EBITDA vs EBIT — Why Mining Companies Demand a Different Standard

Reading the income statement

Both EBITDA and EBIT are derived from the same income statement. The difference is simply where you stop — and whether you treat depreciation & amortisation as a real cost or an artefact of accounting policy.

Illustrative income statement Figures in A$m

Top line
Revenue	1,000
Cost of goods sold (COGS)	(600)
Gross profit	400
Operating expenses
SG&A	(120)
R&D	(50)
Other opex	(30)
EBITDAStop here →	200
Depreciation & amortisation (D&A)	(40)
EBIT — operating profitOr here →	160
Below the line
Net interest expense	(20)
Earnings before tax (EBT)	140
Income tax (30%)	(42)
NPAT — net profit after taxBottom line	98

The gap between EBITDA ($200m) and EBIT ($160m) is entirely the $40m D&A charge. In most industries, adding back D&A is analytically defensible — it neutralises differences in accounting policy across peers. In coal mining, it erases a cost that is economically real.

Why most industries default to EBITDA

Before examining where EBITDA breaks down, it is worth understanding why it became the dominant operating metric in the first place. There are five reasons analysts, bankers, and acquirers reach for it — and each is genuinely defensible in the right context.

1 — It neutralises capital structure

EBIT subtracts D&A but still sits above the interest line. EBITDA strips out interest entirely. When comparing two otherwise identical businesses — one conservatively financed with equity, one leveraged — EBIT still penalises the leveraged one for its interest cost. EBITDA removes that distortion, letting you assess the underlying operating business independent of how the owners chose to finance it. This is especially important in private equity, where the same business might be acquired at varying leverage levels by different buyers.

2 — It removes geography and tax jurisdiction

Tax rates vary across countries, states, and even industries (concessional mining royalties, R&D offsets, accelerated depreciation regimes). A business earning $100m EBIT in Australia faces a different NPAT than one earning $100m EBIT in Singapore — not because the operating business is different, but because of where it is domiciled. EBITDA sits above all of that, enabling genuinely cross-border peer comparisons.

3 — It approximates operating cash flow

In asset-light businesses — SaaS platforms, media companies, consumer brands, professional services firms — working capital movements are modest and capex is minimal. In those settings, EBITDA is a reasonable proxy for free cash generation because D&A is genuinely close to zero in economic terms. The business does not need to spend much to maintain its productive capacity, so the depreciation add-back does not introduce meaningful distortion.

The SaaS illustration: a software company with $200m revenue and $80m in capitalised development costs amortises that over, say, 3 years. But the next three years of development spending is already underway and will replace it. If the business is growing, the amortisation charge is more than offset by the growth in the productive asset base. EBITDA is a reasonable operating metric because the D&A charge does not represent a real declining-balance on a finite asset — it is a rolling capitalisation of ongoing investment.

4 — It is the basis for most transaction multiples

LBO models, M&A deal pricing, and leveraged loan covenants are almost universally structured around EBITDA. This is partly self-reinforcing — investment banks quote EV/EBITDA because buyers and sellers expect EV/EBITDA — but it also reflects that EBITDA is the closest income-statement figure to the pre-tax, pre-financing cash flow a new owner inherits when they acquire a business. Debt capacity in leveraged buyouts is typically sized as a multiple of EBITDA (e.g. 4×–6× Net Debt / EBITDA), so the metric directly determines how a deal is structured and priced.

5 — D&A is genuinely subjective

Two companies with identical physical assets can report meaningfully different depreciation charges depending on their assumed useful lives, residual values, and methodology (straight-line vs declining balance vs units-of-production). EBITDA sidesteps that subjectivity entirely. In a cross-sector screen where you cannot easily normalise each company's depreciation policy, stripping it out produces a cleaner comparison than attempting imperfect adjustments.

The bottom line: EBITDA is the right default metric when D&A is either genuinely immaterial (asset-light businesses), genuinely replaced by ongoing investment (growing tech companies), or when capital structure and tax jurisdiction differences would otherwise dominate the comparison. It is a good tool used in the right context. The problem is applying it universally — particularly to businesses where the asset base physically depletes with production.

