Cost Estimation
This project requires you to use information about Alaska Air Group to examine issues related to cost behavior, revenue drivers, and profit estimation. For those who are unfamiliar with Alaska Air, I have included their summary of business from a recent 10-K below. The Excel file contains historical information from 2004 to 2016 for the consolidated operations, which encompasses subsidiaries Alaska Airlines and Horizon Airlines. The file contains a variety of operating expenses (the costs) and several potential cost drivers (revenue passenger miles, available seat miles, and average number of full-time equivalent employees) as well as historical non-company data (GDP, population numbers etc.).
Excerpt from Alaska Air Group’s 2016 10-K describing their business
Air Group operates Alaska, Virgin America and Horizon Air. We completed the acquisition of Virgin America on December 14, 2016, at which time Virgin America became our wholly-owned subsidiary. Together with our regional partner airlines, we fly to 118 destinations with nearly 1,200 daily departures through our expansive network across the United States, Mexico, Canada, Costa Rica and Cuba.
With our global airline partners, we can provide our guests with a virtual network of more than 900 destinations worldwide. During 2016, we carried an all-time high 34 million guests and earned adjusted net income of $911 million, which includes operating and financial results for Virgin America for the period December 14, 2016 through December 31, 2016, and excludes pretax special items and merger-related costs of $117 million.
Our acquisition of Virgin America positions us as the fifth largest airline in the U.S., with an unparalleled ability to serve West Coast travelers. Virgin America provides a platform for growth of our low-fare, premium product providing a powerful West Coast network for our guests as well as enhanced international partnerships. Additionally, Virgin America provides an opportunity to grow and improve our loyalty program while gaining access to constrained gates, particularly on the East Coast. The combined company now provides more seats from the West Coast than any other carrier, allowing us to serve our guests better.
Our mission is “creating an airline people love.” The “ing” is to recognize we are never done—we are continually working to get better. We believe our success depends on our ability to provide safe air transportation, develop relationships with guests by providing exceptional customer service and low fares, and maintain a low cost structure to compete effectively. It is important to us that we achieve our objective as a socially responsible company that values not just our performance, but also our people, our community and our environment.
Revenue Estimation
Policy on Academic Honesty
This is an individual assignment. You can ask me for assistance, but you are not to work with other students on this project or seek answers from other students. This includes comparing answers. Doing so will constitute a major violation of academic integrity standards and, if detected, will have serious consequences.
By typing your name below, you assert that you have completed this project on your own:
Enter text here Questions (100 points in total)
Use the file “Alaska Air Group Data” to answer the questions in this document. Enter all your numbers and explanations in this Word document. For grading, also submit your Excel calculations on which your answers in this document are based.
You might find it helpful to create scatter plots with a trend lines before you use the “linest” command. However, any graphs that you create will not be graded.
Cost Estimation (53 points)
- Explain Total Operating Expenses using available seat miles (ASM) as the cost driver.
- (6 pts) Compute the slope, t-statistic of the coefficient, and the intercept for a linear cost function.
Slope = 0.092 t-stat = 5.66 Intercept = 852.379
- (2 pt.) How much does an additional seat mile cost according to your regression output?
Answer:
An additional seat mile (ASM) (in 000,000) would cost extra 0.092 unit, of total operating expenses.
- Explain Total Operating Expenses using revenue passenger miles (RPM) as the cost driver.
- (6 pts) Compute the slope, t-statistic of the coefficient, and the intercept for a linear cost function.
Slope = 0.096 t-stat = 7.05 Intercept = 1294.813
- (4 pts) Which of the two cost drivers (ASM or RPM) would you choose if you had to decide for one of them? Name an empirical (based on your numbers) and a theoretical reason.
Answer: Statement: We should choose the linear regression model of “Total Operating Cost” vs. “Revenue Passenger Miles” between these two models.
Evidences: The Value of multiple R2 in the first model (Total Operating cost vs. ASM) is 74.44%. The Value of multiple R2 in the second model (Total Operating cost vs. RPM) is 81.86%. Here, R21 < R22. The Value of F-statistic in the first model is 32.04 and second model is 49.66. Here, F1 < F2. Significant p-values of the two models are respectively, 0.000148 and 0.0. Here, p1 < p2.
Justification: The second model is better as per higher R2 value (coefficient of determination) and lower F-statistic (indictor of ratio of two variables). P-value is more satisfactory for second model. Therefore, we can infer that the second model is better (Montgomery, Peck and Vining 2012).
- (10 pts) Substantiate your suggestion from b. with multiple regression analysis. Also, describe how the results support your choice from b.
Name of cost driver: Revenue Passenger Miles Slope = 0.22456 , t-stat = 2.7926
Name of cost driver: Available Seat Miles Slope = -0.13128, t-stat = -1.6214
Intercept = 2126.93, t-stat = 3.4942
Explanation:
The multiple regression model undertakes “Total Operating Expenses” a dependent variable as well as “Revenue Passenger Miles” and “Available Seat Miles” as independent variables. The now model shows RPM is positively associated with Total Operating Expenses (slope is positive) and ASM is negatively associated with Total Operating Expenses (slope is negative). Hence, if RPM increases, operating expense enhances whereas if ASM decreases, operating expense enhances. Conversely, if RPM decreases, operating expense decreases whereas if ASM increases, operating expense decreases.
