Live Chat
The weighted average cost of capital is the blended rate a company pays for every dollar of financing it uses. You build it in four moves: work out what share of the financing comes from each source, work out what each source costs, cut the debt cost down for the interest tax deduction, then take the weighted average.
The formula takes a minute to memorize. Almost every point lost on a cost-of-capital problem is lost somewhere else — in a weight built from the wrong column of the balance sheet, in a coupon rate standing where a yield belonged, in a market return plugged into a slot that wanted a market premium. So what follows runs the whole chain on one company with the arithmetic on the page, then prices each of those mistakes in basis points.
- Weights come from market values, not the balance sheet. On the company below, book-value weights give 8.41% where market-value weights give 9.31% — same component costs, 90 basis points apart.
- Only debt gets the
(1 − T). Interest is deductible; preferred and common dividends are not. Tax-adjusting the preferred leg is a quiet, common slip. - Use the yield to maturity for the cost of debt, not the coupon rate. The coupon is a historical rate; the yield is what lenders demand today.
- The market risk premium is a premium, not a return. Using an expected market return of 9.78% where the 5.00% premium belonged moves this WACC from 9.31% to 13.42%.
- Make your weights sum to 1.00 before you go further. It is a two-second check, and it catches the one error that silently biases everything after it.
The formula, and the three legs it can have
Most textbooks print the two-leg version:
WACC = (E/V × Re) + (D/V × Rd × (1 − T))
That version assumes the company has only common stock and debt. Plenty of assigned problems hand you preferred stock as well, and then you need the three-leg form. OpenStax's Principles of Finance 2e writes it as D% × rd(1 − T) + P% × rpfd + E% × re, and the shape of that equation carries the single most useful fact in this whole topic: the tax adjustment is attached to the debt term and to nothing else.
The reason is not a convention. Interest is a deductible expense, so borrowing a dollar costs the company less than the lender receives. Dividends — preferred or common — are paid out of after-tax income, so no deduction exists to hand back. If your preferred leg has a (1 − T) on it, you have invented a tax break.
Four steps, in order, and each one feeds the next:
Market value of each financing source, divided by the market value of all of them together. These must sum to 1.00.
Usually CAPM: risk-free rate plus beta times the market risk premium. Some problems hand you a dividend-growth route instead.
Preferred is the fixed dividend over the current price. Debt is the yield to maturity on the company's bonds, then multiplied by (1 − T).
Multiply each cost by its weight and add. Keep the unrounded figures until the last line.
Step 1: build the weights from market values
Take Northgate Materials, a company invented for this walkthrough so every number in it can be checked against every other. It has 40 million common shares trading at $28.50, two million preferred shares trading at $50.00 that pay a fixed $4.00 dividend, and $300 million face value of bonds quoted at 94.5 — meaning the market pays $94.50 for each $100 of face value. Its marginal tax rate is 21% — the federal statutory rate, with state tax set aside for this example.
| Component | How its market value is built | Market value | Weight |
|---|---|---|---|
| Common stock (E) | 40,000,000 shares × $28.50 | $1,140,000,000 | 74.83% |
| Preferred stock (P) | 2,000,000 shares × $50.00 | $100,000,000 | 6.56% |
| Bonds (D) | $300,000,000 face × 0.945 | $283,500,000 | 18.61% |
| Total capital (V) | — | $1,523,500,000 | 100.00% |
Two things in that table are worth slowing down for. First, the equity value is shares outstanding times share price — not the common-equity line on the balance sheet, which records what shareholders paid historically. Second, the debt value is the bonds' market price, not their face value, because $283.5 million is what it would take to retire that debt today.
Now suppose you had built the weights from the balance sheet instead. Give Northgate a book common equity of $520 million, preferred carried at its $100 million par, and debt carried at its $300 million face. Total book capital is $920 million, so the weights become 56.52% equity, 10.87% preferred and 32.61% debt. Run the identical component costs through those weights and the answer comes out at 8.41% instead of 9.31%.
Book weights load the mix toward debt here, and debt is the cheap leg after tax, so the WACC comes out too low. A discount rate that is too low makes weak projects look worthwhile, which is the whole reason the market-value rule exists. Write that reason into your answer rather than just applying the rule.
Market value of equity is easy — shares times price. Market value of debt is often not, because bank loans do not trade. Aswath Damodaran's The Cost of Capital: The Swiss Army Knife of Finance (April 2016) gives two options. The shortcut is to assume the book value of debt equals market value, which he calls not a bad assumption if the debt is recent or if the cost of debt for the company has not changed much since the debt issuance
; the better option, which he sets out immediately after, is to build a market value from the expected interest payments, face value and maturity. Whichever you take, say in your answer which one it was.
