Examining the Limitations of EV/EBITDA

Introduction

You're screening deals or comps and need a fast filter - the direct takeaway: EV/EBITDA is a useful quick screen but often misleads for capital‑heavy industries (utilities, industrials), financial firms, or high‑growth companies with negative or volatile EBITDA. EV (enterprise value) = market cap + net debt; EBITDA = earnings before interest, taxes, depreciation, amortization (a proxy for operating cash flow). Practitioners use it commonly for M&A screening and cross‑company multiples because it strips capital structure, but it ignores capex, working capital needs, and accounting differences - so start here, not stop here. defintely use it as a filter, then dig deeper.

Key Takeaways

• EV/EBITDA is a useful, fast screening multiple-but treat it as a starting filter, not a final verdict.
• It omits depreciation, capex and working‑capital swings, so it can understate true cash needs in capital‑heavy firms.
• Avoid EV/EBITDA for financials, REITs, or high‑growth/negative‑EBITDA companies; use book/NAV, EV/Revenue or EV/ARR where appropriate.
• Normalize for cyclicality and one‑offs (3-5 year EBITDA average) and flag firms with high leverage (e.g., net debt/EBITDA > ~4x).
• Complement EV/EBITDA with cash‑focused metrics (EV/FCF, EV/EBIT, EBITDA‑capex) and a check of capex and net‑debt ratios before valuing.

Accounting and earnings distortions

You're using EV/EBITDA to screen names - that's fine as a first pass, but be careful: EBITDA removes depreciation and amortization (non-cash charges), so it can hide real capital consumption and working-capital needs that change cash flow materially. Start here, not stop here.

EBITDA strips depreciation and amortization, so it hides real capital consumption

Takeaway: EBITDA removes D&A (depreciation and amortization), which masks the cost of the assets that generate revenue; for capital-intensive firms this understates how much cash is actually needed to sustain the business.

Steps to test and adjust

• Pull FY2025 depreciation and amortization from the income statement or notes.
• Pull FY2025 capex (capital expenditures) from the cash flow statement and split maintenance vs growth capex if possible.
• Estimate maintenance capex using one of three methods: historic maintenance-to-revenue ratio, capex-to-depreciation parity, or management guidance.
• Compute adjusted EBITDA-like cash metric: adjusted EBITDA minus maintenance capex.

Practical rules

• Flag if maintenance capex roughly equals or exceeds reported D&A - that implies D&A isn't capturing full replacement needs.
• Prefer EV/EBIT (operating profit) or EV/FCF for capex-heavy industries.

One-liner: If assets wear out, EBITDA lies about cash available for investors.

Ignores working capital swings that can change cash needs materially

Takeaway: EBITDA ignores changes in working capital (receivables, inventory, payables) that can turn strong EBITDA into poor cash flow in a hurry.

Steps to evaluate working-capital risk

• Extract FY2025 year-over-year changes in accounts receivable, inventory, and accounts payable from the cash flow statement.
• Convert those changes into days (DSO, DIO, DPO) and compare to industry peers and multi-year medians.
• Run a sensitivity: assume receivables slow by 10 days - show incremental cash need and impact on free cash flow for the next 12 months.

Best practices

• Treat working-capital outflows as a recurring cash drag if swings repeat 2+ times in 3 years.
• When valuing, use EV/EBITDA only after adjusting EBITDA for normalized working-capital cash flow or use EV/FCF instead.

One-liner: Strong EBITDA ≠ strong cash if customers or inventory tie up cash.

Non-GAAP adjustments and recurring one-offs can inflate EBITDA - quick check

Takeaway: Companies often report adjusted (non-GAAP) EBITDA that removes items they call non-recurring; many are recurring in practice and inflate the multiple's denominator.

