Capex Roi Evaluator

Capex ROI Evaluator

When to use

Use whenever a capital expenditure decision needs a return evaluation — new equipment, a facility investment, a technology buildout — especially replacing an evaluation that computes NPV at WACC and stops there, without checking the result against the company's actual hurdle rate, which is often set higher than WACC to account for risk and opportunity cost.

What it does

Computes NPV, IRR, payback period, and simple average ROIC for a capital project via a bundled calculator, checks IRR explicitly against the stated hurdle rate (not just WACC), and flags the specific case where a project would look approvable on NPV-at-WACC alone while actually failing the higher bar the hurdle rate represents.

Method

1. State the initial capex, useful life, and expected annual cash flows the investment is projected to generate — driver-based where possible, not a single assumed number.
2. State both WACC and the hurdle rate explicitly, as different numbers. WACC is the cost of capital; the hurdle rate is usually WACC plus a risk premium reflecting the specific project's risk and the opportunity cost of capital versus other available projects. Using WACC alone as the approval bar systematically approves riskier projects than the company actually intends to accept.
3. Run the bundled calculator (scripts/capex_roi.py) to get NPV at both WACC and hurdle rate, IRR, payback period, and simple ROIC.
4. Check IRR against the hurdle rate explicitly, not just whether NPV at WACC is positive — a project can show positive NPV at WACC while its IRR sits below the hurdle rate, meaning it clears the cost-of-capital bar but not the company's actual required-return bar.
5. Read the WACC-vs-hurdle NPV comparison as a specific flag, not a rounding difference — when NPV is positive at WACC but negative at the hurdle rate, that's exactly the situation where evaluating on WACC alone would approve a project the company's own stated standard would reject.
6. Check payback period against any stated capital constraints separately from NPV/IRR — a project with strong NPV but a very long payback might still be wrong for a business with tight near-term capital availability, a consideration NPV alone doesn't capture.
7. Document the hurdle rate's basis (why this project gets this premium over WACC) so the evaluation is defensible when someone asks why one project used a 12% hurdle and another used 15%.

Inputs

• Initial capital expenditure
• Useful life in years
• Projected annual cash flows over the useful life
• WACC and the project-specific hurdle rate, with the hurdle's basis stated
• Config saved as JSON matching the format documented at the top of scripts/capex_roi.py

Output format

NPV at WACC and at hurdle rate; IRR; payback period; simple average ROIC; explicit pass/fail read of IRR against the hurdle rate; flag when NPV is positive at WACC but negative at the hurdle rate.

Example

A project shows NPV of $197,266 at a 9% WACC, which alone would suggest approval. Checked against a 12% hurdle rate (the company's actual required return for this risk category), the IRR of 13.7% still clears it, so the project passes on the real bar too. A second project with weaker cash flows shows an IRR of 7.1% against the same 12% hurdle, a clear fail that a WACC-only NPV check might have missed entirely if its NPV at WACC alone had happened to be positive.

Common pitfalls

• Evaluating capex projects against WACC alone, approving projects that fail the company's actual, higher-risk-adjusted hurdle rate.
• Reporting NPV without IRR, losing the direct comparability against a stated required-return threshold.
• Ignoring payback period when the business has real near-term capital constraints that NPV and IRR alone don't capture.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: Natan-Mohart
• Source: Natan-Mohart/24-finance-skills-for-claude
• License: MIT

Install and usage instructions live in the source repository linked above.