Discounted Cash Flow Analysis

Introduction

You're valuing a company or checking an investment and need an anchor for intrinsic value, so start with a discounted cash flow to set a defensible baseline; the quick purpose: DCF estimates the present value of forecasted free cash flows (FCF = cash the business generates after capital reinvestment), and it gives you a single, comparable intrinsic number. One line: DCF turns future cash into a today number you can compare to price. Here's the quick math example: if forecasted FCF in 2025 is $120,000,000 and you use a 8% discount, the one-year present value is roughly $111,111,111 (120,000,000 / 1.08); what this hides is terminal assumptions and execution risk, so treat early-year cash and the terminal value separately - defintely stress-test both. Finance: build the 5-year FCF forecast and WACC input by Friday.

Key Takeaways

• DCF converts forecasted free cash flows into a single intrinsic value - use it as a defensible baseline.
• Accurate FCF inputs (revenue growth, margins, tax, capex, change in NWC) drive value - cash is king.
• Forecast a realistic explicit horizon (typically 5-10 years) with bottoms-up drivers and scenario cases.
• Discount with WACC (CAPM for cost of equity); small changes in the discount rate greatly affect value.
• Terminal value often dominates the model - stress-test terminal growth/exit multiples and reconcile with comps.

Cash-flow fundamentals

Define free cash flow and the clean formula

You're building a DCF and need a reliable measure of cash the business actually throws off to investors. Free cash flow (FCF) is that number: cash from operations after tax, less capital spending, plus or minus working-capital changes.

Use this canonical formula: FCF = NOPAT (net operating profit after tax) + Depreciation & Amortization - Capital Expenditures - Change in Net Working Capital.

Practical steps to compute each line:

• Pull operating profit (EBIT) from the income statement.
• Apply the normalized tax rate to get NOPAT (NOPAT = EBIT × (1 - tax rate)).
• Add back non-cash charges (D&A) from the cash-flow statement.
• Subtract reported capex (cash outflow for PP&E) from investing activities.
• Compute change in net working capital (ΔNWC) as Δ(AR + Inventory - AP) on the balance sheet.

Example - Company Name FY2025 quick math: Revenue = $200,000,000, EBIT margin = 12% → EBIT = $24,000,000. Use tax rate = 21% → NOPAT = $18,960,000. Add D&A = $8,000,000, subtract CapEx = $10,000,000, subtract ΔNWC = $2,000,000 → FCF = $14,960,000.

What this estimate hides: one-offs, timing effects, and working-capital seasonality. Normalize where possible - use trailing twelve months (TTM) or multi-year averages for volatile items, and call out unusual receipts or payables so you don't double-count a one-time cash swing.

Show the inputs you must forecast and how to structure them

You need a driver map that ties top-line drivers to cash. Break forecasts into explicit, traceable inputs: revenue growth, margins, tax rate, capex, and change in net working capital.

• Forecast revenue from units × price or market share × market size; avoid blunt % lifts with no driver.
• Build margins from gross margin, operating expense lines, and expected operating leverage; split fixed vs variable costs.
• Choose a tax approach: use the company's recent effective tax rate if stable; otherwise use statutory federal + state - typically the U.S. blended rate lands near 25%, but rely on the company history for precision.
• Project capex by separating maintenance capex (to sustain current revenue) and growth capex (to support new capacity); express as $ or as % of revenue.
• Model ΔNWC as a percent of revenue or via working-capital days (AR, Inventory, AP); for volatility, use rolling averages or model each working-capital line explicitly.

Best practices and checks:

• Source TTM cash-flow and balance-sheet items; reconcile to reported net cash change.
• Use multi-year averages for D&A and capex when the business cycles interact with investment timing.
• Stress-test margin assumptions against product mix and commodity price moves.
• Flag items to adjust: pension cash contributions, restructuring, asset sale proceeds - pull them out of core FCF if they're non-recurring.

Here's the quick math method: translate each percent into dollars early so your sensitivity tables can flip real cash, not abstract percentages.

One line and the practical implications - Cash is king

Cash is king - accurate FCF drives value.

Actions to make FCF credible and actionable:

• Document each input and its source (line item, footnote, management guide).
• Run three scenarios (base, bear, bull) and build a two-way sensitivity on discount rate and terminal growth.
• Cross-check implied metrics: FCF yield, implied EV/EBITDA from the terminal value, and IRR vs. comparable deals.

Quick sanity checks you should run every time: implied exit multiple within sector precedent, FCF margin vs peers, and cumulative cash vs reported cash balances. If numbers diverge, reconcile the line items - you're likely misclassifying a non-recurring cash flow or missing a smoothing adjustment.