- Analyze the behavior of Wages and Benefits. The obvious cost driver is the average number of full-time employees.
Profit Estimation and Forecast
- (1 pt.) Based on economic intuition, do you expect the cost function to be variable, mixed, or fixed? Why?
Based on the intuition of Economics, we expect that cost function is variable. The cost function is variable as the average number of full-time equivalent employees (FTE) sometimes may vary and sometimes may not vary according to the increment or decrement of wages and benefits. - (4 pts) Based on a linear regression, is the cost function most similar to a variable, mixed, or fixed cost? Explain your answer by referring to numbers from your regression output.
The linear regression model is: Wages and benefits = (-119.8233) + 0.0904 * Average number of full-time equivalent employees (FTE). The R2 is 0.3408. Therefore, FTE can only describe 34.08% variability of Wages and benefits. The coefficient of determination refers that the linear association between dependent and independent variables is weak and positive (slope is positive). Hence, the cost function is most similar to mixed cost as per our assumption. Note that, R2 near to 1 interprets variable cost and near to 0 interprets fixed cost.
- (1 pt.) Based on economic intuition, do you expect the cost function to be variable, mixed, or fixed? Why?
- Analyze the behavior of the cost of Food and beverage service.
- (2 pts) Before performing any analysis, which two operating statistic variables are the most likely to be a cost driver that influences food and beverage costs and why?
Among the five cost drivers, most significant cost-drivers are- Revenue passenger mile (RPM) (000,000) “traffic”: RPM in transportation industry metric indicates the number of miles travelled by paying passengers in an airline traffic industry (Oum and Yu 2012). As the RPM fluctuate, as per our assumption the food and beverages also fluctuate significantly. Load factor: It measures the capacity utilization of public transport services in Airlines. This factor is just the ratio of number of passengers and available seats considering both velocity and number of flights. This break-even load factor would lead to high significance with the amount of food and beverages (Panzar 1979).
- (8 pts)Test your intuition from part (a) by running linear regression analysis. First, perform a separate analysis for each cost driver (8 points).
Cost driver = Revenue passenger mile (RPM) Slope = 0.0004, t-stat = 18.5052 Intercept = -38.3832
Cost driver = Load Factor Slope = 397.665, t-stat = 3.5177, Intercept = -249.1726
(8 pts) Now, perform multiple regression with both cost drivers.
Name of cost driver: Revenue passenger mile (RPM) Slope = 0.00494 t-stat = 14.5654
Name of cost driver: Load Factor Slope =
-90.8041 t-stat = -2.1660
Intercept = 19.9643, t-stat = 0.72746
(2 pts) Which of the three models best predicts food and beverage service costs and why?
The R2 values of the three consecutive models are- 96.89% (“Food and Beverage Service” and “RPM”), 52.94% (“Food and Beverage Service” and “Load Factor”) and 97.88% (“Food and Beverage Service”, “RPM” and “Load Factor”). The third file that is multiple regression model predicts food and beverage service costs with most perfection. Here, dependent factor “Food and Beverage Service” is 97.88% (maximum in three models) is predicted by both “RPM” and “Load Factor”.
Revenue Estimation (23 points)
- Now turn to the non-company data as possible explanatory variables of Total Operating Revenues:
- The Alaska Gross Domestic Product (GDP) per capita in $. Since the Alaska economy is strongly driven by commodity prices, you can assume that a high GDP correlates significantly with a high oil price.
- The population of the state of Alaska in million.
- The US Gross Domestic Product (GDP) per capita in $. Note the dent during the Great Recession in 2008 and 2009.
- The population of the US in million.
- A measure of customer satisfaction for Alaska Airlines (one of the airlines in the Alaska Air Group portfolio).
- A measure of inflation for the US.
- (12 pts) Comment on the economic plausibility of each variable. Which type of relationship (positive/negative/none) with Total Operating Revenue do you expect and why?
Alaska GDP per capita: A positive relationship between GDP per capita and “Total Operating Revenue” might be economically plausible in Alaska. As the GDP per capita of Alaska would increase, the estimated revenue would be more for Alaska Airways.
Alaska Population in millions: No association between population in million and “Total Operating Revenue” might be economically plausible in Alaska. As the population of Alaska would increase, the estimated revenue would may be or not be more for Alaska Airways.
US GDP per capita: A positive relationship between population in million and “Total Operating Revenue” might be economically plausible in USA. As the GDP per capita of US would increase, the estimated revenue would be more for Alaska Airways as both countries are collaborated in the relationship with each other.
Conclusion
US population: No association between population in million and “Total Operating Revenue” might be economically plausible in USA. As the US population would increase, the estimated revenue would may be or not be more for Alaska Airways.
Customer Satisfaction: A positive relationship between customer satisfaction and “Total Operating Revenue” could be economically plausible. For greater number of satisfied customers, the total operating revenue would get increased for their future acceptance.