How to spot and correct for cosmetic EBITDA

• For FY2025, reconcile reported adjusted EBITDA back to GAAP EBITDA using the reconciliation table in the 10-K/10-Q.
• List each add-back and classify as truly non-recurring, recurring, or ambiguous; treat items appearing 2+ times in 3 years as recurring.
• Require management support for large add-backs (> 5% of EBITDA) and check footnotes for related-party or acquisition timing effects.

Quick quantitative check

• Compute D&A + capex as a percent of EBITDA for FY2025. If greater than 30%, EV/EBITDA likely understates investment needs and overstates relative cheapness.
• Example math: EBITDA $200m, D&A $40m, capex $30m → D&A+capex = $70m = 35% of EBITDA → flag for deeper review.

Actionable adjustments

• Build adjusted-EBITDA variants: GAAP EBITDA, adjusted EBITDA (true non-recurring removed), and cash-EBITDA (subtract maintenance capex and normalized working-capital change).
• Compare valuation across multiples: EV/EBITDA vs EV/EBIT vs EV/FCF. If rankings flip, rely on cash-based multiples.

What this estimate hides: the 30% rule is a rule-of-thumb - industry norms differ (software vs utilities), so calibrate by peer group and asset life tables. Also, defintely read the footnotes.

Capital-structure and sector mismatches

EV includes net debt while EBITDA ignores interest-different leverage skews comparability

You should treat EV/EBITDA as a leverage-aware but imperfect shortcut: enterprise value (EV) = market cap + net debt (debt minus cash), while EBITDA ignores interest expense and hence the cost of that debt.

One-liner: EV tells you how much of the business you're buying, EBITDA tells you how the business performed before financing-mixing them without adjustments hides leverage risk.

Steps to use this correctly:

• Compute EV precisely: market cap + net debt + minority interest + preferred stock - cash.
• Calculate leverage: net debt / trailing-12-month (LTM) EBITDA using the 2025 fiscal-year LTM when available.
• Compare peers on net debt/EBITDA, not just EV/EBITDA.
• Run sensitivity: debt costs +1% and EBITDA -10% to test covenant and valuation impact.

Example (illustrative, use your 2025 LTM numbers): EBITDA $200m, net debt $900m → net debt/EBITDA = 4.5x. That changes the story even if EV/EBITDA looks cheap.

What this hides: higher leverage raises default and refinancing risk, and interest tax shields vary across jurisdictions-so two firms with identical EV/EBITDA can have very different equity risk.

Negative equity or large minority interests can make EV/EBITDA meaningless

If a company has negative shareholders' equity, or significant minority interests, headline EV/EBITDA often breaks down. Negative equity can make market cap small or zero while liabilities dominate; minority stakes mean you're valuing unowned cash flows incorrectly.

One-liner: when balance-sheet oddities appear, EV/EBITDA stops being a clean yardstick-do the math, don't trust the single multiple.

Practical checks and fixes:

• Recompute EV adding back minority interest and preferred securities where applicable.
• When equity is negative, prefer debt-adjusted metrics: debt/EBITDA, interest coverage, or EV/FCF.
• For conglomerates with minority stakes, use sum-of-the-parts (SOTP) valuations and value each segment on the right multiple.
• Flag and document anomalies: negative tangible equity, large deferred tax liabilities, or sizeable off-balance-sheet leases.

Example adjustment: if market cap = $50m, net debt = $500m, and minority interest = $120m, the meaningful EV component is debt + minority, so equity multiple alone misleads and you should pivot to cash-flow measures.

What to watch: regulatory capital firms (insurers, banks) or firms with large pension deficits-they need capital-ratio metrics, not EV/EBITDA.

Financials and REITs: EV/EBITDA usually inappropriate; use book-based or NAV metrics instead

Banks, insurers, and real estate investment trusts (REITs) have business models where interest, leverage, and asset values are central; EBITDA removes the items you actually need. For these sectors, use price-to-book, tangible book, or net asset value (NAV) and sector-specific earnings measures like funds from operations (FFO).