If you want a simple next step right now: pull TTM cash flow and balance-sheet items for FY2025, compute the FCF line-by-line, and test a ±200 bps swing in margins to see sensitivity - then iterate from there. Finance should own that first pass; it's how you turn assumptions into a number you can trust (and defintely defend).

Forecasting the explicit period

Choose the horizon

You're setting the forecast horizon to anchor present value - pick the shortest window that captures the company moving toward steady state, and no more. Start with the direct takeaway: match horizon to visibility and industry dynamics so your terminal assumptions don't hide bad forecasts.

Steps to decide the horizon

• Assess visibility: contracts, backlog, regulatory milestones, product launch timelines.
• Map lifecycle: early-stage growth → longer (eight to ten years); mature/volatile firms → shorter (five years).
• Check capital timing: major capex or R&D that finishes inside X years argues for including that period explicitly.
• Align with reporting cadence: use annual for 5-10 year views, quarterly if seasonality or rapid change matters.

Best practices

• Default to a five to ten year horizon; tilt shorter for cyclical, longer for capital-heavy, stable sectors.
• Document the single event that ends the explicit period (patent expiry, capacity build-out, major contract end).
• Run a sensitivity that shifts the horizon by two years - if value swings materially, your terminal assumptions matter too much.

One clean line: Pick a horizon that reflects what you can reasonably forecast, not what you wish you knew.

Build bottoms-up drivers

You build credibility with drivers that tie to observable economics: units, price, share, and margin levers. Start at the smallest measurable element and roll up to revenue, then to operating line items and free cash flow (FCF).

Concrete steps

• Project units and pricing: revenue = units × price × share; forecast unit growth separately from pricing/realization.
• Model market share: use addressable market (TAM), served available market (SAM), and expected share gains/losses.
• Drive margins from the cost structure: gross margin drivers (input costs, productivity), operating expense lines (SG&A, R&D) tied to revenue or headcount.
• Estimate capex and depreciation: split into maintenance vs growth; express as % of revenue or as per-project schedules.
• Model working capital in days: days sales outstanding (DSO), days inventory (DIO), days payable (DPO) → convert to ΔNWC as % of revenue.
• Use tax assumptions tied to jurisdictional rates and expected shielding (loss carryforwards, credits).

Here's the quick math for FCF (illustrative example): start with EBIT, subtract tax to get NOPAT, add D&A, then subtract CapEx and ΔNWC. Example: revenue $100m, EBIT margin 15% → EBIT $15m; tax 21% → NOPAT $11.85m; D&A $5m, CapEx $6m, ΔNWC $1m → FCF ≈ $9.85m.

Best practices and checks

• Link drivers to KPIs (AR days, churn, conversion) so operational teams can own inputs.
• Keep growth and margin paths internally consistent: aggressive top-line with static margins is unlikely.
• Calibrate with historical rates: if revenue CAGR in past five years was 8%, a 30% forecast needs explicit rationale.
• Build monthly/quarterly submodels for material seasonality (>20% swing).

What this estimate hides: driver-based models still rely on judgment - document assumptions, sensitivity ranges, and the operational triggers that would move you between scenarios. Be defintely conservative on uptake curves for new products.

One clean line: Drive revenue and margins from observable units, prices, and cost drivers - cash flows follow the ops.

Craft scenarios and keep forecasts useful

Forecasts are tests, not prophecies. Build three clear scenarios - base, bear, bull - and make each a coherent story tied to measurable triggers (price pressure, customer wins, supply shocks).

How to build scenarios

• Define the pivot variables: top-line CAGR, margin progression, capex intensity, and working capital elasticity.
• Assign plausible ranges: base = management guidance or median consensus; bear = base minus 3-7 p.p. top-line and 2-5 p.p. margin compression; bull = base plus upside tied to expansion or efficiency.
• Attach probabilities and decision triggers (regulatory approval date, market share milestones).
• Produce sensitivity tables for discount rate (WACC) and terminal growth to show value drivers.

Cross-checks to keep forecasts grounded

• Compare implied exit multiples and cumulative FCF to market comps and enterprise value.
• Run payback and IRR checks: if implied payback exceeds industry norms, question the growth or margin path.
• Validate top-down: ensure forecasted CAGR fits within realistic TAM and competitor dynamics.

One clean line: Forecasts should be realistic enough to test scenarios, not crystal balls.

Action: ask Finance to build a seven-year bottoms-up model with base/bear/bull inputs and a sensitivity table for WACC and terminal growth by next Friday - owner: Finance.

