:

The accompanying data set presents simulated financial data of some companies drawn from four different industry sectors. The data include return on capital, sales, operating margin, and debt-to-capital ratio. All the data pertain to the most recent 12 months for which data were available for each company. The specific time period may be different for different companies, but we shall ignore that issue for purposes of data analysis.

Using suitable dummy or indicator variables to represent the sector of each company, regress the return on capital against all other variables, including the indicator variables.

1. Make histograms of each of the four variables. Then do scatter plots among the four variables.

What do the histograms and bivariate plots imply about the computed Pearson correlations among the four variables?

What might you do about possible problems with the correlations or with the resultant multiple regression models?

2. The sectors are to be ranked in descending order of return on capital. What will that ranking be?

3. It is claimed that the sector that a company belongs to does not affect its return on capital. Conduct an appropriate test within multiple regression to see if all the indicator variables can be dropped from the regression model.

4. For each of the four sectors, give a 95% prediction interval for the expected return on capital for a company with the following annual data: sales of $3 billion, operating margin of 31%, and a debt-to-capital ratio of 37%.

5. Use the three metric (continuous) independent variables and appropriate sector indicator variables to create three possible quadratic independent variables and 18 possible simple two way ANOVA interaction effects among the independent variables and sector classification.

Use the hierarchical multiple regression method to explore whether the quadratic effects or the 18 interaction effects are statistically significant. If you found any significant quadratic or two way interaction effects, how should you interpret them?

Part 2: An Anomalous Data Set

Examine the data set called “anomaly.” Y is the dependent variable, and X1 and X2 are the independent variables. Use multiple regression analyses to model Y from X1 and/or X2.

What problems, if any, arise with this data set? What can or should you do to detect and reduce these problems?

Price: $101.75

Solution: The downloadable solution consists of 21 pages, and 1201 words
Deliverable: Word Document and pdf