Tools › Statistics › Uncertainty and Monte Carlo › Sensitivity Analysis › Example
Clemen and Reilly Chapter 5: Eagle Airlines one-way sensitivity, tornado and break-even analysis
| Tool | Sensitivity Analysis |
| Source | Clemen, R. T., and Reilly, T. (2014). Making Hard Decisions with DecisionTools, 3rd ed. South-Western Cengage Learning, Mason, OH. Chapter 5, Eagle Airlines. Verified through van Dorp, J. R., and Mazzuchi, T. A. (2006). EMSE 269 lecture notes, George Washington University. |
| Level | Undergraduate |
Problem
Eagle Airlines is considering buying an aircraft whose annual profit depends on hours flown, charter price, ticket price, occupancy, the proportion of charter flying, operating cost, insurance, the financed fraction, the interest rate and the purchase price, each with a low, base and high value. Compute the base-case profit, the profit with every input at its low and at its high value, rank the inputs in a tornado diagram, and find the occupancy and operating cost at which the profit equals the $4,200 of the money-market alternative.
Inputs to Enter
- Model (default): Revenue = R·H·CP + (1 − R)·H·TP·5·Occ; Cost = H·OC + I + PP·F·IR; Profit = Revenue − Cost; target 4,200
- H 800 (500–1,000); CP 325 (300–350); TP 100 (95–108); Occ 0.50 (0.40–0.60); R 0.50 (0.45–0.70)
- OC 245 (230–260); I 20,000 (18,000–25,000); F 0.40 (0.30–0.50); IR 0.115 (0.105–0.13); PP 87,500 (85,000–90,000)
Answers
- Base profit 9,975: revenue 230,000 (130,000 charter + 100,000 scheduled), cost 220,025 (196,000 variable + 24,025 fixed)
- All inputs low: profit −15,927.50 (revenue 119,750, cost 135,677.50); all high: 51,350 (342,200, 290,850)
- Tornado: occupancy first (swing 40,000 = 200,000 × 0.2, profit 9,975 ± 20,000), operating cost second (±12,000); hours flown 500 → −2,775, 1,000 → 18,475 (slope 42.5 per hour)
- Break-even against the 4,200 alternative: occupancy 0.471 at OC 245 (indifference line Occ = 0.004 OC − 0.509); operating cost 252.25 at occupancy 0.50; hours (4,200 + 24,025)/42.5 = 664
Notes
The tool's default inputs are this example; press Submit to reproduce it. The distributions attached to the variables for the Monte Carlo part are illustrative additions, as the page states; the book gives only low, base and high values.