Consumer Price Index: A negative association between customer satisfaction and “Total Operating Revenue” can be economically plausible. Inflation must cause the decrement of the Total Operating Revenue.
Overall a positive and significant relationship may be found with Total Operating Revenue.
- (7 pts) Pick 3 of the variables above and conduct multiple regression analysis using those variables simultaneously. You might want to experiment with different combinations of variables to see if one “model” works better than other models. Be careful when labeling your regression outputs with the correct variable names. Show your results here (note only 3 of the 6 will be filled in based on which variables you select):
GDP (Alaska): Slope = 0.0876 t-stat = 2.2827
Population (Alaska): Slope = na t-stat = na
GDP (US): Slope = 0.3405 t-stat = 2.8066
Population (US): Slope = na t-stat = na
Consumer satisfaction: Slope = 17.1001 t-stat = 3.1943
Inflation: Slope = na t-stat = na
Intercept = -30549.88435 t-stat = -6.26683
- (4 pts) Considering the results on statistical significance, how do you explain possible discrepancies between your expectations from a. and the directional effect of some variables or their observed t-stats?
The possible discrepancies between expectations from the directional effect of some variables indicate that consumer satisfaction has maximum t-value (3.1943) than t-values of other two factors.The Discrepancy found in this analysis from the assertion is that, GDP of USA and Consumer Satisfaction has significant association with “Total Operating Revenue”. However, Alaska GDP per capita is significantly related with the dependent variable.
Profit Estimation and Forecast (24 points)
- Analyze Operating Income (excluding “other net” revenue, “other” costs, and “restructuring charges”).
- (1 pt.) Prepare a new line in your spreadsheet for Adjusted Operating Income excluding the items listed above so that you can use it as a new dependent variable that represents profit.
- (4 pts) Explain what Load Factor represents (hint: the equation to calculate it in the spreadsheet should help). Consider Load Factor as a potential profit driver. Why is it economically plausible to assume a causal relationship?
Load factor as a potential profit driver is assumed. Load factor may be economically plausible to assume a causal relationship as it is a ratio index performing the percentage of passengers travelling in flight with respect to seats available when both factors, number of flights and speed of flights are accounted. - (8 pts) Compute the slope, t-statistic of the coefficient, intercept, and R2 for a linear profit function using Load Factor.
Slope = 629.024, t-stat = 7.572, Intercept = (-255.666), R2 = 0.83903.
- (4 pts) Despite economic plausibility, you will find that the fit of your model in c. is relatively low and that the statistical significance of the proposed profit driver is borderline. Considering that aircraft fuel is a large item that affects operating income, explain how fuel costs can interfere with the analysis that you conducted.
Despite economic plausibility, it is found in the linear regression model that goodness of fit of the linear regression model is sufficient. The value of R2 = 83.903%, which shows that quality of fitting is not bad.We consider that aircraft fuel is a large item that hampers operating income. The linear regression shows that the value of multiple R2 is 0.005. In this model we assumed “Adjusted Operating Income (Loss)” as dependent variable and “Load Factor” as independent variable. But, the goodness of fit in terms of the multiple R2 is found to be very bad.
- (4 pts) Propose an alternative analysis that overcomes the problem described in d. and conduct this analysis. Show whether the goodness of fit and the statistical significance of Load Factor’s slope coefficient improve.
For improving the fitting of the model between load factor and adjusted operating income, we included another probable significant factor “Aircraft fuel” in the model. The linear regression model becomes multiple regression model. Here, “Adjusted Operating Income (Loss)” is dependent factor and “Aircraft fuels” as well as “Load factor” are the independent factors.
In this multiple regression model, the value of multiple R2 is 83.909%. It is not much greater than the simple linear regression model, when “Adjusted Operating Income” is dependent and “Load factor” as single predictor. Therefore, the statistical significance in terms of multiple R2 has not significantly improved.
- (3 pts) Forecast Alaska Air Group’s (unadjusted) Operating Income for 2017 by using your results from the cost and revenue estimation parts of this project. For estimating Total Operating Expenses, use the cost driver that you have selected. You will find “forecast” data on the second tab in your Excel file. Assume that Alaska Air Group will be able to maintain their J.D. Powers score from 2016.
2017 Operating Income = 1161.9 US dollar.
References:
Feng, C.M. and Wang, R.T., 2000. Performance evaluation for airlines including the consideration of financial ratios. Journal of Air Transport Management, 6(3), pp.133-142.
Montgomery, D.C., Peck, E.A. and Vining, G.G., 2012. Introduction to linear regression analysis (Vol. 821). John Wiley & Sons.
Oum, T.H. and Yu, C., 2012. Winning airlines: Productivity and cost competitiveness of the world’s major airlines. Springer Science & Business Media.
Panzar, J.C., 1979. Equilibrium and welfare in unregulated airline markets. The American Economic Review, 69(2), pp.92-95.
Seber, G.A. and Lee, A.J., 2012. Linear regression analysis(Vol. 329). John Wiley & Sons.
Order an Essay Now & Get These Features For Free:
Turnitin Report
Formatting
Title Page
Citation
Outline
