DCF I: Framework, Cash Flows, and Mechanics

A discounted cash flow analysis answers the one question no market-based method can: what a company is worth on its own fundamentals, independent of what the market happens to be paying for it or its peers. Trading comps read current market pricing and precedent transactions read past deal pricing; the DCF builds value from the cash the business itself is projected to generate, the growth of that cash, and the risk of actually receiving it. That independence makes it the intrinsic anchor of every valuation, and it makes "walk me through a DCF" the single most asked technical question in investment banking interviews.

The machine has three components: a set of projected free cash flows, a discount rate, and a terminal value. This read owns the framework logic, the projections, both definitions of free cash flow, every discounting mechanic, the bridge from enterprise value to a per-share number, and the sensitivity work that turns a point estimate into a defensible range. The construction of the discount rate (WACC) and the terminal value is the subject of the next read; here they appear as named inputs at summary depth.

Why the DCF Works: The Time Value of Money

The entire framework rests on one idea: a dollar in hand today is worth more than the same dollar promised later. Two forces drive this.

• •Opportunity cost. Cash in hand can be invested. $100 today at a 10 percent return becomes $110 in a year, so receiving $100 one year from now is worth only about $91 today (because $91 invested at 10 percent grows to roughly $100). Discounting simply reverses compounding.
• •Risk. A projected cash flow may never arrive: the company misses its plan, the economy turns, regulation shifts. The further out the cash flow, the less certain it is. The discount rate charges for this uncertainty, and riskier cash flows are worth less today.

Discounting at a rate that reflects both forces converts every future cash flow into its present value, and the sum of those present values is what the business is worth today.

Intrinsic, Not Relative

Comps and precedents are relative valuations: they price the company off what similar assets currently fetch. The DCF is an intrinsic valuation, built entirely from the company's own characteristics. That independence cuts both ways. In a sector bubble, comps inherit the inflation while a DCF on realistic assumptions shows a lower, defensible number; in a panic, the DCF can show that fundamentals support more value than the depressed share price implies. When the DCF and the market disagree, that divergence is not a failure of either method. It is information: either the model's assumptions are off, or the market is mispricing the company, and working out which is one of the most valuable analytical exercises in banking.

The catch is that the intrinsic answer is only as good as the assumptions behind it. Two competent analysts can build DCFs on the same company and land 30 percent apart with fully defensible inputs. Do not confuse analytical with objective: every growth rate, margin, and discount rate input is a judgment call. The DCF's real virtue is transparency, not objectivity. Every assumption is explicit and can be challenged one at a time, which is impossible with a comps-derived value where the market's assumptions are invisible.

The Limitations Interviewers Ask About

"What are the weaknesses of a DCF?" is a standard follow-up, and the answer reveals whether you understand the method or memorized it.

• •Extreme assumption sensitivity. A 0.5 percent move in the discount rate or a 0.25 percent move in the terminal growth rate can shift the output by 15 to 25 percent. The DCF produces a range, never a point.
• •Terminal value dominance. The terminal value typically carries 60 to 80 percent of the total, so most of the answer rests on views about the distant future, precisely where forecasts are least reliable.
• •Projection uncertainty. Five-year forecasts are speculative for most companies; for volatile industries even two-year forecasts are shaky.
• •Garbage in, garbage out. No amount of mechanical rigor rescues bad inputs: optimistic management projections or a cherry-picked discount rate produce a wrong answer no matter how clean the model is. The same transparency that makes the DCF auditable also makes every input manipulable.

Despite all of this, the DCF stays essential because it is the only method that forces a complete, assumption-by-assumption view of the company's future. That discipline identifies the true value drivers even when the specific output is uncertain.

The Five Steps

The canonical answer to "walk me through a DCF" has five steps:

1. 1.Project unlevered free cash flow (UFCF) year by year over an explicit forecast period, typically 5 to 10 years.
2. 2.Calculate the terminal value capturing all cash flows beyond the forecast period, using the perpetuity growth method or the exit multiple method.
3. 3.Discount every projected cash flow and the terminal value to present value at WACC.
4. 4.Sum the present values to get implied enterprise value.
5. 5.Bridge from enterprise value to equity value and divide by diluted shares to get an implied price per share.

