Baraka Energy: Project Selection Under a Capital Cap with FX Exposure
Baraka Energy Ltd must choose projects for 2026 from a capital budget capped at KES 100 million by a lender covenant. The board has resolved to keep a 20% strategic reserve of the budget uncommitted. All figures are KES millions, post-tax, in real terms, discounted at a 16% WACC.
Project A (solar plant): cost 60; NPV 18; IRR 22%; payback 3.1 years. Forty per cent of its revenue is contracted in US dollars under a 10-year fixed tariff.
Project B (diesel back-up plant): cost 50; NPV 14; IRR 26%; payback 2.4 years.
Project C (metering upgrade): cost 20; NPV 6; IRR 18%; payback 4.2 years.
The directors favour Project B because it has the highest IRR and the shortest payback. The shilling has depreciated 9% against the dollar over the past year and 12-month forwards are quoted at a 6% annual premium.
1. Determine the combination of projects the board should approve, showing all feasible combinations and the basis for your ranking.
Read the worked answer
Deployable capital. 100 × (1 − 0.20 reserve) = KES 80 million. This is the binding constraint; the reserve is a board policy, not a legal restriction, so it should be challenged only if the marginal project clearly dominates.
Feasible combinations (cost ≤ 80):
| Combination | Cost | NPV |
|---|---|---|
| A only | 60 | 18 |
| B only | 50 | 14 |
| C only | 20 | 6 |
| B + C | 70 | 20 |
| A + C | 80 | 24 |
| A + B | 110 | not feasible |
| A + B + C | 130 | not feasible |
Profitability index (NPV per unit of scarce capital):
- A: 18 / 60 = 0.300
- B: 14 / 50 = 0.280
- C: 6 / 20 = 0.300
Recommendation. Approve A + C: cost KES 80 million, NPV KES 24 million, index 0.300. It exactly uses the deployable capital and produces the highest NPV of any feasible set.
Note on the ranking. With a single-period cap and divisible projects, rank by index; where projects are indivisible (as here), enumerate the feasible sets, because greedily taking the top-index project first would give A, then C, which happens to be optimal - but B + C (index 0.286 blended) beats B alone, showing that ranking by IRR (B first) would have produced a clearly inferior KES 14 million or KES 20 million outcome.
2. Explain why the IRR ranking misleads the directors here, and how you would address the payback preference.
Read the worked answer
Why IRR misleads.
- Scale. IRR is a percentage and ignores the size of the investment. B's 26% is earned on KES 50 million; A's 22% is earned on KES 60 million and creates KES 4 million more absolute value. Value maximisation requires NPV, not the rate.
- Reinvestment assumption. IRR implicitly assumes cash flows are reinvested at the IRR itself. A 26% reinvestment rate is not credible at a 16% WACC; NPV assumes reinvestment at the cost of capital, which is the realistic case.
- Ranking conflict is the decision rule. Where IRR and NPV conflict on mutually exclusive projects, NPV prevails because it measures the change in shareholder wealth directly.
- Lending/borrowing and multiple-sign anomalies. B's diesel plant carries fuel-price risk that could produce negative later cash flows; a non-conventional stream can generate multiple IRRs and make the metric meaningless.
On payback. Payback measures liquidity and risk exposure, not value. B's 2.4-year payback is attractive, but:
- It ignores all cash flows after the cut-off, which is precisely where the solar plant's 10-year contracted tariff sits.
- It ignores the time value of money unless discounted payback is used.
- A 2.4-year payback is a symptom of a short-lived asset, not of superior returns.
What to tell the board. Use payback only as a secondary screen - for instance, as a tie-breaker or as a check on the maximum acceptable exposure period. Present the decision as: NPV KES 24 million for A + C versus KES 14 million for B, with a sensitivity showing the discount rate at which A + C ceases to dominate (test at 18% and 20%). Request the cash-flow schedules for A and B so the crossover rate can be computed and tabled.
3. Advise on managing the US dollar exposure on Project A, quantifying the cost of the forward alternative against the recent depreciation.
Read the worked answer
Exposure profile. 40% of Project A's revenue is USD-denominated over 10 years - a long-dated, recurring economic exposure, not a one-off transaction exposure. The relevant risk is that a weaker shilling increases KES receipts; the risk the board actually faces is shilling appreciation, which would compress KES revenue and could push Project A's NPV below the KES 18 million used in the ranking.
Sensitivity. If the shilling appreciated 9% and the USD portion is 40% of revenue, KES revenue falls by roughly 3.6% at the top line; on a plant with high fixed costs the NPV effect is larger. Indicatively, a 9% adverse move on 40% of revenue could remove KES 4–6 million of NPV - enough to make A + C (24) only marginally better than B + C (20).
Options and cost.
| Instrument | Mechanism | Cost / constraint |
|---|---|---|
| Forward contract | Locks the rate | 6% annual premium; a 10-year forward is not obtainable in the Kenyan market |
| Money-market hedge | Borrow USD, convert, invest KES | Requires USD borrowing lines; replicates the forward |
| FX option | Floor on KES receipts, retains upside | Premium payable upfront; tenor limited to ~2 years |
| Natural hedge | Match USD debt to USD revenue | Cash-flow match over the asset life; cheapest if available |
| Tariff indexation | Negotiate KES or indexed tariff | Contractual, not financial |
Recommendation.
- Do not hedge the full 10 years with forwards - the 6% premium paid annually would consume a significant slice of the NPV, and long tenors are unavailable.
- Adopt a layered programme: hedge 60–70% of the next 24 months with forwards and options, rolling quarterly; leave the residual open, since USD revenue is a natural hedge against imported equipment and dollar-denominated debt.
- Explore USD-denominated project debt sized to the USD revenue stream (natural hedge) - this removes the exposure at source and avoids premium cost.
- Set a written hedging policy with an authorised instrument list, a prohibition on speculative positions, and delegated limits; report mark-to-market monthly to the board.
- Re-run the A + C versus B + C comparison on a hedged basis before final approval, so the ranking is not flattered by an unhedged FX assumption.