Step 2: the cost of equity
Equity has no stated rate, so it has to be estimated. CAPM is the standard route:
Re = risk-free rate + (beta × market risk premium)
Northgate has a beta of 1.15. Take the risk-free rate as 4.78% and the expected return on the market as 9.78%. The market risk premium is the difference between those two — 9.78% − 4.78% = 5.00% — and that difference, not the 9.78%, is what gets multiplied by beta.
Re = 4.78% + (1.15 × 5.00%) = 4.78% + 5.75% = 10.53%
If a problem gives you an expected market return, subtract the risk-free rate before you multiply by beta. Skipping that subtraction here gives Re = 4.78% + (1.15 × 9.78%) = 16.03%, and carries Northgate's WACC to 13.42%. It is the largest single error in this article and the easiest to make, because both numbers are percentages and both are sitting in the same sentence of the question.
Some problems give you a dividend instead of a beta, which means they want the constant dividend growth route: next year's dividend divided by the current share price, plus the dividend growth rate. The two models usually land close to each other and almost never land on the same number. OpenStax puts it plainly — Often, these methods will produce similar estimates of the cost of capital; seldom will the two methods provide the same value.
Use the one your problem gives you the inputs for. If it gives you both, compute both, and say which you carried forward and why. That sentence is usually worth points on its own; if your course publishes a grading scheme for problem sets, it is worth reading how the credit is actually allocated before you decide how much working to show.
It is also worth knowing which model gets used outside the classroom. In the 2022 Duke CFO survey (March 2020 wave) reported in John Graham's Presidential Address: Corporate Finance and Reality (NBER Working Paper 29841, March 2022), Table III records that 82.61% of large firms always or almost always use CAPM to estimate the cost of equity, against 17.50% for the dividend discount model. Among small firms the CAPM figure was 26.58%. CAPM is the default because it is familiar, not because it is settled.
Step 3: the preferred and the debt
Preferred stock is the easiest leg. Its dividend is fixed, so its cost is that dividend divided by the price the market currently pays for it. Damodaran's advice, for a company with substantial preferred stock, is to keep it as a third component in the cost of capital and attach the preferred dividend yield (obtained by dividing the fixed dividend by the current price of preferred stock) to it as a cost
. For Northgate that is $4.00 ÷ $50.00 = 8.00%. No tax adjustment.
Debt takes the most work. What you want is the return lenders demand from this company today, which is the yield to maturity implied by the bond's current price — not the coupon rate printed on it years ago.
Northgate's bonds pay a 5.5% coupon semiannually, mature in 12 years, and trade at 94.5. So a financial calculator gets N = 24 periods, PV = −94.50, PMT = 2.75, FV = 100, and solving for the periodic rate returns 3.0775%. Double it for the annual bond-equivalent yield: 6.155%. The yield sits above the 5.5% coupon because the bond trades at a discount: lenders now want more than the coupon pays, so they will only buy it below face value.
Then the tax shield. The US federal corporate rate is set out in the IRS's Instructions for Form 1120 (2025), which tells a filer at Schedule J, Line 1a to Multiply taxable income (page 1, line 30) by 21% (0.21)
. So:
After-tax Rd = 6.155% × (1 − 0.21) = 4.862%
A tempting shortcut is to take interest expense from the income statement and divide it by debt on the balance sheet. Damodaran's verdict on that in the same 2016 paper: the book interest rate is close to useless as an estimate of the cost of debt
, because it records the rate the company borrowed at historically, not the rate it faces now.
The whole calculation, end to end
Every input is now in hand. Multiply each cost by its weight, using the unrounded weights, and add:
| Leg | Weight | Cost | Contribution to WACC |
|---|---|---|---|
| Common equity | 74.8277% | 10.530% | 7.879% |
| Preferred | 6.5638% | 8.000% | 0.525% |
| Debt, after tax | 18.6085% | 4.862% | 0.905% |
| WACC | 100.0000% | — | 9.31% |
Northgate's weighted average cost of capital is 9.31%. One observation worth carrying into your write-up: the equity leg contributes 7.88 of those 9.31 points, roughly 85% of the answer, from 74.83% of the capital. Equity is both the biggest weight and the most expensive rate, which is why the cost-of-equity estimate deserves most of your attention and most of your justification — and why an error in the debt leg is usually survivable while an error in the equity leg is not.
A note on rounding. Carrying the rounded two-decimal weights instead (0.7483, 0.0656, 0.1861) against the costs as printed in the table gives 9.3092%, which still rounds to 9.31%. That is luck, not method. Keep full precision through the chain and round once, at the end.