One-liner: for financials and REITs, value the balance sheet first and cash flows second-don't force EV/EBITDA.

Actionable alternatives and steps:

• For banks/insurers: use price-to-book (P/B), return on tangible equity (ROTE), and regulatory capital ratios (Tier 1 CET1). Reconcile market cap to tangible equity per share.
• For REITs: use FFO (funds from operations) and AFFO (adjusted FFO), plus NAV per share and cap rates on properties. Convert NAV to per-share and compare to market price to gauge discount/premium.
• For asset-heavy corporates (utilities, telcos): prefer EV/EBIT (operating profit), EV/FCF, or FCF yield; ensure you subtract maintenance capex when computing free cash flow.
• Best practice: when you see EV/EBITDA used on a bank or REIT model, stop and rebuild using the sector-appropriate metric set.

Rule of thumb: flag any firm with net debt/EBITDA > 4x or revenue largely driven by financial spread or real estate-then switch to balance-sheet metrics and stress-test asset values under 10-20% downside scenarios.

Owner action: you should run a three-year normalized EBITDA and an EV/FCF check for targets; Finance: produce that by Friday so you have a cash-focused comparator for the 2025 fiscal-year view. defintely run covenant and NAV sensitivity while you're at it.

Growth, lifecycle, and business-model biases

High-growth and early-stage firms - EV/EBITDA punishes optionality

You're looking at a company growing quickly but with negative or thin EBITDA; EV/EBITDA will usually make it look expensive, often misleadingly so.

One-liner: start here, not stop here.

Why this matters: EV/EBITDA values current operating earnings and ignores future scale benefits and margin expansion that high-growth firms may capture. If EBITDA is negative, the multiple is undefined or meaningless.

Practical steps

• Switch to forward metrics: use next‑12‑month (NTM) EV/Revenue or explicit DCF projections.
• Build a simple option-adjusted model: value core business with a DCF and add scenario upside for new markets.
• Stress-test margins: run cases where gross margin improves by 200-500 bps over 3 years.
• Cap valuation upside: convert revenue growth into free cash flow only after gross margins and S&M efficiency reach plausible bands.

Quick math: if revenue is $200m and growing 40% with current negative EBITDA, an EV/Revenue of 6x implies EV $1.2bn; convert to DCF to see whether that EV implies achievable cash flows.

What this estimate hides: customer retention, unit economics, and future capital needs. If churn is high or payback periods exceed 24 months, the growth option evaporates-so check CAC payback and gross retention.

SaaS and subscription models - prefer ARR and unit economics

SaaS firms often report low or volatile EBITDA early because they invest heavily in sales and R&D; multiples tied to earnings miss recurring revenue value.

One-liner: use EV/ARR or EV/Revenue instead of EV/EBITDA for most SaaS comparatives.

Practical guidance

• Use EV/ARR (enterprise value / annual recurring revenue) and Rule of 40 to balance growth and profitability.
• Run unit-economics: compute LTV/CAC, gross retention, and CAC payback in months.
• Adjust EBITDA for timing: capitalize and amortize onboarding spend if it creates multi-year revenue.
• Segment revenue: separate core subscription ARR from professional services and one-offs.

Concrete thresholds: flag models where LTV/CAC < 3x or CAC payback > 24 months. If ARR growth is 30%+ and retention is > 90%, EV/ARR comparables give clearer valuation signals than EV/EBITDA.

Example framework: for a SaaS with $100m ARR, 30% growth, and 92% retention, compare peers on EV/ARR; translate multiples into implied FCF by modeling margin improvement constrained by historical retention and S&M efficiency.

Mature, capex-heavy industries - prefer EBIT or cash-flow metrics

Utilities, telecoms, and heavy industry sustain high depreciation (D&A) and steady capex; EBITDA strips those costs and overstates free cash available to equity.

One-liner: in capex-heavy sectors, use EV/EBIT, EV/FCF, or regulated NAV metrics over EV/EBITDA.