Discount rate and risk

You're picking a discount rate to turn forecasted free cash flow into a present value so you can compare it to price; pick WACC carefully, run sensitives, and cross-check implied multiples.

Use WACC to discount Free Cash Flow

WACC (weighted average cost of capital) blends the cost of equity and the after-tax cost of debt in proportion to market value of equity and debt; use WACC to discount unlevered free cash flows to enterprise value.

Practical steps:

• Estimate market value of equity (market cap) and net debt (gross debt - cash).
• Compute weights: Equity weight = Market equity / (Market equity + Net debt).
• Use after-tax cost of debt = stated debt yield × (1 - tax rate).
• Combine: WACC = We × Re + Wd × Rd × (1 - Tc).

Best practices and checks:

• Use market values, not book values.
• Recalculate quarterly if market cap swings >20%.
• Use effective tax rate from latest FY2025 filings; if uncertain, prefer marginal statutory rate for forward-looking debt shield.

One-liner: The discount rate prices risk - small changes change value a lot.

Estimate cost of equity with CAPM and adjust beta for leverage

Use CAPM: Cost of equity Re = Risk-free rate + Beta × Equity risk premium (ERP). Beta captures sensitivity to market returns; adjust (re-lever) beta for the company's target capital structure (Hamada formula).

Steps you can follow today:

• Pick a risk-free rate anchored to a long-term sovereign yield (10-year Treasury) near FY2025.
• Choose an ERP-use published consensus (e.g., 4.5-6.0%) and be explicit which source you used.
• Start with an industry raw beta (unlevered) from comparable firms; unlever using each peer's debt-to-equity and then re-lever to the Company Name target D/E: Levered beta = Unlevered beta × (1 + (1 - Tax rate) × D/E).
• Compute Re = Rf + Levered beta × ERP.

Example (illustrative 2025 inputs): assume risk-free = 4.0%, ERP = 5.5%, unlevered beta = 0.90, target net debt / equity = 0.20, tax = 21%. Re ≈ 9.7%. What this estimate hides: sensitivity to ERP and beta choice.

One-liner: Adjust beta for leverage so the cost of equity reflects the capital structure risk you're pricing.

Practical considerations: calibrate, stress, and document assumptions

Small changes in WACC change value nonlinearly. Run explicit sensitivity tables across WACC ±100-300bps and terminal growth ±50-100bps, and translate into implied multiples to sanity-check.

Concrete validation steps:

• Compute implied EV/EBITDA and P/E from your DCF outputs and compare to sector comps (use FY2025 comps data).
• Run a perpetuity sanity check: PV = CF1 / (WACC - g). Example: CF1 = $100, WACC = 8.9%, g = 2.0% → PV ≈ $1,449. If WACC rises to 9.9%, PV ≈ $1,266 (-12.6%).
• Keep a change log: each change to Rf, ERP, beta source, or tax rate needs a short rationale tied to FY2025 facts.

Best practices:

• Favor market-derived inputs where possible; if you use judgment, state it explicitly.
• Use three WACC cases (base, conservative, aggressive) and map them to realistic financing scenarios.
• Check model outputs against transaction and trading multiples for FY2025 peers; if implied multiples look far off, revisit forecasts or discount inputs - don't force-fit.

One-liner: Reconcile WACC-driven values with market evidence - implied multiples catch mistakes fast and defintely save time.

Finance: update the WACC inputs, produce a 3-case WACC table, and deliver sensitivity outputs by next Friday.

Terminal value and method choice

Gordon growth (perpetuity) versus exit multiple (comps)

You're picking how to convert year‑N cash flow into a single continuing value; pick the method that matches how stable and predictable cash flows will be.

Practical steps:

• Use Gordon growth when cash flows are stable, low-growth, and capital intensity is steady.
• Use an exit multiple when the business will revert to a market multiple (industry norms) or when comparables exist.
• Combine both: run a base using Gordon and a cross-check using exit multiples.

Quick math example (how each works):

Assume final-year free cash flow $120 million, terminal growth 2.0%, WACC 8.0%. Gordon TV = 120 × (1+0.02) / (0.08-0.02) = $2,040 million. If year‑N EBITDA = $180 million and sector exit = 10×, exit-multiple TV = $1,800 million.

What this estimate hides: Gordon locks in a perpetual growth assumption; exit multiples hide reversion timing and accounting differences.

One line: Use Gordon for predictability, exit multiples when market comparables drive value-run both.

Test terminal growth and exit multiples against sector precedent

Don't pick terminal growth or a multiple in a vacuum-benchmark to macro and sector evidence and document sources.

