Valuation & DCF

Because a share price means nothing without the share count. A 10 dollar stock with a billion shares outstanding is far bigger than a 500 dollar stock with a million. You need market capitalization, and usually enterprise value, to size the company. Then you have to scale that value against something the business actually produces, like EBITDA or earnings. That ratio is the multiple, and multiples are the thing that is genuinely comparable.

• Share price is meaningless without shares outstanding
• Market capitalization sizes the equity, enterprise value sizes the whole business
• Value must then be scaled by an operating metric to compare across companies
• That scaled ratio is the multiple
• A stock split changes the price and changes nothing about the value

Whether you can explain a stock split as the clean proof. Nothing about the business changes and the share price halves.

EV/EBITDA ignores depreciation, so it effectively treats a company's assets as if they were free. EV/EBIT charges depreciation, which is a rough proxy for the capital the business has to keep spending just to stand still. So I would reach for EV/EBIT when comparing businesses with very different asset intensity, an asset heavy manufacturer against an asset light services firm. EBITDA stays more common because it is less distorted by differing depreciation policies.

• EBITDA adds back depreciation and so ignores capital intensity
• EBIT charges depreciation as a crude proxy for maintenance capex
• Use EBIT when asset intensity differs materially across the peer set
• EBITDA is more comparable when depreciation policies differ
• Both are pre-interest, so both pair with enterprise value not equity value

Whether you tie the choice to capital intensity rather than reciting that one adds back depreciation.

Trailing uses the last twelve months of actual results, forward uses next year's or the year after's estimates. I would lead with forward, because share prices reflect what investors expect rather than what already happened, and forward multiples are usually lower for a growing company. The catch is that forward multiples depend on consensus estimates, so I need every company in the set on the same basis.

• Trailing uses actual last twelve months results
• Forward uses consensus estimates for next year or beyond
• Forward is generally preferred: price reflects expectations
• Forward multiples are lower for a growing company
• Estimates must come from a consistent source across the whole set

Whether you flag the consistency requirement. Mixing sources across a comps set is the mistake that gets caught in review.

A business is worth the cash it will hand its owners over its life, counted in today's money. So I forecast the cash it throws off each year after paying for what it needs to keep running, then shrink each year's number, because a dollar promised later is worth less than a dollar now, and more so the riskier the business. Add those up, add a lump for the years beyond the forecast, and that is the value of the whole company. Take off its debt and you have what the shares are worth.

• Value is future cash, not profit, counted in today's money
• Forecast cash after the spending needed to keep the business running
• Later and riskier cash is shrunk more, which is all a discount rate does
• Most of the value sits in the years beyond the forecast
• Subtract debt to get from the whole company to the shares

Whether you can drop the vocabulary without dropping the logic. Interviewers use this to see if you understand discounting or only the formula.

Comparable public companies, precedent transactions, and the discounted cash flow. Precedent transactions usually gives the highest value, because acquirers pay a control premium and expect synergies. Comps generally come in lower since they reflect minority stakes trading at market prices. The DCF varies most, because it depends entirely on your assumptions.

• Name all three cleanly
• Precedents highest: control premium plus synergies
• Comps reflect minority stakes at market prices
• DCF is assumption-driven and can land anywhere
• An LBO analysis usually sets the floor

Whether you can say why precedents run high, not just that they do.

You value each asset at what it would fetch in a sale, apply a recovery haircut by asset class, then subtract liabilities. Receivables might recover most of face value, inventory far less, goodwill nothing. It matters in distress, where it sets a floor: if break-up value exceeds going concern value, the business is worth more dead than alive, and creditors will say so.

• Value assets at realizable sale proceeds, then subtract liabilities
• Recovery haircuts differ sharply by asset class
• Intangibles and goodwill usually recover close to nothing
• Used in distress and in recovery analysis for creditors
• Acts as a floor: it is what a going concern must beat

Whether you frame it as a floor rather than as another opinion of value.

It is the extra return investors demand for holding equities rather than risk free government bonds. It is not observable, so it is estimated two ways: historically, by averaging the excess return of an equity index over bonds across a long period, or implied, by solving for the return that reconciles current index prices with expected cash flows. Both land broadly in the five to six percent region.

• Compensation for bearing equity risk over the risk free rate
• Not observable, so it must be estimated
• Historical approach: long run average excess return over bonds
• Implied approach: solve from current index prices and expected cash flows
• Use a published, consistent source rather than inventing a number

Whether you know it is an estimate with a method behind it, not a constant you memorized.

It means estimating what the business is worth, either to all capital providers as enterprise value or to shareholders as equity value. You do it either intrinsically, by projecting and discounting the cash the business will produce, or relatively, by looking at what similar companies or similar transactions are worth. Every method is an attempt to answer the same question with different evidence.

• Estimating worth, either enterprise value or equity value
• Intrinsic approach: project and discount future cash flows
• Relative approach: compare to similar companies or transactions
• A valuation is a range and a judgment, not a single correct number
• Different methods answer the same question with different evidence

Whether you frame valuation as a range and a judgment. Candidates who describe it as a calculation usually struggle with the follow-ups.

