Ed Draycutt is the engineering manager of Airway Technologies, a firm that makes computer systems for air traffic control installations at airports. He has proposed a new device whose success depends on two separate events. First, the Federal Aviation Administration (FAA) must adopt a recent proposal for a new procedural approach to handling in-flight calls from planes experiencing emergencies. Everyone thinks the probability of the FAA accepting the new method is at least 98%, but it will take a year to happen. If the new approach is adopted, radio makers will have to respond within another year with one of two possible changes in their technology. These can simply be called A and B. The A response is far more likely, also having a probability of about 98%. Ed’s device works with the A system and is a stroke of engineering genius. If the A system becomes the industry standard and Airway has Ed’s product, it will make a fortune before anyone else can market a similar device. On the other hand, if the A system isn’t adopted, Airway will lose whatever it has put into the new device’s development.

Developing Ed’s device will cost about $20 million, which is a very substantial investment for a small company like Airway. In fact, a loss of $20 million would put the firm in danger of failing.

Ed just presented his idea to the executive committee as a capital budgeting project with a $20 million investment and a huge NPV and IRR reflecting the adoption of the A system. Everyone on the committee is very excited. You’re the CFO and are a lot less excited. You asked Ed how he reflected the admittedly remote possibility that the A system would never be put in place. Ed, obviously proud of his business sophistication, said he’d taken care of that with a statistical calculation. He said adoption of the A system required the occurrence of two events, each of which has a 98% probability. The probability of both happening is (.98 × .98 = .96) 96%. He therefore reduced all of his cash inflow estimates by 4%. He maintains this correctly accounts for risk in the project.

Does Ed have the right expected NPV? What’s wrong with his analysis? Suggest an approach that will give a more insightful result. Why might the firm consider passing on the proposal in spite of the tremendous NPV and IRR Ed has calculated?