When does a DCF/NPV understate value, and how do real options capture what it misses?

A standard DCF discounts a single expected cash-flow path and implicitly assumes management commits today and does nothing as the future unfolds — so it ignores the value of flexibility under uncertainty. Real options capture managerial flexibility: the option to expand if a project succeeds, to abandon and recover salvage if it fails, to defer/wait for information, to switch inputs/outputs, or to stage investment. Because these are asymmetric (you invest more only in good states, cut losses in bad ones), they add value that a static NPV — which averages over outcomes — misses, and the value rises with uncertainty (more volatility = more option value), the opposite of how a discount-rate-penalized DCF treats risk. They matter most for R&D, natural resources (develop when prices are high), staged/phased investments, and start-ups, where flexibility is the main value. You quantify them with option-pricing logic (Black-Scholes-style or binomial/decision trees). The caveat: real options can be over-applied to justify weak projects, and the inputs (volatility, the option's exercisability) are hard to estimate — so use them where genuine flexibility and uncertainty exist.