The forecast horizon follows the business. Mature, stable companies (utilities, consumer staples) reach steady state quickly, so 5 years is usually enough. High-growth or transforming businesses need 7 to 10 years so the model captures the full trajectory before maturity. Whose projections sit in the model also matters: a sell-side engagement typically runs on the client's management case (the most optimistic defensible view), a buy-side analysis runs on the acquirer's own diligence-based projections (usually more conservative), and a pitchbook runs on consensus estimates and public data because nothing internal is available yet. One company can produce three very different implied values depending on whose numbers drive the model, so always know which basis you are looking at.

A compact end-to-end pass shows the shape of the output. Take projected UFCFs of $100M, $110M, $121M, $133M, and $146M over five years, a terminal value of $2,000M (a 10x exit multiple on Year 5 EBITDA of $200M), and a 10 percent WACC:

• •Each year's present value works out to roughly $91M (about $454M in total), because in this example the cash flows grow at exactly the discount rate.
• •The present value of the terminal value is $1,242M, giving total implied enterprise value of approximately $1,697M. The terminal value is 73 percent of the total, inside the typical 60 to 80 percent band.
• •Subtracting $300M of net debt and dividing by 50 million diluted shares gives an implied $27.94 per share.

Every remaining section of this read unpacks one of those steps.

Step 1: Projecting the Business

The projections are the engine of the model. Every downstream number flows from them, and almost every line item is modeled as a function of revenue, so errors in the revenue forecast compound through the whole model. A sophisticated discount rate on top of an unrealistic revenue line is worthless, which is why experienced bankers spend their time on growth and margin assumptions rather than on the mechanical steps.

Revenue: Top-Down vs Bottom-Up

A top-down forecast starts from the size of the market and works down to the company:

The total addressable market (TAM) is the full revenue opportunity if the company captured its entire target market, sourced from industry research firms, management materials, and the analyst's own work. TAM is only useful when scoped honestly: a startup claiming a $500 billion TAM by defining its market as all enterprise software spending has said nothing. Top-down works best for large, well-defined markets and early-stage companies, but it is imprecise because the link between the broad market and one company's revenue is indirect.

A bottom-up forecast builds revenue from the company's own operating drivers:

• •Units times price, for product businesses: project volumes from capacity, demand, and competitive position, then apply expected pricing.
• •Customers times ARPU, for subscription and recurring-revenue businesses: project the customer count (new adds minus churn) and average revenue per user separately.
• •Segment by segment for diversified companies, since divisions grow at different rates.

Practitioners prefer bottom-up because each assumption is independently defensible. "8 percent growth" is one opaque number; "5 percent volume growth plus 3 percent pricing" is two transparent claims that can each be challenged. Strong models run both and compare: if the bottom-up build implies 15 percent growth while the market grows 5 percent, the company must be taking meaningful share, and the analyst has to decide whether that is credible. When asked why you assumed a given growth rate, the answer is never "that is the historical rate"; it is a decomposition into named drivers.

Where the Assumptions Come From

• •Historical performance: the most objective anchor. A company that has grown 8 to 12 percent for five years rarely jumps to 25 percent without a specific catalyst.
• •Management guidance: a useful baseline, but biased; management tends to guide conservatively on revenue and optimistically on margins, so stress-test it.
• •Consensus estimates: sell-side forecasts cover the next 2 to 3 years; beyond that window the analyst is on their own.
• •Industry and macro context: sector growth, competitive dynamics, regulation, and macro assumptions frame what is plausible for the company.

Margins and Operating Leverage

Gross margin is projected off the historical trajectory, adjusted for mix shift, input costs, pricing power, and scale. Operating expenses are usually modeled as a percentage of revenue, with three standard adjustments. Operating leverage is the big one: when fixed costs dominate the cost structure, incremental revenue drops disproportionately to profit, so margins expand as revenue scales (software is the classic case, where serving one more customer costs almost nothing). Investment-phase dynamics cut the other way: heavy sales and marketing spend depresses margins today in exchange for growth. Management guidance on headcount and cost structure fills in the rest.

The EBITDA margin trajectory is one of the most scrutinized assumption sets in the model: stable for mature businesses, expanding where operating leverage or efficiency programs bite, compressing under competition or input-cost pressure. The terminal-year margin must be a normalized, steady-state level, not a cyclical peak or trough, because it feeds the terminal value and therefore drives a disproportionate share of the answer. Beware the margin expansion trap: projecting margins from 20 to 30 percent over five years without naming the driver (leverage, cost program, pricing, mix) inflates the output and will not survive review.

