The base-case NPV in our five-warehouse model was strongly positive. The more useful part of the analysis was understanding how close the recommendation came to changing when the operating assumptions moved.
The final model produced a $140.96 million base NPV and 25.58% IRR at an 8.18% WACC. Those numbers support the project under the base assumptions, but they are not the decision by themselves.
Capital budgeting is an assumption map
The model separated merchandise revenue, membership revenue, warehouse ramp, cannibalization, variable operating costs, initial and maintenance CapEx, depreciation, working capital, salvage, taxes, and financing. Each one creates a different way for the project to disappoint.
What mattered was not only whether the spreadsheet could calculate NPV. It was which assumptions could move enough to erase the margin of safety.
The pessimistic case stayed positive—barely
In the combined pessimistic scenario, NPV fell to about $3.99 million and IRR to 9.14%. That is technically positive, but economically it is a very different decision from a $140.96 million base case.
A model with a barely positive downside case tells management that execution quality matters. Small misses in demand, cost, ramp timing, or capital spending can consume the remaining value.
Mature warehouse sales were the largest one-way driver
The final one-way sensitivity analysis moved mature sales from 80% to 120% of base. NPV moved from roughly $94.27 million to $187.66 million.
That does not mean revenue is the only risk. It means site-level demand and mature-store performance deserve especially strong operating controls: location selection, traffic assumptions, membership conversion, ramp monitoring, and early intervention when a store is tracking below plan.
The model changed the recommendation into a set of controls
Financing was a separate question from project value
The model also evaluated whether the company needed external financing at all. The $128.91 million project represented about 0.84% of the cash plus short-term investments used in the course analysis and about 1.64% of annual simple free cash flow.
That supported an internal-financing recommendation—not because debt is always bad, but because the project did not appear large enough to justify adding financing complexity for liquidity reasons alone.
Model hygiene changed the answer before sensitivity did
Earlier versions contained a membership-revenue overlap and tax-rate treatment that needed correction. That experience was a useful reminder: sophisticated scenario analysis does not rescue a model whose base cash-flow logic is wrong.