Steps to validate assumptions:

• Set terminal growth below long‑run GDP; market practice often uses 1.5-3.0% depending on inflation expectations.
• Build a comparables table (public peers + recent M&A) for EV/EBITDA, EV/FCF, and revenue multiples using latest fiscal‑year 2025 figures where available.
• Adjust comparables for accounting (capital leases, operating leases, minority stakes), and for differences in growth and margin profiles.
• Triangulate: if Gordon implies a terminal multiple outside the peer 25-75th percentile, revisit growth/WACC or extend the explicit forecast.

Practical sourcing: pull 2025 LTM (last twelve months) metrics from filings and use transaction comps from 2023-2025; record medians and 10th/90th percentiles for stress tests.

One line: Test growth and multiples against 2025 sector precedent and pick assumptions inside a defensible peer range.

Why terminal value often dominates and how to manage that dominance

Terminal value frequently represents the majority of enterprise value, so control it explicitly and test how much it drives your output.

Concrete checks and mitigations:

• Calculate TV share: present-value explicit FCF sum versus PV of terminal value. If terminal > 50-70% of EV, flag model risk.
• Run sensitivity tables across WACC (±0.5-1.0ppt) and terminal growth (e.g., 0.5-3.0%) and show implied exit multiples.
• If terminal share is too high, extend the explicit forecast from 5 to 8-10 years, lower terminal growth, or use a staged (two‑step) terminal that phases growth down.
• Cross-check implied multiples: implied EV/EBITDA should sit within peer ranges; if it's outside, either forecasts or discount rates are the issue.

Quick math to illustrate dominance: using the earlier Gordon TV $2,040 million and an explicit PV of FCF sum of $300 million, terminal share = 2,040 / (2,040+300) = 87%. That's defintely a red flag to test assumptions.

One line: Terminal value usually dominates-stress test it, extend explicit years, and validate implied multiples before you rely on the model.

Model execution and validation

Build a three-case model and run WACC / terminal-growth sensitivities

You're finalizing a valuation and need a defensible range, so start with a compact three-case model: base, bear, bull. The direct takeaway: build clear, traceable assumption sets and run a sensitivity grid to see which inputs move value the most.

Steps to build the three cases

• Set a fiscal anchor: use FY2025 actuals as Year 0. For this example use starting free cash flow (FCF) $12.0m in FY2025.
• Define scenarios: base = conservative growth (example: FCF +10%/yr), bear = downside (FCF +5%/yr), bull = upside (FCF +15%/yr).
• Choose explicit horizon: 5-10 years. This example uses 5 years (FY2026-FY2030) for visibility and faster sensitivity work.
• Project line items bottom-up: revenue drivers → margins → capex → change in net working capital → FCF. Keep the same logic across scenarios; only change driver assumptions.
• Pick discount rates: base WACC 9%, bear WACC 11%, bull WACC 7%. Document how you derived beta, credit spread, and tax rate.

Quick math example (5‑year explicit + Gordon terminal): projected FCFs (base) = $13.2m, $14.52m, $15.97m, $17.57m, $19.33m. With WACC 9% and terminal growth 2%, PV of explicit FCF ≈ $61.7m, PV of terminal ≈ $183.0m, implied enterprise value ≈ $244.7m.

What this example hides: sensitivity to terminal assumptions and capital structure effects on equity value. Run the same calculation for bear/bull to get a credible range; show all assumptions in a single tab so reviewers can trace changes.

Cross-check with implied multiples, comps, payback and IRR

You're done with arithmetic; now sanity-check the outputs. One clean rule: if implied multiples or returns look outlandish, something in the forecast or discounting is wrong.

Specific cross-checks and how to run them

• Compute implied exit multiple: terminal value / last-year EBIT or EBITDA. Example: terminal value $281.7m divided by assumed Year 5 EBITDA $30.0m → implied exit multiple ≈ 9.4x.
• Compare EV/EBITDA and P/E to sector comps: pull median and 25-75th percentile multiples for your industry; if your implied 9.4x sits outside that band, re-examine margins, growth, or WACC.
• Calculate enterprise value vs. market value: EV from DCF vs current market EV shows implied upside/downside as a percent; show both absolute and percent gaps.
• Run payback and IRR checks: compute discounted payback (cumulative PV of cash flows) and an IRR to the implied entry price. If discounted payback > investment horizon or IRR < expected hurdle, the thesis is weak.

Concrete thresholds to watch (example): implied EV/EBITDA substantially above sector 75th percentile or implied IRR below 7-8% (for lower-risk investments) signals over-optimism or under-discounting. If numbers feel off, adjust inputs and re-run - don't fudge to hit a target multiple.

