NPV: Why the Discount Rate You Pick Matters More Than the Cash Flows You Forecast
A project requiring a $100,000 investment today, expected to return $28,000 a year for five years, looks like a clear yes at an 8% discount rate: net present value comes out to roughly $11,800. Discount those same, identical cash flows at 14% instead, nothing about the project changed, just the rate used to evaluate it, and the NPV flips to roughly negative $3,900. Same investment, same forecast, opposite decision.
This calculator discounts future cash flows to their present value. Here's the mechanics behind that discounting, why the rate chosen matters more than most people assume, and the places this calculation quietly goes wrong.
The formula and what it's actually doing
Net present value sums up every future cash flow, each one discounted back to today's dollars, and subtracts the initial investment: NPV = Σ [CFt / (1 + r)^t] − Initial Investment, where CFt is the cash flow in period t, r is the discount rate, and t is the number of periods from now. The core idea is that a dollar received in the future is worth less than a dollar in hand today, because that dollar could otherwise be earning a return elsewhere, and the discount rate is the stand-in for that foregone return.
Working the $100,000 example at 8%: each year's $28,000 gets divided by (1.08) raised to that year's power, and the five resulting present values sum to about $111,800. Subtract the $100,000 investment and NPV comes to roughly $11,800, positive, meaning the project is expected to create value above what the discount rate itself represents. Run the same cash flows at 14% and the present value of the inflows drops to about $96,100, for an NPV of roughly negative $3,900, below the bar the discount rate sets. Nothing about the project's actual cash flows changed between those two calculations. The entire swing came from the discount rate assumption alone.
Where the discount rate is actually supposed to come from
This is the part a bare NPV calculator can't supply on its own, and it's arguably more consequential than the cash flow forecast itself. The discount rate is meant to represent the return available on the next-best alternative use of the money, sometimes called the opportunity cost of capital, not an arbitrary number.
For a company evaluating an internal project, the standard building block is the weighted average cost of capital (WACC): a blend of the cost of equity and the after-tax cost of debt, weighted by how much of each the company actually uses to fund itself. The cost of equity is commonly estimated using the Capital Asset Pricing Model: risk-free rate plus a company-specific beta multiplied by the equity risk premium. As a current reference point, the 10-year U.S. Treasury yield, a common proxy for the risk-free rate, was trading around 4.7% in mid-August 2026, itself elevated compared to recent years and the subject of active market debate about whether it moves higher still. Add a typical long-run equity risk premium in the range of 5% to 6%, and a cost of equity for an average-risk company lands somewhere around 10% to 11% before blending in any (generally cheaper) cost of debt. Established, stable companies often land in the 7% to 10% range for a blended WACC; early-stage ventures or higher-risk projects are frequently evaluated at 15% to 25% or more, precisely because more uncertain cash flows warrant a bigger discount.
Plugging a personal savings account interest rate, a generic "safe" number, or a round figure picked without any basis into an NPV calculation defeats the purpose of the calculation. The discount rate needs to reflect the actual risk and opportunity cost of the specific decision being evaluated, not a convenient default.
NPV and IRR can disagree, and it's not a calculation error
The internal rate of return (IRR) is the discount rate at which NPV equals exactly zero, in the example above, working out to approximately 12.3%, the rate directly between the 8% and 14% scenarios shown earlier where NPV crosses from positive to negative. IRR is often used alongside NPV as a second way to evaluate a project, phrased as a percentage return rather than a dollar figure, which makes it easier to compare against a required hurdle rate at a glance.
The two methods can rank competing projects differently, and this isn't a bug, it reflects a real difference in what each measures. A small project with a high percentage return can show a higher IRR than a large project with a lower percentage return but a bigger absolute NPV, precisely because IRR ignores scale. When choosing between mutually exclusive projects (only one can be selected, not both), finance theory generally favors NPV, since it measures actual value created in dollar terms, which is what shareholders or a business ultimately care about, while IRR can favor a smaller project that happens to have a flashier percentage return but adds less total value.
