The Case Packer Capital Request
You are the operations finance manager at a regional beverage co-packer operating three plants. The plant manager at the largest site has requested $1.4 million for an automated case packer, supported by a five-year DCF showing an NPV of +$95,000 at the company's 11% hurdle rate and a payback of 3.9 years against a 3.5-year threshold for this risk class. The model assumes six FTE of labour savings at $52,000 fully loaded, an 8% throughput gain, $85,000 per year of added maintenance, $120,000 salvage value, and $90,000 of one-time working capital. Two facts were not reflected. First, the collective agreement bars layoffs, so four of the six operators would be retrained and retained; only two positions would actually disappear. Second, the sales forecast underpinning the throughput assumption is 3%, not 8%. The plant manager argues that safety, scrap reduction and shorter customer lead times justify approval regardless, and points out that the capital committee meets once more this fiscal year with $3.0 million of requests against a $1.8 million budget.
1. Which adjustments to the model are warranted, and how does each one move the NPV? Show the arithmetic where you can and state where you cannot.
Read the worked answer
1. Labour savings. The model claims 6 x $52,000 = $312,000 per year. Under the collective agreement only two roles are eliminated, so the defensible recurring saving is $104,000, a reduction of $208,000 per year. Discounted at 11% over five years (annuity factor 3.696), that removes about $769,000 of NPV, taking +$95,000 to roughly -$674,000. If the plant manager wants to claim value for redeploying the other four operators, he must document an alternative use that avoids a hire elsewhere; an assertion of 'capacity' is not cash.
2. Throughput. The 8% gain was modelled against demand growth of 3%, so five points of the assumed volume have no order behind them. The direction is unambiguous (downward) but the magnitude cannot be computed from the file: it depends on contribution margin per case, not revenue, and on whether the extra cases displace overtime or add sales. Ask for the volume-to-contribution bridge before re-running.
3. Salvage and working capital. Check the mechanics rather than the assumptions. The $90,000 of working capital should be an outflow at t0 and a recovery at t5, whose present value is 90,000 x 0.593 = about $53,000. If the model expensed it with no recovery, NPV is understated by $53,000; if it omitted it entirely, NPV is overstated by about $37,000. Salvage of $120,000 at t5 carries a present value of about $71,000 and should be confirmed against dealer quotes.
4. Payback. With the corrected cash flows, cumulative inflow does not recover $1.4 million plus working capital within five years, so the 3.9-year payback fails the 3.5-year threshold as well as the NPV test.
The model as submitted is not decision-grade. Do not approve on the strength of the qualitative list until those items are quantified or explicitly accepted as strategic spend outside the capital-rationing pool.
2. Recommend what you put to the capital committee, given that only $1.8 million of $3.0 million in requests can be funded.
Read the worked answer
Recommendation: defer, with a defined path back, rather than reject outright.
Reasoning:
- The corrected economics are negative on the numbers available, so the project cannot be ranked against competing requests on profitability index. Funding it would require displacing a project with a demonstrated positive NPV from a fixed $1.8 million pool - a real opportunity cost, not a bookkeeping one.
- A flat rejection is also wrong. Three inputs are genuinely unresolved: the contribution effect of the throughput change, the safety case, and whether scrap and lead-time improvements carry contractual value (customer penalties, expedite freight). Each is quantifiable within two to three weeks if someone is made accountable for it.
- Offer a decision structure: return to the next committee with (a) a re-run model using two eliminated positions and the 3% forecast, (b) a safety and quality business case with costed risk (incident history, scrap dollars, expedite spend), and (c) a phased option - lease the packer on one line, or defer to a smaller configuration - which reduces committed capital and produces real operating data.
- Set the test in advance: if the revised NPV is negative, the project proceeds only if the safety and customer-retention case is worth more than the negative NPV, and that judgement is made explicitly by the committee, minuted, not buried in an optimistic spreadsheet.
- Separate the sponsor from the analyst. Ask an independent finance resource to rebuild the model so the plant manager can advocate without owning the numbers.
Handled this way, you protect the capital allocation process and keep a plausible project alive on evidence.
3. The plant manager tells you privately that he will 'take it to the CEO' if you block him. How do you respond, and what do you do next?
Read the worked answer
Respond without defensiveness and without conceding the analysis. Say plainly: he should absolutely have access to the CEO; what you are asking is that both of you walk in with the same corrected numbers rather than two versions. Offer to co-present. This keeps the disagreement about the model, not about you, and removes the appearance of a finance function protecting turf.
Then do three things in order.
- Pre-brief the CFO the same day. The CFO should not learn about a contested $1.4 million request from the CEO. Give the two-sentence version plus the corrected NPV and the open items.
- Document. Circulate a short memo to the plant manager and CFO: assumptions reviewed, adjustments made, questions outstanding, and the date the revised submission is due. Contemporaneous, factual, unemotional.
- Preserve the decision rights. If the CEO decides to fund it anyway, that is a legitimate executive call, but it should be recorded as a strategic override with the negative NPV disclosed - not as a 'finance-approved' project. Ask for the override in writing.
What not to do: do not soften the analysis to avoid conflict, and do not withhold information from the committee to win. Both convert a technical disagreement into an integrity issue.
Watch for the pattern. If escalation becomes the standard response to any adverse analysis, that is a governance signal for the CFO and audit committee, not just a personality clash on one project.