CapEx and Working Capital

Capital expenditure splits into maintenance CapEx, the unavoidable spend that sustains the current asset base, and growth CapEx, the discretionary spend that expands it. Growth CapEx is typically front-loaded in the projection and fades as the company matures. In the terminal year the company is assumed to be at steady state, so CapEx should roughly equal D&A plus a modest increment for terminal growth; carrying an elevated investment program into the terminal year artificially depresses terminal cash flow and understates value. CapEx is modeled as a percentage of revenue, as a ratio to D&A (1.0x means maintaining the asset base, above 1.0x means expanding), or bottom-up from known project budgets. Intensity varies enormously: capital-heavy industries spend 10 to 20 percent of revenue, asset-light ones 1 to 3 percent.

Net working capital (NWC), current operating assets minus current operating liabilities, consumes cash as the business grows: more sales mean more receivables and more inventory to fund. The standard build projects each component on its own driver: receivables on days sales outstanding (DSO), inventory on days inventory outstanding (DIO, driven by COGS), payables on days payable outstanding (DPO). The cash flow impact each year is the change in NWC versus the prior year, so a growing company with stable ratios consumes cash roughly in proportion to its growth. Two traps: projecting flat NWC in the terminal year (implying zero working capital investment despite perpetual growth) overstates terminal cash flow, and some business models break the stable-ratio assumption entirely. Retailers swing seasonally, project businesses have milestone-driven receivables, and subscription models collecting cash upfront can run negative working capital, where growth actually generates cash rather than consuming it.

Unlevered Free Cash Flow

UFCF is the cash flow at the heart of the standard DCF: the cash the operations generate for all capital providers, debt and equity alike, after operating costs, taxes, and reinvestment.

Equivalently, defining NOPAT (net operating profit after taxes) as EBIT times one minus the tax rate:

Because no interest is deducted, UFCF is capital-structure-neutral: it pairs with WACC, the blended cost of all capital, and produces enterprise value. Starting from EBIT rather than EBITDA changes nothing, since D&A is added back in the next line; the two routes are algebraically identical.

The Tax Nuance Interviewers Probe

The formula taxes EBIT at the full rate, computing the hypothetical taxes the company would pay with no debt. The company's actual tax bill is lower, because interest is deductible and creates a tax shield. The unlevered DCF does not ignore that benefit; it simply enters elsewhere. The after-tax cost of debt inside WACC lowers the discount rate and raises present value, so the shield's value arrives through the denominator. Deducting interest before taxes in the cash flows as well would count the shield twice: once as lower taxes in the numerator, once as a lower rate in the denominator.

The rate itself should be the marginal tax rate, not the effective rate off the income statement. The US statutory rate is 21 percent federal, typically 25 to 27 percent combined with state taxes. The effective rate is polluted by timing differences, one-off items, and international structures that will not persist, and in the terminal year especially, only the sustainable marginal rate makes sense.

D&A: An Add-Back, Not Free Cash

D&A is a non-cash expense: recording depreciation moves no cash. It is added back purely to reverse a non-cash deduction embedded in EBIT, not because it creates cash. The real cash left when the asset was bought, through CapEx. That is why the add-back and the CapEx subtraction are linked: CapEx minus D&A is the company's net investment in its asset base. CapEx above D&A means the asset base is expanding; D&A persistently above CapEx means assets are wearing out faster than they are replaced, a signal of underinvestment that eventually erodes earning power.

CapEx and the change in NWC are subtracted because both are real cash outflows required to keep the business running and growing, even though CapEx hits the income statement only gradually through depreciation. An NWC increase consumes cash; a decrease (faster collections, leaner inventory, longer payment terms) releases it.

A Worked Build

Take a Year 1 projection: revenue $1,000M, EBITDA $250M (a 25 percent margin), D&A $50M so EBIT is $200M, a 25 percent tax rate, CapEx of $60M, and a $15M increase in NWC.

1. 1.NOPAT = $200M times 0.75 = $150M
2. 2.Add back D&A: plus $50M
3. 3.Subtract CapEx: minus $60M
4. 4.Subtract the NWC increase: minus $15M
5. 5.UFCF = $125M

That $125M is what the business hands its capital providers after paying its costs, its taxes, and its reinvestment bill. It is the number discounted at WACC for that year.

Beyond the Core Formula

Detailed models layer in further items, all governed by one principle: include every real operating cash flow, exclude everything non-cash or financing-related.