EBITDA applicability by industry Stylised assessment

Industry	Asset intensity	Asset depletes?	EBITDA usefulness
SaaS / software	Low	No	High — D&A minimal
Consumer brands / media	Low–medium	No	High — capex light
Retail / distribution	Medium	No	Moderate — lease D&A distorts (AASB 16)
Industrials / manufacturing	High	Partial	Moderate — equipment replacement real
Infrastructure / utilities	Very high	Partial	Low — long-life assets, regulated RAB
Oil & gas	Very high	Yes — depletion	Low — DD&A is real cost
Coal / hard rock mining	Very high	Yes — finite reserve	Low — use EBIT or EBITDA-SX

Charlie Munger called it bullshit earnings

The five reasons above explain why EBITDA became universal. They do not make it intellectually honest. Charlie Munger — Buffett's partner at Berkshire Hathaway and one of the most rigorous minds in the history of capital allocation — had a characteristically direct view on the matter.

Every time you see the word EBITDA, you should substitute the words bullshit earnings.

Charlie Munger · Poor Charlie's Almanack

This was not a throwaway line. Munger and Buffett returned to the EBITDA critique repeatedly over decades — at Berkshire annual meetings, in shareholder letters, and in interviews — because they believed the metric was not merely imprecise but actively deceptive. Understanding their argument requires unpacking three specific objections.

Objection 1 — Depreciation is a real cost, full stop

Munger's core charge was simple: the "D" in EBITDA represents money that has already been spent — on trucks, plant, equipment — the economic benefit of which is being consumed with every hour of production. Adding it back does not make the cost disappear. It just hides it from the income statement.

His illustration, delivered after a Berkshire annual meeting, was blunt: you have to replace the trucks all the time. You paid for them in advance and received the benefit gradually over years. That timing mismatch is precisely why depreciation exists as a concept — to match the cost to the period of benefit. Stripping it out via EBITDA unmatches the cost and the revenue again, which is the opposite of what accrual accounting was designed to achieve.

Buffett framed it as a question: does management think the tooth fairy pays for capital expenditures? The asset does not maintain itself. Someone pays. D&A is the income statement acknowledging that someone.

Munger's truck analogy: a haul truck bought for $5m and depreciated over five years costs $1m per year to "use up." That $1m depreciation charge is not a fiction — it is the annual slice of a real cash outlay already made. Adding it back to earnings makes the business look $1m more profitable per year. But the truck still wore out. Someone still paid $5m. The cash left the building five years ago.

Objection 2 — It was invented to serve deal-makers, not investors

EBITDA first rose to prominence in the leveraged buyout boom of the 1980s. Investment banks needed a metric that could justify paying high acquisition prices for asset-heavy cable and telecom businesses that were consuming enormous capital. By stripping out depreciation, the businesses looked far more profitable — and far more capable of servicing the debt piles that LBO sponsors were loading onto them.

Munger was explicit about this origin and explicit about his contempt for it. The metric was constructed to make transactions easier to execute, not to give investors an accurate picture of economic reality. The investment banking community then promoted it into mainstream financial discourse because their fee income depended on the deals it enabled.

80s

LBO boom — EBITDA invented as a debt-service proxy

Investment banks needed to justify high leverage on asset-heavy cable and telecom acquisitions. EBITDA gave a pre-D&A earnings figure that could service debt on paper — even when the business was consuming more capital than it generated.

90s

Telecoms adopt it universally — WorldCom collapses

EBITDA becomes the de facto operating metric for capital-intensive telecoms. WorldCom hides $3.8 billion in costs by misclassifying operating expenses as capex — inflating EBITDA while burning cash. The fraud is only visible when cash flow statements are read.

now

Adjusted EBITDA — Munger's "horror squared"

Companies now routinely strip out stock-based compensation, restructuring charges, acquisition costs, and other items to arrive at "adjusted EBITDA." Munger described this escalation as announcing to the world that you are a flake. Each adjustment reduces the metric's connection to cash reality further.

