What is an EBITDA bridge?

An EBITDA bridge is a waterfall that explains how a company's EBITDA moved from one point to another, and why. It starts at one EBITDA figure, ends at another, and fills the gap with bars that each attribute part of the change to a single driver: volume, pricing, cost, acquisitions, one-time items. Instead of "EBITDA went from $40M to $52M," a bridge tells you exactly which $12M came from where.

What the bridge actually answers

A single EBITDA number tells you the result. It doesn't tell you the story. A bridge converts "what happened to earnings?" into "which specific levers moved, and by how much?" — which is why it shows up wherever someone has to explain or defend an earnings number: board decks, budget-vs-actual, investor presentations, and private-equity value-creation plans. In PE especially, the bridge is often the single most important page in the plan.

The discipline it enforces is ownership. Each bar is a commitment with a name attached — pricing to the commercial team, cost to operations, bolt-ons to the deal team. When the board asks why EBITDA is behind, the bridge turns a vague question into a precise one: which bar missed, by how much, and who owns the recovery.

The two bridges people confuse

There are two different things called an "EBITDA bridge," and they answer different questions.

The change-over-time bridge tracks movement; the reported-to-adjusted bridge cleans a snapshot. A diligence pack needs both, and the classic error is mistaking a one-time adjustment for a recurring driver of growth.

The standard drivers

A good bridge uses a clean, non-overlapping set of drivers so the bars add up exactly to the total change with nothing double-counted. The common categories are:

For multinationals, a foreign-exchange bar usually comes first, isolating currency translation so the operating drivers are measured on a constant-currency basis.

A worked example: a value-creation bridge

A sponsor buys a business at $40M EBITDA and underwrites $65M at exit. The bridge decomposes the planned $25M of growth into bars that each add up to the target:

Now each line is a testable commitment rather than a hopeful total. If pricing delivers only +$4M by mid-hold, the board sees immediately that the gap is $4M on the pricing bar specifically, and that the remaining quarters have to carry more growth than planned. That visibility is the whole point.

A real-company bridge: NVIDIA Q1 FY2027

The hypothetical above keeps the drivers clean; a real bridge is messier, but the logic holds. The walk from NVIDIA's Q1 FY2027 10-Q $58.3B net income to ~$72.9B adjusted EBITDA is short: add back taxes and a thin depreciation line to reach EBITDA, then layer the debated add-backs (stock-based compensation, one-time items) to reach the adjusted figure.

Context matters as much as the number. An ~89% adjusted EBITDA margin only means something set next to peers:

And the single most contested line is the one every reported-to-adjusted bridge argues over — stock-based compensation. Add it back and the margin is 89%; leave it in and it's 86%. The discipline is to show both and state which you used.

See the full interactive example: EBITDA Bridge — NVIDIA Q1 FY2027 ↗

The reported-to-adjusted bridge in M&A

The reported-to-adjusted version is where the bridge becomes the most contested document in a deal, because valuation runs straight through it.

Buyers don't pay a multiple on reported EBITDA. They pay it on adjusted EBITDA, the figure both sides agree reflects sustainable, repeatable earnings. So sellers (usually with their advisors) build an adjusted EBITDA bridge that walks the buyer from statutory EBITDA up to the adjusted number, with each add-back justified: excess owner compensation normalized to a market-rate salary, personal expenses removed, one-time legal or consulting fees stripped out, above-market related-party rent restated. The math is leverage: at a 5x multiple, a defensible $100,000 add-back is worth $500,000 of enterprise value.

That's also why the bridge gets tested hard. It isn't a financial statement so much as the opening position in a negotiation, and every number on it will be challenged. In any institutional process the buyer commissions a quality of earnings (QoE) review, which is effectively an independent audit of the seller's adjusted EBITDA, testing each step against the general ledger, revenue recognition, and margins.

Edge cases and common errors

• Keep the drivers MECE. Bars must be mutually exclusive and collectively exhaustive, or the same dollar gets counted twice. Price and volume bleed into each other — fix a convention (price = realized rate at constant volume) and hold it.
• Fix the sign convention and reconcile. Decide up front what's positive and negative, and make the bridge foot back to the actual EBITDA delta; a bridge that doesn't reconcile is wrong somewhere.
• Put FX first. For a multinational, isolate currency translation in the first bar so operating drivers are on a constant-currency basis.
• Don't treat an adjustment as a driver. A reported-to-adjusted add-back is a one-time restatement, not recurring growth — mixing the two overstates the run-rate.
• Stock-based comp is the most contested line. Add it back and margins look higher; leave it in and they don't. Show both and state which you used.
• Beware order-dependence in mixed bridges. When a walk contains both operating drivers and normalizations, the sequence changes how each bar reads — group and label them so a reviewer can follow the logic.

Where each figure comes from

MacrosLM's QoE & EBITDA Bridge Automator builds the reported → adjusted → normalized bridge from the general ledger, ERP exports, invoices, and contracts you drop in — not from memory. It drafts the add-back schedule, flags items that recur across periods (the ones a buyer will challenge), and ties every figure back to its source document. The judgment stays with you: whether an add-back is truly non-recurring, whether a driver is sustainable, and which adjustments are worth defending before a buyer reverses them.

Bottom line

An EBITDA bridge explains how EBITDA got from one number to another, driver by driver, and assigns each piece an owner. In FP&A and PE it tracks performance against plan; in M&A it walks reported EBITDA to adjusted EBITDA, the most scrutinized number in the deal. The chart is simple. The defensibility is in the documents behind each bar.

This article is for general information and is not accounting, tax, or investment advice. EBITDA adjustments and valuation depend on your facts and should be prepared or reviewed by a qualified professional.