M&A Valuation Methods in 2026: DCF, Comps, and Accretion/Dilution Worked Out

M&A valuation methods in 2026: DCF, comps, and accretion/dilution worked out

I'm Sean Yu, co-founder of Peony. I spend most of my time on the investor side of deals, and the single most common request I get from analysts, MBA students, and founders heading into a sale is some version of the same thing: "walk me through the valuation methods, but with numbers I can actually re-compute." The textbooks are dense, and half the accretion/dilution examples online have arithmetic that does not tie out.

So this is that walkthrough. Three worked examples, every figure shown, all of it internally consistent to the cent. You should be able to open a spreadsheet, follow along, and land on the same numbers I do. I run Peony, a data room company, and I see how much of this math ends up living inside the deal room, so at the end I will show where the valuation work actually happens. The bulk of this, though, is the mechanics: DCF, trading comps, precedent transactions, an LBO as a floor, and accretion/dilution as the deal-math overlay on top. One discipline note: there is exactly one live market number in this piece (an equity risk premium, attributed and dated), and I flag every rule of thumb that is a convention rather than a law of nature.

What are the main M&A valuation methods?

The four methods you see on almost every deal are discounted cash flow (DCF), trading comparables, precedent transactions, and an LBO analysis, with accretion/dilution laid on top as the deal-math overlay. Each one answers a slightly different question, which is exactly why bankers run several at once and plot them side by side.

A few things to notice. DCF is the only one of these that is intrinsic, meaning it does not lean on what anyone else paid or is trading at; the other three are relative or structural. Trading comps reflect minority, no-control pricing (you are buying one share on an exchange), while precedent transactions reflect control pricing and therefore carry a premium. The LBO is not really a "what is it worth" method so much as a "what would a private equity buyer pay" backstop, which makes it a natural floor. And accretion/dilution is not a valuation of the target at all; it is a test of what a proposed price and financing structure do to the acquirer's earnings per share.

The output of all of this is usually a football field: a horizontal bar chart showing the valuation range from each method (DCF, trading comps, precedents, LBO, and the 52-week trading range) stacked on top of each other. Where the bars overlap is the zone where a deal is likely to get done, and it frames the negotiation. No single method is "the answer"; the range is the answer.

How does the discounted cash flow method work?

DCF projects a company's unlevered free cash flow over an explicit horizon, discounts each year back to today at the weighted average cost of capital, then adds the present value of a terminal value that captures everything beyond that horizon. Enterprise value is the present value of the explicit cash flows plus the present value of the terminal value; from there you subtract net debt to get to equity value. The whole thing rests on a handful of assumptions, and it swings hard on two of them (the discount rate and the terminal growth), which is why the sensitivity table matters as much as the point estimate.

Here are the steps in order:

1. Project unlevered free cash flow (FCF) over an explicit horizon, typically 5 to 10 years. Unlevered FCF is roughly EBIT times (1 minus the tax rate), plus D&A, minus capex, minus the increase in net working capital. "Unlevered" means before debt financing, so the cash flow is available to all capital providers.
2. Discount at WACC. WACC = (E/V) times cost of equity plus (D/V) times the after-tax cost of debt, where E and D are the market values of equity and debt and V is their sum. Cost of equity typically comes from CAPM: risk-free rate plus beta times the equity risk premium.
3. Compute a terminal value at the end of the horizon, using either the Gordon growth (perpetuity growth) formula or an exit multiple, then discount it back to the present.
4. Sum to enterprise value. Enterprise value = present value of the explicit FCFs plus present value of the terminal value. Bridge to equity value by subtracting net debt and any other claims that rank ahead of common equity.
5. Sensitize. Show a WACC by terminal-growth (or WACC by exit-multiple) matrix, because the output moves a lot on both.

On the terminal value, the two standard formulas are worth stating cleanly, because most of the answer ends up here:

• Gordon growth (perpetuity growth): Terminal Value = Final-Year FCF times (1 + g), all divided by (WACC minus g).
• Exit multiple: Terminal Value = a final-year metric such as EBITDA, times an assumed exit multiple.

