Investment decision

Should we do this project? Run a capital allocation decision (equipment, expansion, acquisition, build-vs-buy) through the textbook CFO workflow: NPV at your hurdle rate, IRR vs. cost of capital, opportunity cost against a passive benchmark, and levered cashflow analysis if the project is debt-financed. Deterministic answers, not a gut call.

When to use this pack

You're evaluating a $500,000 equipment purchase returning $150,000/year for 5 years; a market expansion with $2M upfront and an uncertain return; an acquisition target with a forecasted cashflow stream; or a build-vs-buy decision with different upfront costs and operating profiles. Standard capital-budgeting rules say accept if NPV > 0 at your hurdle rate AND IRR > cost of capital — but the inputs (especially hurdle rate and the cashflow forecast) deserve sanity checks, which this pack walks the agent through layer by layer.

Tools in this pack

• Net present value (NPV) $0.001 POST /api/npv Compute the net present value of a stream of cashflows at a given discount rate. Index 0 is treated as t=0 (today, not discounted); indices 1..n are discounted by (1+rate)^t. Matches Excel's NPV but with the conventional t=0 treatment most finance textbooks use (Excel itself starts discounting at t=1 — see notes). Use for capital-budgeting decisions: positive NPV = creates value at the discount rate; negative = destroys value.
• Internal rate of return (IRR) $0.001 POST /api/irr Compute the internal rate of return (IRR) of a cashflow stream — the discount rate at which NPV = 0. Index 0 is treated as t=0 (typically the negative initial investment); indices 1..n are inflows in subsequent periods. Solved via Newton-Raphson with bisection fallback. Requires at least one positive and one negative cashflow (otherwise IRR is undefined). Multiple sign changes in the cashflows can produce multiple IRR roots — we return the first one found.
• Compound interest $0.001 POST /api/compound-interest Compute future value of a principal under compound interest. Returns future value, total interest earned, and the effective annual rate (APY) given the compounding frequency. Matches Excel's FV(rate, nper, 0, -principal) and the classic (1+r/n)^(nt) textbook formula.
• Loan payment $0.001 POST /api/loan-payment Compute the monthly (or per-period) payment on a fully-amortizing loan: mortgage, auto, student loan, business loan. Returns the periodic payment, total paid over the term, and total interest. Matches Excel's PMT(rate, nper, -principal). Use this when you just need the payment number, not the full per-period schedule (see amortization).
• Amortization schedule $0.001 POST /api/amortization Build the full per-period amortization schedule for a fully-amortizing loan. Each row reports the period number, payment, the principal vs. interest split for that payment, and the remaining balance after that payment. Use this when the user wants to see how interest tapers over the life of the loan, or to model an extra-payment scenario by reading the balance at any period.

Workflow

1. Build the cashflow stream: index 0 = upfront investment (negative), 1..n = expected annual cashflows (positive), with any salvage/terminal value rolled into the final year. Pass to npv with discountRate = your hurdle rate (typically 8-12% for a small business, 10-15% for VC-backed risk, your weighted average cost of capital if you have one). Positive NPV = the project creates value above your hurdle; negative = it destroys value. This is the primary accept/reject signal.
2. Call irr on the same cashflow stream. The IRR is the discount rate at which NPV = 0 — i.e., the project's effective annualized return. Accept if IRR > hurdle rate; reject if IRR < hurdle. If the response has converged=false, the cashflow shape has multiple sign changes (common with mid-project re-investments) and NPV is the more reliable metric — flag the IRR as indicative not definitive.
3. Sanity-check against the passive alternative with compound-interest. Take the same upfront capital, invest at your benchmark rate (7-8% for long-run equity, 4-5% for bonds, your actual savings rate for cash), project forward over the same horizon. If the project's NPV + initial investment doesn't beat the passive future value, the project is destroying value relative to doing nothing — even if NPV at hurdle rate is positive. This catches projects that 'pass NPV' only because the hurdle rate is set unrealistically low.
4. If the project will be debt-financed (most real-world deals are not all-equity), call loan-payment to compute the periodic debt service. Subtract this from the project's annual operating cashflow to get the levered free cashflow to equity. Then re-run npv and irr on the *levered* stream (index 0 = your equity check, not the full purchase price). Leverage almost always boosts IRR (positive leverage when project yield > debt cost) and increases risk — surface both numbers so the user sees the trade-off.
5. Call amortization on the financing loan to get the year-by-year interest + principal split. The interest expense is typically tax-deductible — multiply by your tax rate to get the annual tax shield, which improves the levered cashflows. The remaining balance at each year is what you'd owe if you sold/refinanced — useful for modeling an early exit or refinance scenario. Skip if the project is all-equity; required if you want to model the levered IRR honestly.

Run it in Claude

claude mcp add agent402 -s user -- npx -y agent402-mcp@latest

Then paste this prompt into Claude:

Evaluate this capital project using Agent402: $500,000 equipment purchase returning $150,000/year for 5 years with $50,000 salvage value at the end. Use a 10% hurdle rate. (1) Build cashflows = [-500000, 150000, 150000, 150000, 150000, 200000] (year 5 includes salvage). Call npv at discountRate=0.10 — record the NPV. (2) Call irr on the same cashflows — record the IRR (it should be ~17-18% on these numbers; converged should be true). (3) Sanity-check the passive alternative: call compound-interest(principal=500000, annualRate=0.07, years=5, compoundingPerYear=1) — compare the future value of the cashflow scenario (cumulative undiscounted = $750k + $50k = $800k) against the passive S&P 7% future value (~$701k). If the project beats passive even before discounting, that's a real positive signal beyond NPV. (4) Model financing: if a $400k loan at 8% for 5 years funds most of it, call loan-payment(400000, 0.08, 5). Compute the annual debt service (payment × 12); subtract from $150k cashflow → levered cashflow. Build levered stream = [-100000, leveredCF, leveredCF, leveredCF, leveredCF, leveredCF + 50000] and re-run npv + irr on this — the levered IRR will be meaningfully higher than the unlevered, reflecting the equity returns. (5) Call amortization(400000, 0.08, 5, maxRows=5) for the per-year interest schedule (for tax-shield modeling). (6) Return: {unleveredNpv, unleveredIrr, passiveAlternativeFV, leveredNpv, leveredIrr, recommendation: "ACCEPT"|"REJECT", reasoning}. All five tools are free over PoW.

← All skill packs