It is the summary page: one horizontal bar per methodology, all plotted on a single axis, usually implied share price. You would show DCF, comps, precedents, and often a fifty two week trading range and an LBO floor. Each bar is labeled with what drives it, such as eight to ten times EBITDA. The current share price and any offer price are marked with a vertical line.

• One bar per methodology on a common axis, usually share price
• Typically DCF, comps, precedents, trading range, sometimes an LBO value
• Each bar labeled with the assumption range that produced it
• Current share price and any offer marked as a reference line
• Bar widths come from real sensitivities, not from arbitrary padding

Whether you know each bar has to be traceable to a stated assumption range.

Unlevered free cash flow is the cash available to all capital providers, before any debt service. Discounting it at WACC, which is the blended cost of all capital, produces enterprise value directly. It also makes the output independent of capital structure, so you can compare companies with different leverage. If you used levered free cash flow you would discount at cost of equity and get equity value instead.

• Unlevered is pre-interest, available to all providers
• WACC is the blended cost of all capital, so they match
• Output is enterprise value
• Capital-structure neutral, so comparisons work
• Levered plus cost of equity gives equity value, which is also valid

Whether you understand the numerator and the denominator have to describe the same capital providers.

Companies have different fiscal year ends, so a raw comps set compares periods that do not line up. Calendarization restates everyone onto a common period, normally the calendar year, by stitching quarters together. For a March year end company, calendar 2024 is April to December from the fiscal year ending March 2025, plus January to March from the year ending March 2024. Skip it and you are comparing different economic conditions, which is severe in a cyclical sector.

• Different fiscal year ends make raw periods non-comparable
• Restate each company onto a common calendar period by adding quarters
• Requires quarterly data, which sometimes is not disclosed
• Matters most in cyclical sectors where conditions shift within a year
• Apply the same convention to the target and every peer

Whether you can actually construct the stitched period out loud rather than just define the word.

Because the lines have different drivers. Receivables track revenue, but inventory and payables track cost of goods sold, so a single percentage of revenue quietly mixes them. Days are also an operating assumption a client can challenge: saying receivable days go from sixty to fifty five is a commercial claim about collections. Percentage of revenue is a plug that nobody can argue with, which is exactly the problem.

• Receivables driven by revenue, inventory and payables by COGS
• Days make each assumption explicit and testable
• Lets you model a genuine change in terms or collections
• A blanket percentage hides which line is moving
• Hold days flat in the terminal year so investment grows only with the business

Whether you can defend an assumption commercially rather than mechanically.

Not necessarily. A high terminal share is normal, most DCFs land somewhere between 60 and 80% of value in the terminal year, simply because a five year projection captures only a fraction of a going concern. At 85% I would check whether the projection period is too short to reach a steady state, whether the perpetuity growth rate sits too close to the discount rate, and whether the exit multiple implies more growth than my forecast supports.

• A high terminal share is expected, not automatically an error
• Typical range is roughly 60 to 80% for a five year projection
• A short forecast horizon pushes more value into the terminal year
• Perpetuity growth too near the discount rate inflates terminal value sharply
• Cross-check the exit multiple against the implied perpetuity growth rate

Whether you extend the projection horizon before touching assumptions. Reaching straight for a lower growth rate to hit a target answer is the wrong instinct.

The denominator goes negative and the formula returns a negative terminal value, which is nonsense. Economically the model is saying the business grows faster than the return investors require, forever, so it eventually becomes larger than the whole economy. The fix is the input, not the formula: growth in perpetuity should sit at or below long run nominal GDP, so roughly one to three percent.

• WACC minus growth turns negative, so terminal value flips negative
• As growth approaches WACC, terminal value explodes toward infinity
• No company can outgrow the economy forever
• Cap perpetuity growth at long run nominal GDP, roughly 1% to 3%
• A negative or absurd output means an input is wrong, not the method

Whether you can explain the economic absurdity, not just the arithmetic breakdown.

Three effects, in order. The risk-free rate feeds straight into the cost of equity and the cost of debt, so WACC rises and every discounted cash flow is worth less, with the terminal value hit hardest. Multiples compress for the same reason, because a multiple is a discount rate turned upside down, and long-duration growth stocks fall most. Deals slow, because debt costs more, so sponsors can pay less, and the gap between what buyers offer and what sellers remember opens up.

• The risk-free rate lifts both the cost of equity and the cost of debt, so WACC rises
• Present values fall, and the terminal value falls most
• Multiples compress, growth stocks most, because more of their value is far out
• Debt is dearer, so leveraged buyers can pay less
• Bid-ask spreads widen and M&A volume falls until sellers reset expectations

Whether you connect the three rather than answering only the DCF part. The transmission from rates to deal volume is what a banking interviewer wants.

Three routes, best first. Look at the pricing on its most recent borrowing, a credit facility or a term loan, since that is a real marginal rate. Failing that, use its credit rating and take the current yield on comparably rated bonds. If it is unrated, build a synthetic rating from interest coverage, then add that spread to the risk free rate. I would then tax effect it.