The perpetuity growth rate g is capped near long-run nominal GDP growth, roughly 2 to 3% in developed economies, on the principle that no company can grow faster than the economy indefinitely. If you plug in a g that implies your target eventually becomes larger than the whole economy, the model is telling you something is wrong, not that you have found a great business.

One market input, sourced and dated, because DCF needs it: for a U.S. valuation, Kroll lowered its recommended U.S. equity risk premium from 5.5% to 5.0%, effective as of September 2, 2025 and thereafter until further guidance. Kroll pairs that ERP with a risk-free rate equal to the higher of a normalized 3.5% or the spot 20-year U.S. Treasury yield as of the valuation date. That 5.0% flows into cost of equity through CAPM, then into WACC. Use a published, dated figure like this one rather than inventing a market number.

When does DCF mislead? A worked terminal-value example

DCF misleads when the terminal value is doing almost all of the work, which in most going-concern models it is. When the majority of enterprise value sits in the terminal value, the valuation is really a bet on the perpetuity assumptions (the growth rate, the exit multiple, the discount rate) rather than on the cash flows you carefully modeled. A commonly-cited rule of thumb is that terminal value represents roughly 60 to 80% of total enterprise value in going-concern DCFs. Rather than lean on the rule, let me show it with exact numbers.

Assume explicit FCF grows 10% per year for five years starting at $100, WACC is 10.0%, terminal growth g is 2.5%, and we discount at year-end (no mid-year convention).

There is a tidy artifact here: because FCF growth (10%) exactly equals WACC (10%), every explicit-year present value works out to the same $90.91, so the sum is easy to check (five times $90.91 is $454.55). That is a coincidence of the round assumptions, not a general rule.

Now the terminal value, using Gordon growth:

Even with a full five-year explicit forecast, about 73% of the value here lives in the terminal value, and only 26.8% comes from the modeled years. The DCF is mostly a bet on the perpetuity assumptions, not on the cash flows you spent your week forecasting. Nudge g from 2.5% to 3.0%, or WACC from 10% to 9.5%, and the answer moves a lot, all of it flowing through that terminal value. This is why a DCF without a sensitivity table is close to useless: the point estimate looks precise, but it is balanced on two inputs you are guessing at.

How do trading comps and precedent transactions differ?

Both methods apply multiples, but to different data, and the difference is control. Trading comps use the current market multiples of comparable public companies and reflect minority, no-control pricing. Precedent transactions use the multiples actually paid in past M&A deals for similar targets, so they embed a control premium. Because of that premium, precedent multiples typically run higher than trading comps.

Trading comparables work like this: you assemble a peer set of similar public companies, pull their multiples, and apply the peer-set multiples to your target's metrics. The common multiples are EV/EBITDA, EV/Revenue, EV/EBIT, and P/E. A discipline point that trips people up constantly: enterprise-value multiples (EV/EBITDA, EV/Revenue, EV/EBIT) are capital-structure-neutral and pair with pre-interest metrics, while P/E is an equity multiple and pairs with net income. Do not apply EV/EBITDA to net income, and do not apply P/E to EBITDA. Match the numerator to the denominator: enterprise-value multiples to enterprise-level metrics, equity multiples to equity-level metrics.

Precedent transactions apply the multiples paid in past deals for comparable targets. Because acquirers pay above the unaffected trading price to gain control and to capture synergies, those multiples carry a control premium, the extra a buyer pays over the standalone share price for control. That premium is real and observable in announced-deal offer premiums, but its size is deal-specific; I am not going to quote an "average control premium is X%" number, because a credible figure has to be sourced to a current dataset and I do not have one to stand behind here.

So the mental model is: trading comps tell you what the market pays for a share of a business like this today; precedent transactions tell you what a buyer paid for the whole business, control included. Both go on the football field.

Where does an LBO fit as a valuation floor?

An LBO analysis solves for the maximum price a financial sponsor could pay while still hitting its target return, given assumed leverage and an exit multiple. You fix the return you need and the debt you can raise, then back into the price. Typical inputs are illustrative rather than sourced, but a sponsor might target something like a 20 to 25% IRR or a 2 to 3x return on invested capital over roughly five years; treat those as conventional ranges, not as facts I am asserting about any specific fund.