Run sensitivity tables and interpret outliers

Run a two-way sensitivity table for WACC vs terminal growth and present the EV results; this shows where value concentration lives. One line: If implied multiples look crazy, either forecast or discount assumptions are wrong.

Steps to set the table

• Choose WACC range: example columns 7%, 9%, 11%.
• Choose terminal-growth range: example rows 1%, 2%, 3%.
• Populate table cells with enterprise value for each WACC/g pair using the same explicit FCF stream. Highlight cells that change valuation by >±30% vs base.

Example sensitivity grid (EV, millions):

How to read this: value swings > 40-50% across realistic WACC/g pairs mean the model is driven by terminal assumptions; tighten your forecast horizon, add scenario detail, or justify the WACC/g choices with market data.

Also run sensitivity on individual drivers (revenue growth, margin, capex) to show which operational levers move value most. If one small change flips your verdict, call that out clearly - it's where the model is fragile or where you should do deeper due diligence.

Action: Finance - build the base DCF and a full WACC/terminal-growth sensitivity table using FY2025 actuals, then deliver the model and a one-page sanity-check (implied multiples vs comps) by next Friday.

Discounted Cash Flow Analysis - Actionables

Action

You're preparing a valuation and need a solid, repeatable DCF you can stress-test against market prices.

Steps to build the model:

• Start with actual FY2025 results as the base year - revenue, EBITDA, capex, working capital, tax paid.
• Pick an explicit forecast horizon: use 5 years for low-visibility businesses, 7-10 years for steady industrials and utilities.
• Project revenue drivers (units, price, share), then margins to derive operating profit and free cash flow (FCF) = operating cash after tax - capex ± change in net working capital.
• Create three scenarios: base (realistic), bear (stress: slower growth, margin compression), bull (faster growth, margin recovery). Use concrete yearly rates - e.g., base revenue growth tapering from 5% to 3%, bear 0% to 2%, bull 8% to 5%.
• Discount explicit FCFs with WACC (see next section) and compute terminal value either by Gordon growth or exit multiple.

Here's the quick math for the terminal value example: if terminal-year FCF = $20.0m, terminal growth g = 2.0%, and WACC = 9.0%, TV = 20.0×(1.02)/(0.09-0.02) = $291.4m. What this estimate hides: terminal assumptions drive most value - so vary g and WACC.

One-liner: Build a 5-10 year DCF with three cases and you'll have a defensible intrinsic anchor, not a crystal ball.

Quick check

You've got a DCF output - now reconcile it to market reality before you act.

Practical reconciliation steps:

• Compute implied enterprise value (EV) = PV of explicit FCF + PV of terminal value + cash - debt.
• Derive implied multiples: EV/EBITDA, EV/Revenue, P/E (if you convert to equity value). Use the company's LTM FY2025 EBITDA and revenue as denominators.
• Compare to sector comps (median and 25th/75th percentiles). If implied EV/EBITDA differs by > 25% from the comp median, revisit forecasts or WACC.
• Run sanity checks: implied IRR vs required return; payback period; implied long-term FCF margins vs historical peaks.
• Document the single biggest driver of discrepancy (forecast growth, margin, or discount rate) and adjust only after written rationale.

Example check: implied EV = $400m, FY2025 LTM EBITDA = $30m → implied EV/EBITDA = 13.3x. If sector median is ~10x, that's a ~33% premium - dig into margin or multiple assumptions.

One-liner: If implied multiples look crazy, either your forecasts or your discount assumptions are wrong - don't ignore the gap.

Owner

You need a named owner, clear deliverables, and a tight deadline so the DCF drives decisions.

Finance deliverables and checklist (owner: Finance lead - model owner name):

• Deliver an Excel DCF for FY2026-FY2032 (or FY2026-FY2030 if 5 years) using FY2025 actuals as the base.
• Produce three scenarios (bear, base, bull) with explicit yearly FCFs and a terminal value by both Gordon growth and exit multiple.
• Include a sensitivity table: WACC from 6.0% to 12.0% (0.5% steps) vs terminal growth from -0.5% to 3.0% (0.5% steps) - an 11×8 grid is fine.
• Provide reconciliations: implied EV/EBITDA, EV/Revenue, implied IRR, and a short note listing the top three model risks and mitigants.
• Source inputs: FY2025 audited financials, debt schedule, capex plan, working capital schedule, and current market inputs for risk-free rate, equity premium, and beta sources (cite providers).
• Deliverables due: base DCF + sensitivity workbook by next Friday; a one-page commentary with action recommendation (buy/hold/sell or strategic options) on the same day.

One-liner: Finance: build the base DCF and sensitivity table by next Friday - then we'll use implied multiples to choose the action.