Nominal and real cash flows have to match the discount rate, not mix
This is one of the more common and least obvious NPV errors: cash flow forecasts and the discount rate both need to be either nominal (including expected inflation) or real (stripped of inflation), consistently, never mixed. A nominal discount rate applied to inflation-adjusted (real) cash flows systematically overstates the discount applied, understating NPV; the reverse mismatch overstates it. If a project's cash flow forecasts already build in expected price increases year over year, the discount rate needs to reflect a nominal cost of capital (inclusive of expected inflation) to match. If the cash flows are expressed in today's purchasing power with no inflation assumption baked in, a real discount rate (nominal rate minus expected inflation) needs to be used instead. Mixing the two isn't a small rounding issue, it's an internally inconsistent calculation that produces a specific, predictable direction of error depending on which way the mismatch runs.
Timing assumptions move the number more than they look like they should
The standard NPV formula assumes each period's cash flow arrives in a single lump at the end of that period, which is a simplification, since real cash typically flows in throughout the year. Some practitioners use a "mid-year convention" instead, treating each year's cash flow as if it arrived at the midpoint, which reduces the effective discounting applied and increases NPV. Applying that convention to the $100,000 example at 8% raises the present value of the inflows enough to push NPV from about $11,800 to roughly $16,200, a difference of more than a third, without changing a single dollar of the underlying forecast. Whichever convention gets used, consistency matters more than which one is "correct," since comparing a project modeled with end-of-year timing against one modeled with mid-year timing isn't an apples-to-apples comparison, even if every other assumption is identical.
What NPV doesn't automatically account for
Sunk costs, money already spent before the decision point, don't belong in the cash flow forecast at all; NPV is meant to evaluate only the incremental cash flows still within the decision-maker's control going forward. Financing cash flows (loan proceeds, interest payments, debt repayment) are typically excluded from the operating cash flows being discounted, since the discount rate itself already accounts for the cost of financing through WACC; including both double-counts the cost of capital. And a positive NPV under one discount rate assumption says nothing about how sensitive that conclusion is to being wrong about the rate. Given how dramatically the earlier example flipped between 8% and 14%, running the calculation across a range of plausible discount rates, rather than trusting a single point estimate, is usually worth the extra few minutes.
Common questions
What discount rate should I actually use if I don't know my company's WACC? A reasonable starting point is the risk-free rate (a current Treasury yield) plus a risk premium that reflects how uncertain the specific project's cash flows are, generally landing somewhere between roughly 7% for very safe, established cash flows and 20%+ for speculative or early-stage ventures. This is a rough estimation approach, not a substitute for an actual cost of capital calculation for anything with real money at stake.
Why did my NPV change so much when I only adjusted the discount rate slightly? Because the discounting effect compounds over each future period; a small change in the annual rate gets raised to increasingly higher powers for cash flows further in the future, so longer-dated projects are considerably more sensitive to discount rate assumptions than short-dated ones.
Is a positive NPV always the right decision? Not automatically. NPV assumes the discount rate correctly captures the project's risk and that the cash flow forecasts are reasonable; a positive NPV built on an unrealistically low discount rate or an overly optimistic forecast isn't actually evidence of a good investment, it's evidence of optimistic inputs.
Should I use IRR or NPV to compare two different projects? NPV, when the projects are mutually exclusive and differ meaningfully in scale or cash flow timing, since it reflects the actual dollar value created rather than a percentage return that ignores size. IRR remains useful as a supplementary check, especially against a minimum required hurdle rate.
Does inflation need to be included in the cash flow forecast? Only if the discount rate is also expressed in nominal terms. The two need to match; a real discount rate pairs with inflation-free cash flow forecasts, and a nominal discount rate pairs with cash flows that already include expected inflation.
The 10-year Treasury yield (approximately 4.7% in mid-August 2026) is used here as an illustrative risk-free rate reference and changes daily with market conditions; check a live source such as the Federal Reserve's H.15 release or FRED (fred.stlouisfed.org) for the current figure before using it in an actual calculation. Typical WACC and equity risk premium ranges cited are general market approximations, not specific to any company, and vary by industry, capital structure, and market conditions. The NPV, IRR, and discounting formulas themselves are stable mathematical relationships and do not change over time. This is not financial or investment advice; a real capital budgeting decision should be evaluated using your organization's actual cost of capital, ideally with input from a finance professional.