The reason an LBO acts as a valuation floor is structural. A strategic buyer can usually pay more than a financial sponsor: the strategic captures synergies a sponsor cannot, often has a lower return hurdle, and does not need to sell the business in five years to realize a return. So the LBO-implied price is a "what a private equity buyer would pay" backstop. On the football field the LBO bar tends to sit toward the lower end of the range, which is exactly why deal teams use it as a sanity check on the floor rather than as the headline number.

What is accretion/dilution analysis?

Accretion/dilution analysis tests whether a deal raises or lowers the acquirer's pro-forma earnings per share in the first year or two after closing. A deal is accretive when pro-forma EPS is higher than the acquirer's standalone EPS, and dilutive when pro-forma EPS is lower. The mechanic is straightforward: build pro-forma net income and a pro-forma share count under the proposed structure, divide to get pro-forma EPS, and compare it to standalone EPS. The formula for the change is simply (pro-forma EPS minus standalone EPS) divided by standalone EPS.

Before running the full model, two rules of thumb tell you the sign:

• All-stock rule of thumb: the deal is accretive if the acquirer's P/E is higher than the target's P/E, because the acquirer is using expensive stock to buy cheaper earnings. It is dilutive if the acquirer's P/E is lower than the target's.
• All-cash (debt-funded) rule of thumb: the clean liftable version is this. A cash deal funded with debt is accretive when the target's earnings yield (net income divided by the equity purchase price, which equals 1 divided by the P/E) is greater than the after-tax cost of the acquisition financing. If the target's yield clears the financing cost, EPS goes up.

Those costs, for reference, are: after-tax cost of debt = interest rate on debt times (1 minus the buyer's tax rate); after-tax cost of cash = the foregone interest rate on that cash times (1 minus the tax rate); and the after-tax cost of issuing stock = 1 divided by the buyer's P/E multiple. The two worked examples below show the all-stock and the all-cash cases with the arithmetic fully out, and in both the rule of thumb and the full EPS math agree.

Worked example: an all-stock deal

Take a concrete pair. The acquirer has 100.0M shares and $200.0M of net income, so its EPS is $2.00; at a $40.00 share price its P/E is 20.0x. The target has $50.0M of net income and an agreed equity purchase price of $600.0M, paid entirely in stock. Assume no synergies.

The deal is accretive by +8.70%. And the rule of thumb calls it correctly: the acquirer's P/E of 20.0x is higher than the target's implied P/E of 12.0x, which predicts accretion. The intuition is that the acquirer is paying with stock the market values at 20 times earnings to buy a stream of earnings priced at only 12 times, so the swap adds to EPS.

Worked example: an all-cash deal funded by debt

Keep the same acquirer and target, but now fund the $600.0M purchase entirely with new debt at 6.0% pre-tax, with the acquirer's tax rate at 25%. No new shares are issued, so the whole effect runs through interest expense.

First, the rule of thumb:

The target's earnings yield of 8.33% comfortably clears the 4.50% after-tax cost of the debt, so the rule predicts accretion. Now the full EPS proof confirms it:

The deal is accretive by +11.50%, and the rule and the EPS math agree. The earnings-yield spread of 8.33% minus 4.50%, or 3.83% over the financing cost, is what drives the accretion. Notice this debt-funded version (+11.50%) beats the all-stock version (+8.70%) on the same purchase, which is the usual pattern when debt is cheaper than the target's earnings yield and no dilutive shares are issued. That is also exactly where the next section's warning comes from.

Why accretive does not mean value-creating

Here is the caveat that separates people who understand deal math from people who just run the model: accretion is not value creation. EPS accretion only means the deal's earnings-per-dollar-of-financing math is favorable in year one, often because debt is cheap or the target trades at a lower multiple. It says nothing about whether the acquirer overpaid, whether the projected synergies will actually show up, or whether the target's cash flows justify the price.

The debt-funded example above is a perfect illustration of the trap. It came out +11.50% accretive, better than the all-stock version, entirely because the debt was cheap relative to the target's earnings yield. But "cheap debt made EPS go up" is not the same statement as "this was a good acquisition." A deal can be EPS-accretive and still value-destructive: overpay for a business whose earnings are declining, fund it with cheap debt, and your EPS ticks up in year one while you have quietly loaded the balance sheet with debt against a shrinking asset. Conversely, a strategically excellent, genuinely value-creating deal can be short-term dilutive, for example a high-growth acquisition funded with equity, where EPS dips at first because you issued shares against earnings that have not compounded yet.