• •Stock-based compensation: non-cash, so some analysts add it back like D&A; others treat it as a real cost borne through dilution. The debate is unresolved and treated fully in the normalization read.
• •Deferred taxes: adjusting for the gap between cash taxes and book tax expense converts UFCF to an actual-cash-taxes basis.
• •Other non-cash items buried in EBIT (write-downs, non-cash restructuring) get added back.
• •Changes in other long-term operating items (deferred revenue, pension obligations) that the standard NWC lines miss.
• •One-time charges may be added back to normalize UFCF, with judgment: a genuine one-off factory closure qualifies; "one-time" restructuring taken every other year does not.

A useful owner's test resolves edge cases: if you owned the whole business and pulled out every dollar it generated while keeping it running and growing on plan, what would you have at year-end? That number is UFCF.

UFCF as a Quality Metric

The UFCF conversion ratio, UFCF divided by EBITDA, measures how much operating profit becomes actual cash. Above 50 to 60 percent is healthy for most industries; below 30 to 40 percent signals a business that eats its own earnings through reinvestment. This is why identical EBITDA does not mean identical value: a software company and an industrial manufacturer each earning $500 million of EBITDA might convert 70 to 80 percent and 35 to 45 percent respectively, so the software business generates nearly twice the cash per EBITDA dollar and deserves the premium multiple. The same logic explains why EV/EBITDA misleads for capital-intensive businesses, and why the terminal-year conversion ratio deserves a sanity check before you trust the output.

Levered Free Cash Flow and the Equity DCF

The levered counterpart answers a different question: not what the whole business is worth, but what the equity is worth. Levered free cash flow (LFCF), also called free cash flow to equity (FCFE), is the cash left for shareholders after all debt obligations are serviced.

Or, starting from UFCF:

Two structural differences from UFCF: the build starts from net income, so interest has already been deducted and taxes are the actual bill including the debt tax shield; and net borrowings (new debt raised minus debt repaid) are added, because financing inflows are cash available to equity holders. LFCF is discounted at the cost of equity, not WACC, and produces equity value directly, with no bridging step.

Why the Unlevered DCF Dominates

The direct path to equity value looks simpler, but the inputs are far messier, which is why the unlevered approach is the investment banking standard.

• •The capital structure must be projected forward. A levered DCF needs a full debt schedule: balances, maturities, refinancings, interest rates, and new issuance for every projection year. The unlevered DCF captures all of that in a single WACC set once.
• •Circularity. Equity value depends on levered cash flows, which depend on interest, which depends on debt levels, which depend on the capital structure, which depends on equity value. Breaking the loop requires iteration or a fixed debt schedule.
• •Comparability. LFCF embeds each company's specific leverage, so identical businesses with different debt loads show different LFCFs. The unlevered approach strips that out, which matters when the DCF sits next to capital-structure-neutral comps.

Built correctly on consistent assumptions, the two approaches must agree: enterprise value from the unlevered DCF minus net debt should equal the equity value from the levered DCF. A material gap means one of the models is wrong.

When the Levered DCF Is the Right Tool

• •Financial institutions. For banks and insurers, debt is not financing; it is the operating asset. Deposits and borrowings fund the lending book, and interest expense is a core operating cost, so separating operating from financing cash flows is impossible. FIG groups value banks on levered cash flows, most commonly through the dividend discount model, a levered DCF variant that discounts dividendable earnings at the cost of equity.
• •LBO modeling. An LBO model is essentially a levered DCF with a finite holding period and an explicit exit: the entire exercise is the debt schedule and what it does to equity returns.
• •Rapidly changing capital structures. A company de-levering after a buyout or gearing up for an acquisition is captured more precisely with explicit debt modeling than with a constant WACC.

The pairing rule is absolute, and interviewers test it directly. UFCF pairs with WACC and yields enterprise value; LFCF pairs with the cost of equity and yields equity value. The cost of equity always exceeds WACC (equity holders are paid last and bear more risk), so discounting UFCF at the cost of equity understates value, and discounting LFCF at WACC overstates it.

Terminal Value and WACC, in Brief

Both inputs are built in the next read; the DCF cannot run without them, so here is what each one is.

