Real options

Let K be a commitment cost and V(s) the project value in state s, measured on the same present-value date. Immediate commitment compares expected V(s)−K with the other feasible actions. A stage that produces a signal z lets the decision maker choose an action after observing that signal. Its value depends on the actions and information actually available. Valuing the choice requires evidence about the states, their probabilities, project values and commitment costs.

A symbolic staged comparison is expected best feasible net payoff conditional on z, less the cost of obtaining the signal and any delay, duplicated work or exit cost not already counted in that payoff. Count each cost once. If the tree already includes the choice to stop or continue, do not add another unsupported option premium. The symbolic expression cannot be evaluated without evidence about signals, states, payoffs, timing and authority.

Flexibility can exist without making staging worthwhile. Waiting may lose customers, capacity or an investment opportunity; a pilot may require assets that cannot be reused. A contract may continue to impose payments after the operational work stops. Compare these consequences with the commitment that learning could avoid. No claim about the typical success rate or monetary benefit of pilots follows from the framework.

Information is rarely perfect. A signal can be noisy, biased or too narrow to predict scale. Identify how evidence changes the action and whether the organization can obtain it in time. If signal reliability or feasible alternatives are unknown, state the research needed instead of presenting a precise valuation.

NASA’s actual CCiCap funding amounts and later optional increments form the reported-data exhibit. They document a funded choice within a specific programme phase. They are agreement amounts, not measured option value, actual spending or a full project return. Later certification contracts and oversight findings answer different questions.

The right must remain usable. A contractual exit can be undermined by a supply dependency, a shortage of alternatives, a regulatory obligation or a mission that cannot be abandoned. Budget authorization for the pilot does not guarantee follow-on funding. A stage gate should specify who can decide, what evidence they need and which alternative is actually available.

The paired [NASA case](/en/breakdowns/nasa-commercial-crew-stages) offers a concrete caution. NASA exercised optional milestones during capability development, but the June 2026 OIG release also reports failures to exercise limited simulator-data access rights. A formal right did not automatically become useful information. The source does not establish that using that right would have prevented the later technical difficulties. [OIG release, oversight findings](https://oig.nasa.gov/news/nasas-commercial-crew-program-faces-schedule-delays-cost-increases-and-safety-challenges/).

Do not equate risk reduction with value creation. A pilot can reduce uncertainty about a weak project and appropriately lead to stopping. It can also cost more than the avoided mistake. Compare staging with immediate action, a smaller permanent scope and declining the project entirely. Use the same objective and valuation date in each path.

Avoid importing financial-option assumptions without checking their fit. A non-traded project may lack a replicating asset, stable volatility or a price process resembling a security. A transparent decision tree can be more informative for management than a precise-looking option price whose probabilities and market analogues are unsupported.