So treat accretion/dilution for what it is: an EPS-mechanics screen that tells you the first-year earnings direction under a given structure. Boards care about it, but it is not a verdict on value. The verdict comes from the DCF, the comps, the synergy case, and honest diligence on whether the earnings you are buying are real, which is the natural bridge to where all of that actually gets pressure-tested.

Where valuation work actually happens

A valuation is only as good as the inputs, and on a real deal those inputs do not live in the model. They live in the data room: the audited financials, the quality-of-earnings report that decides which add-backs survive and therefore what EBITDA the multiple gets applied to, the customer and contract data behind the revenue projections, and the working model itself. The DCF, the comps, and the accretion/dilution screen all sit downstream of whatever is in those files. If the QoE reverses an add-back you assumed, your terminal value moves. If a churned-cohort schedule undercuts the growth rate, your whole football field shifts left.

That is where I spend my day. I run Peony, a data room company used by 6,800+ customers, and the pattern I see is that valuation work is really document work wearing a spreadsheet costume. Two things matter for the numbers you have just walked through. First, buyers do not read a data room evenly; they re-open the schedules that make or break the price, and page-level analytics let a seller see exactly which financial exhibits a buyer keeps returning to (the QoE, the cohort file, the debt schedule) so you can get ahead of the questions. Those analytics are free on every Peony tier, including Free; they are not a paywalled add-on. Second, the model has to survive being shared. Emailing an .xlsx leaks your formulas and dies the moment someone edits a cell, which is why sharing a live financial model with an investor through an identity-bound link, rather than an attachment, keeps version control and lets you revoke access if the deal changes.

On specifics, since people always ask: link expiry and revoke work on every tier including Free; the $52/admin/month Data Room tier adds dynamic per-viewer watermarks that stamp each viewer's name and email over the render; and viewers are always free, so you are never paying for the buy-side to look. For how the financials fit the broader workstream, the M&A due diligence process guide walks the full sequence, and for a distressed target there is what happens when a company is liquidated. The honest starting point is the free tier: spin up a room, drop your model and QoE in, and watch what a reviewer actually opens. Full pricing is on the pricing page.

Peony is used by 6,800+ customers running exactly this kind of financial-diligence workflow, and the reason I keep the analytics free rather than gated is that seeing which schedule a buyer re-opens is often the difference between defending your number and getting re-traded.

Frequently asked questions

What are the main M&A valuation methods?

The four you see on almost every deal are discounted cash flow (DCF), trading comparables, precedent transactions, and an LBO analysis, with accretion/dilution laid on top as the deal-math overlay. DCF values a business off its own projected cash flows. Trading comps apply peer market multiples. Precedent transactions apply multiples paid in past deals, which embed a control premium. An LBO solves for what a financial sponsor could pay, so it acts as a valuation floor. Bankers plot all of them on a football field to frame the negotiating range.

How does the discounted cash flow method work?

DCF projects a company's unlevered free cash flow over an explicit horizon (typically 5 to 10 years), discounts each year back to today at the weighted average cost of capital (WACC), then adds the present value of a terminal value that captures everything beyond the horizon. Enterprise value equals the present value of the explicit cash flows plus the present value of the terminal value; you subtract net debt to reach equity value. Terminal value is computed with the Gordon growth formula or an exit multiple, and the answer swings hard on the WACC and growth assumptions.

What is the terminal value formula in a DCF?

There are two standard formulas. Gordon growth (perpetuity growth): Terminal Value = Final-Year FCF times (1 + g), divided by (WACC minus g), where g is the long-run growth rate. Exit multiple: Terminal Value = a final-year metric such as EBITDA times an assumed exit multiple. The perpetuity growth rate is capped near long-run nominal GDP growth (roughly 2 to 3% in developed economies), because no company can outgrow the economy forever. Whichever you use, discount the terminal value back to the present before adding it to the explicit cash flows.

Why can DCF be misleading?