Actionable checklist

• Compare capex to EBITDA: if D&A + capex > 30% of EBITDA, EV/EBITDA understates investment needs.
• Compute EV/FCF (enterprise value / free cash flow) and FCF yield to capture true cash conversion.
• Separate maintenance vs growth capex; model maintenance capex as an ongoing cash outflow tied to revenue or assets.
• When regulated, use book or NAV multiples and cross-check with regulated ROE or allowed returns.

Quick math: a telco with EBITDA $1bn, D&A $400m, and capex $500m has operating cash constrained-EBITDA ignores the $900m of non-discretionary charges, so EV/EBITDA will understate capital needs.

Practical caveat: if leverage is high or assets are aging, depreciation policies and capex timing can swing FCF sharply-so forecast 3-5 years of capex and run sensitivity scenarios; defintely stress the case where maintenance capex rises by 20%.

Owner and next step: you or Finance should build a 3‑year EV/FCF and EV/EBIT bridge for each target, flagging where EV/EBITDA deviates by > 25% from cash-flow-based valuations.

Cyclicality, one-offs, and normalization

You're valuing a company with lumpy results and wondering why single-year EV/EBITDA swings so wildly; the quick answer is that one-year EBITDA often reflects cycles, one-offs, or accounting moves - use multi-year normalization and targeted adjustments instead.

Here's the quick math you'll use repeatedly: take a 3-5 year EBITDA average, remove clear one-offs, and reconcile for accounting changes before you plug EBITDA into any multiple or DCF.

Cyclical earnings and multi-year normalized EBITDA

If a firm's revenue or margins move with commodity prices, end-market demand, or investment cycles, one-year EBITDA can be a bad guide to sustainable earning power. Pick a normalization window that matches the cycle length: commodity/energy might need 5 years; industrials often 3 years.

Practical steps:

• Collect EBITDA for the last 3-5 fiscal years and current-year trailing twelve months (TTM).
• Calculate simple average and median; compare to TTM and latest fiscal year.
• If average and median diverge >10%, investigate cyclical drivers (prices, volumes).
• Use the average for multiples, and run sensitivity at ± one standard deviation.

Example: FY EBITDA (FY2021-FY2025) = $100m, $200m, $50m, $150m, $120m; 5-year average = $124m. Use $124m as the normalized EBITDA baseline, not the FY2025 $120m alone - defintely check the drivers.

One-liner: use a 3-5 year average, not a single-year snap shot.

Adjusting for cyclical peaks/troughs and material one-offs

Companies report gains, losses, disposals, and restructuring that distort EBITDA. You must separate economic recurring earnings from transactional noise before comparing multiples.

Practical checklist:

• Flag items labeled non-recurring, one-time, or extraordinary in notes.
• Quantify cash vs. non-cash impact (e.g., disposal gain $50m is non-cash gain).
• Reclassify recurring items mislabelled as one-offs (warranty, litigation, carve-outs).
• Create pro forma EBITDA removing true one-offs and reinstating recurring costs.

Best practices: document each adjustment with citation to the 10-K/10-Q note, show cash flow impact, and present both reported and adjusted EV/EBITDA in your model so reviewers see the bridge.

Example: reported EBITDA $200m with a disposal gain of $50m and a restructuring charge of $20m (cash out next 12 months). Adjusted EBITDA = $200m - $50m + (exclude restructuring only if non-recurring) = $150m.

One-liner: strip true one-offs and show reported vs adjusted EBITDA side-by-side.

Watch for accounting changes that shift EBITDA without economic change

Accounting rule changes or policy shifts (lease accounting, revenue recognition, capitalization of development) can move EBITDA mechanically. Don't treat these as economic improvement unless cash flow changes too.