The terminal value captures every cash flow beyond the explicit forecast, out to infinity, and typically carries 60 to 80 percent of total enterprise value. Two methods exist. The perpetuity growth method assumes UFCF grows at a constant rate $g$ forever:

with $g$ typically 2 to 3 percent, in line with long-run GDP growth and inflation, and brutally sensitive: half a point of $g$ moves the terminal value 15 to 20 percent. The exit multiple method assumes a sale at the end of the forecast at a market-based multiple:

with the multiple taken from the current peer trading range. It is more intuitive but imports market pricing into a supposedly intrinsic method. Practitioners compute both and cross-check each against the other: the implied perpetuity growth rate behind an exit multiple (above 4 to 5 percent means the company eventually outgrows the economy) and the implied exit multiple behind a growth rate (25x EBITDA is a red flag in most sectors). Many desks present the exit multiple as the base case because clients find it tangible, with perpetuity growth as the sanity check; either order is fine if both are shown and the assumptions are transparent.

The weighted average cost of capital is the blended return all capital providers require:

where E and D are the market values of equity and debt, $r_e$ is the cost of equity (built from CAPM), $r_d$ is the pre-tax cost of debt, and $t$ is the marginal tax rate; the $(1 - t)$ term on debt is where the tax shield lives. After the terminal assumptions, WACC is the most powerful variable in the model: one percentage point of WACC moves implied enterprise value 10 to 15 percent, because it compresses every year's cash flow and the terminal value simultaneously. This is also how interest rates transmit into valuations: a higher risk-free rate raises both the cost of equity and the cost of debt, the higher WACC hits the terminal value hardest (it is discounted over the full period and is the largest component), and equity values fall disproportionately.

Discounting Mechanics and the Mid-Year Convention

Each projected cash flow converts to present value through the discount factor:

where $t$ is the number of years from the valuation date to receipt. The terminal value discounts from the end of the forecast period:

with $n$ the length of the projection period. Implied enterprise value is the sum of everything:

Year-End vs Mid-Year

The exponent $t$ encodes an assumption about when cash arrives. The year-end convention treats each year's cash flow as a lump sum on December 31: exponents of 1, 2, 3. Simple, but wrong in a specific direction, because businesses generate cash continuously through the year, and pretending it all lands at year-end discounts it for too long.

The mid-year convention, standard in investment banking, assumes each year's cash flow arrives at the midpoint of the year: exponents of 0.5, 1.5, 2.5, and so on. Every cash flow is discounted for half a year less, so present value comes out higher, typically by 3 to 5 percent of enterprise value.

The terminal value is the debated case. One school (the more common one) keeps the terminal value at exponent $n$, treating it as a point-in-time lump sum at the end of the forecast even while the explicit flows use mid-year exponents. The other applies the mid-year logic throughout and uses $n - 0.5$. Both are defensible, the gap between them is usually under 2 to 3 percent of enterprise value, and the real rule is consistency: pick a convention, apply it everywhere, and document it. Mixing conventions inside one model is the actual error.

A Worked Discounting Table

A 5-year DCF at 10 percent WACC, UFCF starting at $100M and growing 5 percent annually, terminal value of $2,000M, mid-year convention with the terminal value at year-end (exponent 5.0):

Under the pure year-end convention the total is approximately $1,657M, about 1.2 percent lower. The uplift here is smaller than the typical 3 to 5 percent because the terminal value, three quarters of the total, keeps its year-end exponent under this treatment; when the mid-year shift is applied to the terminal value as well, the whole valuation moves by close to the half-year discount saving, which is where the larger quoted range comes from.

One implementation note: never hardcode the exponents. Build them as a formula from the year number (year minus 0.5 for mid-year, year for year-end, ideally behind a toggle), so extending the forecast period cannot silently leave stale exponents behind.

From Enterprise Value to Equity Value per Share

The DCF's raw output is implied enterprise value, but the question clients, MDs, and interviewers actually care about is what it means for the share price. The conversion is the EV bridge applied in reverse:

Every bridge item comes from the most recent balance sheet as of the valuation date: total debt is all interest-bearing obligations, preferred equity at liquidation value, minority interests at book (or fair value if estimable), cash adjusted for restricted or operating balances where appropriate, and the share count is diluted shares from the treasury stock method at the current share price.

The timing rule matters: use current balance sheet data, never terminal-year projections. The DCF expresses value as of today (every cash flow was discounted to the present), so the claims subtracted from it must also be measured today. The terminal year's projected debt balance is irrelevant; the company's future cash generation and financing capacity are already inside the discounted cash flows.