Because in most going-concern DCFs the terminal value is the majority of enterprise value, so the output is mostly a bet on perpetual assumptions rather than the years you actually modeled. Terminal value commonly lands around 60 to 80% of total enterprise value. In the worked example on this page, a five-year forecast still puts 73.2% of enterprise value in the terminal value, leaving only 26.8% in the explicit cash flows. Small changes in the growth rate or WACC then move the valuation a lot, which is why every credible DCF ships with a sensitivity table.

How do trading comps and precedent transactions differ?

Both apply multiples, but to different data. Trading comps use the current market multiples of comparable public companies (EV/EBITDA, EV/Revenue, P/E) and reflect minority, no-control pricing. Precedent transactions use the multiples actually paid in past M&A deals for similar targets, so they embed a control premium, the extra an acquirer pays over the unaffected share price to gain control and synergies. Because of that premium, precedent multiples typically run higher than trading comps. One is what the market pays for a share; the other is what a buyer paid for the whole company.

What is accretion/dilution analysis?

Accretion/dilution analysis tests whether a deal raises or lowers the acquirer's pro-forma earnings per share (EPS) in the first year or two after closing. If pro-forma EPS is higher than the acquirer's standalone EPS, the deal is accretive; if it is lower, the deal is dilutive. You build pro-forma net income and pro-forma share count under the proposed structure, divide to get pro-forma EPS, and compare. It is a fast screen bankers run before deeper valuation work, but it measures EPS mechanics, not whether the price makes economic sense.

What is the accretion/dilution formula and rule of thumb?

Accretion equals (pro-forma EPS minus standalone EPS) divided by standalone EPS. Two rules of thumb predict the sign without a full model. All-stock: the deal is accretive if the acquirer's P/E is higher than the target's P/E, because expensive stock buys cheaper earnings. All-cash funded with debt: the deal is accretive if the target's earnings yield (net income divided by equity purchase price, which is 1 divided by the P/E) exceeds the after-tax cost of the financing. If the target's yield clears the financing cost, EPS goes up.

When is a deal accretive in an accretion/dilution model?

In an all-stock deal, it is accretive when the acquirer's P/E exceeds the target's P/E. In an all-cash deal funded by debt, it is accretive when the target's earnings yield beats the after-tax cost of debt. The worked example on this page shows both: a 20.0x acquirer buying a 12.0x target all-stock comes out +8.70% accretive, and funding the same purchase with 6.0% debt at a 25% tax rate (4.50% after tax) against an 8.33% target yield comes out +11.50% accretive. The earnings-yield spread over the financing cost is what drives the accretion.

Is an accretive deal always a good deal?

No. EPS accretion only means the deal's earnings-per-dollar-of-financing math is favorable in year one, often because debt is cheap or the target trades at a lower multiple. It says nothing about whether the acquirer overpaid, whether projected synergies will materialize, or whether the target's cash flows justify the price. A deal can be EPS-accretive yet value-destructive (overpaying for a declining business with cheap debt), and a strategically excellent deal can be short-term dilutive (a growth acquisition funded with equity). Accretion/dilution is a mechanics screen, not a verdict on value.

What equity risk premium should I use in a DCF in 2026?

For a U.S. valuation, Kroll lowered its recommended U.S. equity risk premium from 5.5% to 5.0%, effective as of September 2, 2025 and thereafter until further guidance. Kroll pairs that with a risk-free rate equal to the higher of a normalized 3.5% or the spot 20-year U.S. Treasury yield as of your valuation date. The equity risk premium feeds cost of equity through CAPM (risk-free rate plus beta times ERP), which in turn feeds WACC. Use a published, dated figure rather than inventing a market number, and note the as-of date in your model.

Where do deal teams share valuation models and financials securely?

Most share them inside the deal's data room rather than over email, because a live model or a quality-of-earnings file emailed as an attachment leaks formulas and loses version control the moment someone edits a cell. Peony, the data room company I co-founded, is used by 6,800+ customers; page-level analytics are free on every tier (so you can see which schedules a buyer re-opens), link expiry and revoke work on every tier including Free, the $52/admin/month Data Room tier adds dynamic per-viewer watermarks, and the free tier is the natural way to start. Viewers are always free.