Steps to handle accounting shifts:

• Read the footnotes for ASC 842/IFRS 16, ASC 606, and capitalization policies for R&D or software.
• Restate historical EBITDA where possible to pre-change basis, or create a consistent pro forma series.
• Adjust for lease reclassification: convert lease expense to EBITDA-neutral by adding back interest and depreciation equivalents if you want comparability.
• Note tax and cash implications separately - EBITDA moves may not change free cash flow (FCF).

Example: a company adopts capitalization of software development, reducing expense and increasing EBITDA by $30m. If capitalized spend still produces the same cash outflow, restate prior years to add back the capitalization to create a consistent EBITDA series.

One-liner: reconcile accounting-driven EBITDA jumps to a cash-consistent baseline before valuing.

Examining the Limitations of EV/EBITDA - Practical adjustments and alternatives

Adjust EBITDA: normalize non-recurring items, working capital, and capex

You're using EV/EBITDA to screen targets, so start by cleaning the earnings line before you compare multiples.

One-liner: adjust first, compare second.

Practical steps:

• Isolate true non-recurring items: disposals, litigation settlements, and restructuring costs that are clearly one-off.
• Normalize working capital swings: convert seasonal or invoice-timing moves into a multi-year average change in working capital (use 3-5 years).
• Separate maintenance capex from growth capex: maintenance capex is the spend required to keep current EBITDA intact; treat growth capex as investment, not as recurring EBITDA support.

How to do the math (quick example): assume FY2025 reported EBITDA = $150m, add-backs for one-offs $12m, normalized working capital drain +$8m, and maintenance capex estimated at $50m. Adjusted EBITDA for cash-generation purposes = $150m + $12m - $8m - $50m = $104m.

What this estimate hides: maintenance capex is judgement-heavy; if you understate it, you overstate adjusted EBITDA. Always document capex drivers (replacement cycles, useful lives, regulatory spend).

Use EV/EBIT, EV/FCF, or FCF yield when cash matters

If you care about the cash that can service debt, pay dividends, or fund buybacks, move from an earnings multiple to a cash multiple.

One-liner: cash metrics beat EBITDA for cash questions.

Best practices:

• Use EV/EBIT (enterprise value to operating profit) when depreciation maps to real economic wear-and-tear-utilities, telcos, industrials.
• Use EV/FCF (enterprise value to free cash flow) or FCF yield (FCF/market cap) for buyback/dividend capacity and solvency analysis.
• Standardize FCF: start from operating cash flow, subtract maintenance capex, and adjust for normalized working capital.

Example methodology (FY2025 style): calculate FCF = cash from operations $220m minus maintenance capex $60m = $160m. If EV = $2.4bn, then EV/FCF = 15x and FCF yield = 6.7% (FCF/market cap if market cap = EV - net debt).

Limits to watch: one-year FCF is noisy in cyclical sectors; use a 3-year average or normalized run-rate for valuation comparisons.

Adjust for capex and use segmented or sum-of-the-parts multiples for complex firms

Diversified or capital-intensive businesses hide value when you roll everything into a single EV/EBITDA multiple.

One-liner: break it up, or you'll miss the cash drivers.

Practical tactics:

• Compute EV/EBITDA-CAPEX or EBITDA minus maintenance capex as an alternative denominator. Label it EBITDA-M to be clear.
• Run segmented multiples: value divisions on their relevant metrics-EV/ARR or EV/Revenue for SaaS, EV/EBIT for heavy industry, NAV per share for REIT-like assets.
• When one division dominates cash, use sum-of-the-parts (SOTP): value each segment on its best multiple, then consolidate and subtract net debt.

Concrete example (illustrative FY2025 split): Segment A (SaaS): revenue $300m, use EV/ARR = 6x; Segment B (Industrial): EBITDA $120m, maintenance capex $40m, use EV/(EBITDA - maint. capex). If EV apportioned before SOTP = $3bn, SOTP shows a different picture than a single EV/EBITDA multiple of 10x.