Objection 3 — Adjusted EBITDA is where it gets truly dangerous

If raw EBITDA troubled Munger, adjusted EBITDA — now standard in earnings releases, investor presentations, and analyst models — was what he called the full horror. Companies strip out stock-based compensation (a real cost of retaining employees), restructuring charges (sometimes recurring), acquisition costs (the price of growth), and any number of other items that management deems non-recurring. The resulting figure can bear almost no relationship to the cash the business actually generates.

Seth Klarman, another value investor of the same school, observed that those who used EBITDA as a cash flow proxy either ignored capital expenditures or assumed that plant and equipment do not wear out. The adjusted EBITDA variant compounds the problem by also ignoring the dilution of equity compensation, the cost of periodic restructuring, and the integration costs of acquisitions — which, in serial acquirers, are essentially permanent.

Where Munger's critique leaves room for EBITDA: Munger was not arguing that the concept is useless in every context. He was arguing that it is intellectually dishonest to present EBITDA as a substitute for earnings — which is exactly how it is routinely used. In a business where D&A genuinely does not represent a real recurring cash cost (because the assets are not wearing out or depleting), the metric is defensible. The problem is the universality of its application, not the concept itself.

Why the critique hits hardest in coal mining

Munger's general objection — that depreciation is a real cost — becomes a structural objection in extractive industries. In coal mining, every one of his three charges is amplified:

✕

Equipment D&A is not optional. You cannot run a coal mine without replacing draglines, trucks, and wash plant. The depreciation charge is a certainty, not an estimate.

✕

Mine development amortisation is production-matched. Under units-of-production methodology, every tonne sold triggers an amortisation charge — the accounting is directly tied to the depletion of the reserve.

✕

The asset is finite and irreplaceable. Unlike a SaaS company whose code base can be updated indefinitely, the coal seam is gone when mined. EBITDA implicitly treats the reserve as if it has infinite life.

Munger's critique, applied to a general business, is a philosophical objection about intellectual honesty. Applied to a coal mine, it becomes a quantitative one: the D&A charge in a coal producer can represent 15–40% of EBITDA. Ignoring it does not produce a slightly misleading multiple — it produces a structurally wrong one.

Why D&A is not noise in a coal mine

The conventional argument for EBITDA is that depreciation is a non-cash accounting estimate — a distortion to be stripped away so you can compare operating performance across companies with different asset bases. That argument has merit in asset-light businesses. It collapses in extractive industries.

In a coal mine, D&A captures three economically real costs that compound one another:

01

Equipment wear & replacement

Draglines, haul trucks, and wash plants are consumed producing every tonne. The depreciation charge is the honest signal of an inescapable future capex obligation. EBITDA pretends it doesn't exist.

02

Mine development amortisation

Pre-strip costs, development drilling, and decline construction are capitalised under AASB 116 and amortised on a units-of-production basis. This is genuine cost-matching — not a policy artefact.

03

Resource depletion

Unlike a software company, the asset base is the coal itself. Every tonne sold is permanently removed from the reserve. EBITDA treats the mine as if it could generate that cash forever.

The fundamental test: would you buy a coal mine and expect to earn EBITDA in perpetuity without spending another dollar? No — you'd need to replace equipment, redevelop faces, and eventually find a new orebody. D&A is the income statement's attempt to price that reality.

The comps problem — three distortions EBITDA introduces

When building a peer universe for a company like New Hope Corporation, EBITDA multiples systematically mislead because the mines being compared have materially different D&A profiles for reasons entirely unrelated to operating quality.

1 — Mine age and depreciation curves

A mature mine with largely written-down assets carries very low D&A. A recently expanded or greenfield operation carries high D&A on fresh capex. EV/EBITDA will make the new mine look expensive even if the two operations are economically equivalent — the multiple difference reflects the asset's age, not its quality.

2 — Mining method

Open-cut and underground operations have fundamentally different capex intensities. Underground mines require continuous development expenditure — roadways, ventilation, longwall moves — that capitalises and amortises continuously. EBITDA blurs this distinction entirely; EBIT preserves it.