Where the Bridge Goes Wrong

• •Basic instead of diluted shares. Options, warrants, RSUs, and converts dilute the equity; a basic share count makes the per-share output look better than it is.
• •Skipping preferred equity or minority interests. Small lines that are easy to miss and material for utilities, financial institutions, and conglomerates; leaving them out inflates equity value.
• •Double-counting non-operating assets. If the projections already include cash flows from a non-core asset, adding that asset's value again in the bridge counts it twice.
• •Stale data. Presenting the DCF a quarter later without refreshing debt, cash, and share counts anchors the per-share output to numbers that no longer exist.

A full pass: the DCF implies $5.2 billion of enterprise value; the balance sheet shows $1.2 billion of total debt, $100 million of preferred, $50 million of minority interests, $400 million of cash, and 200 million diluted shares. Implied equity value is 5.2 minus 1.2 minus 0.1 minus 0.05 plus 0.4, or $4.25 billion, and the implied share price is $21.25. Against an $18.00 market price the DCF says the stock is undervalued by roughly 18 percent; against $24.00 it says the stock is overvalued by roughly 11 percent.

Sensitivity, Scenarios, and Sanity Checks

A DCF that produces one number is not finished. Given how hard the output leans on WACC and the terminal assumptions, a single figure implies precision the method cannot deliver, and every DCF that reaches a pitchbook, board presentation, or fairness opinion ships with sensitivity analysis attached.

The Two-Variable Sensitivity Table

The standard exhibit is a sensitivity table: a grid, built with Excel's data table function, showing the implied output across ranges of two inputs, base case highlighted in the middle. The classic pairings match the terminal method: WACC against the terminal growth rate (WACC in 0.5 percent steps, growth in 0.25 percent steps) when using perpetuity growth, and WACC against the exit multiple (in 0.5x steps) when using an exit multiple. An implied-share-price grid across WACC and exit multiples looks like this:

With a base case of 9.0 percent WACC and a 10.5x exit multiple (roughly $24.40 per share), the reader sees at a glance that defensible assumptions span $17.10 to $33.30. The width of that range is itself information: a tight grid signals a robust valuation, a wide one signals an answer hostage to its assumptions.

Scenario Analysis

Sensitivity tables move one or two dials mechanically. Scenario analysis builds whole alternate futures: a base case on the best available judgment, an upside case, and a downside case, each with an internally consistent set of assumptions. Internal consistency is the entire point. A recession does not just cut revenue growth; it compresses margins and can widen risk premiums, raising WACC. An upside case pairs faster growth with the margin expansion that operating leverage would actually deliver. A scenario that moves one variable and freezes the rest is not a scenario, just a mislabeled sensitivity.

Presenting the Output

On the football field, the DCF bar spans the downside-to-upside scenario range or an interior band of the sensitivity grid (commonly the 25th to 75th percentile), with the base case at the midpoint. The numbers travel with a narrative that names the drivers: an enterprise value range of $4.5 to 5.8 billion around a $5.1 billion base case, driven primarily by revenue growth of 7 to 12 percent and exit multiples of 9 to 11x, with roughly $300 million of sensitivity per 50 basis points of WACC. Presenting a single-point DCF without this is analytically irresponsible; the first question from any senior banker or board member is what happens when an assumption moves, and if the answer is not ready, the analysis is not done.

Sanity Checks Before You Present

• •Terminal value share. 60 to 80 percent of total enterprise value is normal. Above 85 percent, the forecast period is too short or the terminal assumptions too aggressive; below 50 percent, the reverse.
• •Cross-check the terminal assumptions. The implied exit multiple from a perpetuity-based terminal value should sit within 2 to 3 turns of the current trading multiple; the implied perpetuity growth rate from an exit multiple should stay below 4 to 5 percent.
• •Revenue reasonableness. Translate the final-year revenue into implied market share. Projecting $30 billion of revenue in a $50 billion addressable market means 60 percent share, rarely credible outside a monopoly.
• •Margin convergence. Terminal margins should converge toward industry norms; a business going from 15 to 45 percent EBITDA margins needs an extraordinary explanation.

The finished DCF then takes its place in the triangulation. A DCF above comps says the fundamentals see more value than the market prices in, or the projections are optimistic; a DCF below comps says the reverse; convergence with precedent transactions builds confidence in the range. Its unique contribution is the one it started with: an anchor that is independent of market pricing, which no other methodology on the page can provide.