Where each input actually comes from
Assigned problems hand you the inputs. Case studies and capstone projects do not, and this is where students stall, so here is where each number lives.
The risk-free rate is a US Treasury yield, published daily by the Treasury itself in its Daily Treasury Par Yield Curve Rates table. The 10-year point is the usual choice for corporate work. On September 4, 2026, that rate was 4.78% — the figure used above. It moves every trading day, so pull the current one rather than reusing this one, and put the date in your answer.
The market risk premium is an estimate, and it is the input your answer is most exposed to. OpenStax reports that Typical market risk premiums used by financial managers range from 5% to 8%.
Damodaran's country risk data, last updated January 5, 2026, puts the US equity risk premium at 4.46%. Run 4.46% and 8% through Northgate and its WACC lands anywhere between 8.84% and 11.89% on identical company facts. A one-point change in the premium moves this WACC by 0.86 points. State the premium you used and where it came from; the number matters less than the fact that you sourced it.
Beta is not a fact either. OpenStax's own illustration is that in April 2021, the beta for IBM was reported as 0.97 by MarketWatch and as 1.25 by Yahoo! Finance
. At this article's 4.78% risk-free rate and a 5% premium, those two betas give costs of equity of 9.63% and 11.03%. Name your source.
The tax rate is the marginal rate on the next dollar of interest deduction, not the effective rate you might compute from a filed return. The federal statutory rate is 21%, and companies also pay state corporate income tax where they operate — California, for one, states in FTB Publication 1123 that C corporations are taxed on their net income at a rate of 8.84 percent by California.
Combined marginal rates therefore run above 21%. For coursework, use the rate the problem gives you and note the assumption.
Six errors, priced against the same base case
Each row below changes exactly one thing about the Northgate calculation and leaves everything else alone, so the last column is the cost of that mistake on its own.
| The mistake | WACC it produces | Error |
|---|---|---|
| Expected market return used where the market risk premium belonged | 13.42% | +411 bp |
| Book-value weights instead of market-value weights | 8.41% | −90 bp |
| No tax adjustment on the debt leg at all | 9.55% | +24 bp |
| Preferred stock folded in with common equity at the cost of equity | 9.48% | +17 bp |
| Tax adjustment applied to the preferred leg as well as the debt leg | 9.20% | −11 bp |
| Coupon rate of 5.5% used in place of the 6.155% yield to maturity | 9.21% | −10 bp |
| Correct calculation | 9.31% | — |
The ranking is the useful part. The bottom four errors are conceptual mistakes that barely move the number, which means a grader spots them by reading your working, not by checking your answer. The top two move the number enough to change a recommendation. If you have time for one check, make it the market-risk-premium subtraction, and then confirm your weights sum to 1.00.
That last check is not busywork. OpenStax's own worked example is a demonstration of it: Principles of Finance 2e establishes a debt weight of 24.4% and an equity weight of 75.6% in §17.1, restates both at the top of §17.3, then in the very next sentence prints the WACC line as WACC = 24% × 4.99% + 75.6% × 13.4% = 1.20% + 10.13% = 11.33%. Those printed weights add to 99.6%, and the book's own 24.4% figure gives 11.35% rather than 11.33%. The gap is two basis points and nothing in the chapter's method is wrong — but it is a published textbook losing a decimal in the one place a summing check would have caught it.
When a single WACC is the wrong rate
A company's WACC is the required return on assets of that company's average risk. Use it to discount a project riskier or safer than the firm as a whole and you are answering a question nobody asked. A stable utility evaluating a software venture at its own WACC will approve things it should not; a divisional analysis usually needs a rate built from comparable firms in that division's line of business instead.
The gap between theory and practice here is documented. Graham's NBER paper reports that firms' hurdle rates have carried roughly a six-point buffer above their cost of capital for two decades — in the 2022 survey (March 2019 wave), an average hurdle rate of 15% against an average WACC of 9% — and that in a separate June 2017 Duke CFO survey, only 1-in-5 companies say they accept all projects whose expected return exceeds the stated hurdle rate
. So the number you have just computed is the cost of capital, which is the input to a net present value calculation, not by itself the bar a real firm applies. If your assignment asks you to comment on that, the buffer is what to comment on.
Cost-of-capital questions are unforgiving in a specific way: one wrong input propagates silently to the end, and the final number gives no signal that it happened. Reconciling the weights, sourcing each input and writing down the assumptions is the part of the answer that still earns credit when a step goes wrong. If the deadline has closed the gap and you would rather have someone work through the cost-of-capital chain with you, our finance specialists prepare the full solution with the working shown, so what comes back is something you can follow and check rather than a number to copy.