3 — Strip ratio and capitalised stripping

High strip ratio mines capitalise more pre-strip under the stripping activity asset model, which flows back through as higher amortisation over the mine's life. EBITDA treats a high-strip and a low-strip mine as identical on the operating line. They are not identical — the amortisation charge is the accounting trace of a real economic difference.

The effect of all three distortions is illustrated below with a hypothetical peer set:

Illustrative comps — thermal coal peers Multiples may be distorted by D&A profile

Company	Mine type	D&A / EBITDA	EV/EBITDA	EV/EBIT	Signal
NHC (New Hope)	Open-cut, mature	18%	4.2×	5.1×	Comparable
Peer A	Open-cut, expansion	34%	5.8×	5.3×	EBITDA overstates cost
Peer B	Underground, high strip	41%	6.4×	5.2×	Looks expensive on EBITDA
Peer C	Open-cut, written-down	9%	3.9×	4.3×	Comparable

Peers A and B look meaningfully more expensive on EBITDA. On EBIT, the spread compresses considerably — reflecting that much of the apparent difference was the D&A profile, not underlying business quality. An analyst who screens purely on EV/EBITDA may incorrectly rank these peers.

EBITDA vs EBIT — head to head

EV / EBITDA

Neutralises capital structure differences
Useful for asset-light, high-margin businesses
Strips out D&A as accounting "noise"
Industry standard for cross-sector screens
Good for SaaS, media, consumer brands

vs

EV / EBIT

Preserves real depreciation cost signal
Penalises capex-heavy operations appropriately
Better peer comparability in extractives
Reflects mine age and development profile
Preferred for mining, energy, infrastructure

The verdict for NHC: EV/EBIT is the more reliable comparator when building a thermal coal peer universe. It is not perfect — D&A still depends on useful life assumptions and methodology — but it is materially less distorted than EBITDA across a peer set with divergent mine ages, strip ratios, and mining methods.

The superior metric — EBITDA minus sustaining capex

EBIT is a better proxy than EBITDA in mining, but it is still an accounting-derived figure. The metric many mining analysts prefer for comps is EBITDA minus sustaining capex — sometimes written as EBITDA-SX or EBITDA-capex.

The logic: sustaining capex is the cash equivalent of economic depreciation. If you subtract it from EBITDA, you get a proxy for true free cash generation that is less sensitive to accounting policy than EBIT but more economically grounded than raw EBITDA.

Preferred metric — mining & extractives

EBITDA − Sustaining Capex

Sustaining capex is the actual cash a mine must spend to maintain productive capacity — equipment replacement cycles, ongoing development, tailings facility works. It is the cash expression of what D&A tries to estimate. Subtracting it from EBITDA gives you a figure that is comparable across the peer set without relying on each company's depreciation methodology.

EV / (EBITDA − Sustaining Capex)

The practical limitation: sustaining vs growth capex classification varies by company and is often disclosed with inconsistency. For NHC and ASX-listed coal peers, the sustaining capex split is typically available in operational updates and investor day presentations — but it requires normalisation before use in a comp table.

When sustaining capex disclosure is unreliable, EV/EBIT remains the most accessible and defensible comparator for thermal coal peer analysis.

Operating cost metrics as a complement

For NHC specifically, EV/EBIT or EV/(EBITDA−SX) should be read alongside cost-per-saleable-tonne and AISC (all-in sustaining cost). The earnings yield of a coal producer is fundamentally a function of the spread between realised coal price and mine-gate cost — that is where genuine peer differentiation shows up, and no multiple captures it directly.

Summary

The EBITDA add-back rests on the premise that D&A is accounting noise with no economic content. That premise is defensible for businesses where the asset base does not physically deplete with use. For a coal mine, it is incorrect — and using EBITDA as the primary comps metric can introduce systematic ranking errors across a peer universe.

The hierarchy of metrics for thermal coal comps, in order of analytical preference:

01

EV / (EBITDA − Sustaining Capex) — most economically accurate; requires reliable capex segmentation

02

EV / EBIT — good comparator; better than EBITDA; sensitive to depreciation policy

03

Cost per saleable tonne / AISC — operational benchmark; complements any multiple

—

EV / EBITDA — use with caution; adjust for D&A divergence or use only within tightly matched sub-groups