• Start with the rate on recent borrowings, the true marginal cost
• Otherwise use yields on comparably rated debt
• Unrated: synthetic rating from interest coverage, then add the spread
• It must be forward looking and marginal, not historical average interest
• Tax effect it, since interest is deductible

Whether you reach for a forward looking marginal rate rather than dividing last year's interest expense by debt.

Perpetuity growth assumes the business grows forever at a modest rate, so terminal value is the final year cash flow times one plus g, divided by WACC minus g. Exit multiple assumes you sell the business at a multiple of a terminal year metric, usually EBITDA, taken from where comparable companies trade. Perpetuity growth is more theoretically grounded, exit multiple is more market-grounded, and in practice you cross-check one against the other.

• Perpetuity: final year cash flow times (1+g) over (WACC minus g)
• Exit multiple: terminal metric times a comparable trading multiple
• Growth rate should not exceed long-run GDP, typically 2 to 3%
• Exit multiple imports current market sentiment into a long-run number
• Standard practice is to run both and check they imply each other

Whether you know to sanity-check them against each other, and that a growth rate above GDP implies the company eventually becomes the whole economy.

It values a company using multiples paid in past acquisitions of similar businesses, rather than where similar companies currently trade. The key difference is that transaction multiples include a control premium and expected synergies, so they run higher than trading comps. The weakness is that they reflect market conditions at the time of each deal, which may be nothing like today.

• Multiples paid in completed M&A deals for similar targets
• Includes a control premium and buyer synergies
• Generally produces the highest valuation of the three methods
• Data is historical, so market conditions may have shifted
• Deal-specific factors like competitive auctions distort individual multiples

Whether you explain why it runs higher rather than only stating that it does.

Start with EBIT. Multiply by one minus the tax rate to get net operating profit after tax. Add back depreciation and amortization and any other non-cash charges. Subtract capital expenditure. Subtract the increase in net working capital. That is unlevered free cash flow. Nothing related to financing appears anywhere: no interest, no debt repayment, no dividends, because this cash belongs to all capital providers.

• EBIT times one minus tax equals NOPAT
• Add back D&A and other non-cash items
• Subtract capex
• Subtract the increase in net working capital
• No financing items at all, since the cash flow is unlevered

Whether you can produce the bridge without hesitating, and whether you keep every financing item out of it.

Usually that it is heavily levered. Interest expense compresses net income, so the equity multiple looks expensive, while enterprise value spreads the same business across debt and equity and looks cheap. Heavy depreciation does the same thing, since EBITDA sits above it and net income below. So before calling it mispriced I would check the capital structure, the D&A load and the effective tax rate.

• The gap is driven by leverage, D&A and tax, the items between the two
• Interest compresses net income, inflating P/E
• EV/EBITDA is capital structure neutral, so it looks cheaper
• Heavy depreciation widens the same gap
• Check leverage and D&A before concluding it is mispriced

Whether you can name what actually sits between EBITDA and net income, and reason from that rather than guessing.

Earnings before interest, taxes, depreciation and amortization. It is used as a rough proxy for operating cash flow that is comparable across capital structures and tax regimes. The criticisms are real: it ignores capital expenditure, so it flatters asset-heavy businesses that must constantly reinvest; it ignores working capital swings; and because it is not a defined accounting measure, companies adjust it aggressively.

• Earnings before interest, taxes, depreciation and amortization
• A proxy for operating cash flow, comparable across structures
• Ignores capex, which is a real and recurring cost
• Ignores working capital movements
• Not GAAP-defined, so "adjusted EBITDA" is often generous

Whether you volunteer the capex criticism. It is the substantive one and the reason EBITDA overstates cash generation for industrials.

EV/EBITDA is capital-structure neutral, so it lets you compare companies with different leverage. It also sits above interest, taxes, depreciation and amortization, which strips out financing decisions, tax jurisdictions and differing depreciation policies. P/E is distorted by all of those, and breaks entirely when earnings are negative. EV/EBITDA is the default for comparing operating businesses.

• Capital-structure neutral, so leverage differences do not distort it
• Removes tax jurisdiction and depreciation policy differences
• Still meaningful when net income is negative
• P/E is a levered, post-tax, equity-holder metric
• P/E is still preferred for banks, where leverage is the business

Whether you can name a case where P/E is the better choice. Banks and insurers are the standard answer.

It is the fallback when EBITDA is negative or meaningless: early stage businesses, companies investing heavily ahead of profitability, or sectors where accounting policy distorts profit. The hidden assumption is that this company will eventually earn margins like its peers. Revenue is the only line that ignores how profitably it is earned, so I would pair it with growth and margin, and show the implied out-year EBITDA multiple.

• Used when EBITDA is negative or not meaningful
• Also when accounting differences distort profit comparability
• The implicit assumption is margin convergence with peers
• Always pair it with growth and margin, since revenue ignores both
• Cross-check by showing the multiple on a normalized future EBITDA

Whether you name the margin convergence assumption. Anyone can say EBITDA is negative; the assumption is the real answer.