Execution checklist:

• Document segment accounting and intercompany allocations.
• Estimate maintenance capex per segment from fixed-asset schedules and capex history.
• Apply conservative exit multiples for loss-leading or cyclical units.

Small note: segmented SOTP is data-intensive and often reveals valuation gaps worth a deeper due diligence spend - defintely run it for diversified targets or if net debt/EBITDA > 4x.

Examining the Limitations of EV/EBITDA - Practical Actions

You're using EV/EBITDA as a quick screen to compare targets, and that's fine - but you need clear checks so it doesn't steer you wrong. Below are direct, actionable steps to stop treating a single multiple as the final answer and to force-capture the cash, capital, and growth differences that matter.

EV/EBITDA is a fast but has clear blind spots

One-liner: EV/EBITDA is a good starting point, not the decision.

Use EV/EBITDA to shortlist names, then immediately test four blind spots: capital intensity, leverage, growth profile, and cyclicality. If a company's combined depreciation plus capex consistently exceeds 30% of EBITDA, the multiple will understate the true investment need. If net debt/EBITDA is above 4x, leverage differences distort comparability. For high-growth firms with revenue growth above 20% and negative EBITDA, EV/EBITDA throws away optionality and future margin expansion. For cyclical businesses, a single-year EBITDA can be misleading - prefer a 3-5 year averaged EBITDA to smooth cycles.

• Flag D&A + capex > 30% of EBITDA
• Flag net debt/EBITDA > 4x
• Flag revenue growth > 20% with negative EBITDA

Normalize EBITDA, check capex and net debt, and run a cash metric

One-liner: Adjust the earnings number and check the cash math before you value.

Practical steps you can run this week:

• Collect FY2023-FY2025 income statements, cash flow, and balance sheets.
• Calculate 3-year normalized EBITDA = (EBITDA2023 + EBITDA2024 + EBITDA2025) / 3.
• Adjust EBITDA: remove true one-offs, reinstate recurring restructuring, and normalize working-capital swings.
• Estimate maintenance capex (industry rule or % of revenue) and compute EBITDA minus maintenance capex for a cash-focused proxy.
• Compute EV/FCF: EV divided by free cash flow to firm (post-maintenance capex, pre-debt service).

Here's the quick math with a compact example (illustrative): FY2023 EBITDA $900m, FY2024 $1,100m, FY2025 $1,200m. Three-year normalized EBITDA = (900 + 1,100 + 1,200) / 3 = $1,067m. If enterprise value = $12,000m, EV/normalized EBITDA ≈ 11.25x. If FCF (post-maintenance capex) = $800m, EV/FCF = 15x. What this estimate hides: maintenance capex assumptions and working-capital normalizations; be explicit about both.

Owner: run a three-year normalized EBITDA and EV/FCF comparison for targets

One-liner: Assign one owner, produce a short pack, and use explicit flags.

Action steps for you or Finance (owner):

• Gather FY2023-FY2025 reported EBITDA, D&A, capex, working-capital changes, net debt, and reported EV.
• Build a one-page table per target with columns: EBITDA by year, normalized EBITDA, D&A+capex as % of EBITDA, net debt/EBITDA, EV/normalized EBITDA, FCF (post-maintenance capex), EV/FCF.
• Highlight flags: D&A+capex > 30%, net debt/EBITDA > 4x, revenue CAGR > 20% with negative EBITDA.
• Deliverable: a three-sheet pack - (1) raw data, (2) normalized metrics and ratios, (3) one-page recommendation (buy/watch/skip) per target.
• Timeline: compile data within 3 business days, deliver pack in 5 business days.

Quick sample table (illustrative numbers):

Owner: you or Finance - run the three-year normalized EBITDA and EV/FCF comparison for each target, flag the three checks above, and present the pack in five business days. If maintenance capex guesses exceed 25% of EBITDA, defintely escalate for revisit of valuation assumptions.