Staged Injectable Line After a Sunk Regulatory Dossier
Lahore Pharma has spent PKR 45 million over two years preparing a dossier for a new sterile injectable line and has not yet received DRAP registration. Management must decide whether to build the line now or contract manufacture. Building requires PKR 320 million immediately for clean rooms and PKR 480 million in two years for filling and lyophilisation, only if registration is granted. If DRAP approves on the expected timeline, the line generates a present value of PKR 1,250 million of after-tax operating cash flows at the project's 17% WACC. If approval slips 18 months, competition and price erosion reduce that present value to PKR 900 million while the phase-two cost stays the same. The company's regulatory consultant puts the probability of on-time approval at 65%. Contract manufacturing requires no capex and has a present value of PKR 210 million. Tax is 29%, and the plant would be depreciated on a reducing-balance basis.
1. Should the PKR 45 million dossier cost influence the decision? Recompute the expected NPV of the phased build and compare it with contract manufacturing.
Read the worked answer
The PKR 45m is sunk. It was incurred before the decision point and does not change with the choice between building and contract manufacturing, so it should not be added to either branch. The only exception is any future tax deduction or amortisation that the 45m generates; then include the tax effect, not the gross cost.
Phase-two capex at year 2 has a PV of 480 / 1.17^2 = PKR 350.7m. On-time approval gives 1,250 - 320 - 350.7 = PKR 579.3m. Delayed approval gives 900 - 320 - 350.7 = PKR 229.3m. Expected NPV = 0.65 x 579.3 + 0.35 x 229.3 = PKR 456.9m. Contract manufacturing is PKR 210m. The phased build is worth about PKR 247m more on these assumptions, so proceed with phase one, but keep phase two as a separate gate. If the company included the 45m in the build branch only, it would understate the build by 45m and might wrongly prefer contract manufacturing. If it included the 45m in both branches, the ranking would not change but the reported NPV would be misleading.
2. Explain why a decision tree may give a different answer from a single 17% discounted NPV, and how you would avoid double-counting risk.
Read the worked answer
A single 17% discount rate applied to expected cash flows treats the project as a fixed set of cash flows with a constant risk profile. The staged project is not fixed: management can stop after phase one if registration is delayed or denied, so the downside is truncated and the value of that flexibility is not in a static NPV. A decision tree separates the regulatory outcome from the investment decision and values the option to abandon or defer.
The risk of double counting arises if the probabilities are already reflected in a risk-adjusted discount rate. Use one of two consistent approaches. Either use actual probabilities and discount each branch's certain cash flows at a rate that reflects only time value and non-diversifiable market risk, or use risk-neutral probabilities and the risk-free rate. Do not apply 17% and also apply a further haircut for regulatory risk. The 17% may be appropriate for the operating phase after approval, but phase one is largely a call option on phase two and may warrant a different rate. State the rate for each branch and test the decision at 15% and 19%.
3. DRAP offers a fast-track route that would cut the delay to six months at an extra PKR 25 million in phase one. Is it worth it?
Read the worked answer
The fast-track fee buys earlier cash flows, not a higher probability of approval, so value it as a timing benefit. If approval is granted, moving the operating cash flows forward by roughly twelve months from the 18-month-delay case to a six-month delay is worth approximately 1,250 x (1 - 1/1.17) = PKR 181.6m in present value, before considering the probability of approval and the extra 25m. The net value is positive even after tax if the fee is deductible.
However, compare the correct counterfactual. The base case already assumes an expected timeline; fast-track reduces a delay from 18 months to six, a twelve-month acceleration, not from zero. If the expected timeline is already six months, the incremental benefit is smaller. Also, the extra PKR 25m is incurred whether or not approval is granted, so it belongs in phase one and lowers the expected NPV by 25m, not by 25m times the approval probability. If fast-track also raises the approval probability, or shortens the delay only in the approved branch, model that separately. On the stated facts, the fee is worth paying if financing is available.