Halcyon Foods: Capital Rationing with a Strategic Option
Halcyon Foods, a listed European convenience-food producer, has a capital budget of €40 million for the coming year, capped by a covenant limiting net debt to 3.0× EBITDA. Four proposals have been approved in principle: Project A, a chilled-distribution automation programme, outlay €22m, NPV €7.4m at the group's 9% WACC, six-year life; Project B, a plant-based ready-meal line, outlay €16m, NPV €5.0m, four-year life, but its cash flows sit in a new category whose appropriate discount rate is 11%; Project C, a €10m digital traceability system, NPV €2.7m, which can be deferred one year with no loss of strategic value; and Project D, a €6m Gulf export hub, NPV €0.8m, which the commercial director calls a platform for later expansion. Projects A and B are mutually exclusive because they would occupy the same factory footprint. The board wants a recommendation before the covenant test on 31 December.
1. Recommend which projects Halcyon should fund this year. Show the portfolio arithmetic and explain why a simple profitability-index ranking can mislead.
Read the worked answer
Profitability indices: A = (7.4 + 22)/22 = 1.336; B = 21/16 = 1.313; C = 12.7/10 = 1.270; D = 6.8/6 = 1.133.
Because A and B are mutually exclusive and projects are indivisible, rank feasible portfolios by total NPV within €40m:
- A + C + D = €38m, NPV €10.9m
- A + C = €32m, NPV €10.1m
- B + C + D = €32m, NPV €8.5m
- A + D = €28m, NPV €8.2m
- B + C = €26m, NPV €7.7m
Best is A + C + D. A PI queue would place B second and might tempt the board to drop C or D, but B cannot coexist with A. B's NPV is also quoted at the 9% group rate although its risk warrants 11%; its risk-adjusted NPV must therefore be below €5.0m, so it does not overturn the result. PI ignores scale, timing and indivisibility. Recommend funding A, C and D, leaving €2m contingency. Confirm B's risk-adjusted NPV before any future re-run.
2. Should Project C be deferred? What is the value of the deferral option, and what conditions would justify it?
Read the worked answer
Deferring C releases €10m this year but sacrifices €2.7m of NPV now. Because the €40m limit is hard and unused budget cannot be banked, A + D would use only €28m and leave €12m idle; the released cash earns nothing. Deferral therefore destroys €2.7m unless a qualifying project can use the €10m this year.
The deferral is a real call option: it has value if waiting resolves uncertainty or avoids locking into a technology, and if the project can still be exercised next year at a known cost. Value it as expected NPV next year minus the NPV forgone now, discounted, less any cost of delay. Here the stated no-loss condition means the option's time value is small; the exercise price is the €2.7m forgone. Defer only if (i) next year's budget is expected to be tighter and C would displace a project worth more than €2.7m, (ii) a pilot or regulatory decision would raise C's expected NPV by more than €2.7m, or (iii) the €10m can be carried forward or used to repay debt yielding more than the deferred NPV. On the facts, fund C now.
3. The commercial director wants D funded even if it means dropping C. Under what conditions, if any, should a low-NPV strategic project be accepted in a hard-rationing regime?
Read the worked answer
Accept a low-NPV strategic project only when its real-option value exceeds the NPV of the project it displaces. Here D's €0.8m NPV must be compared with C's €2.7m, so D needs an option value above €1.9m to justify dropping C. Estimate that option value with a decision tree: a 30% chance of a follow-on project with NPV €10m gives expected value of €3.0m, enough to justify D; a 15% chance gives €1.5m, not enough.
Other legitimate conditions are a regulatory licence or route to market that cannot be bought later; learning that reduces the cost of a larger subsequent investment; an abandonment option that limits downside; or a staged investment where Halcyon can stop after a pilot. Vague claims of platform or strategic presence are not enough. Require stage-gated funding, a named expansion decision and a sensitivity showing the option value. If those cannot be evidenced, rank on NPV and fund C.