Champagne's 2033 Forecast Rests on Recovering the Price It Just Lost

A market forecast is a claim with a shape. Champagne’s is unusually easy to see, and unusually exposed, because most of it is not a forecast about demand at all.

The category is sized at USD 6.16 billion in 2025, reaching USD 7.13 billion by 2033 at 1.85 per cent a year, on volume growing 0.27 per cent. Strip the currency convention out and the model says something very specific: bottles barely move, and nearly all of the growth is a recovery in the price of a bottle.

The arithmetic the headline hides

Two growth rates that look similar describe two completely different businesses. Volume growth of 0.27 per cent is a line so flat that a single bad winter can move it. So the value has to come from revenue per bottle, and the model’s own numbers say where it has to get to: EUR 24.26 in 2033, from EUR 21.42 in 2025.

That is a compound 1.57 per cent a year for eight years, and 1.57 plus 0.27 is 1.84 — which is the headline 1.85. The model is internally clean. That is precisely what makes it worth interrogating: the arithmetic is a tautology, and the entire forecast is a single assumption about price dressed as a growth rate.

Now look at what EUR 24.26 actually means. Champagne’s revenue per bottle was EUR 23.95 in 2024. The 2033 forecast is 1.29 per cent above the level the appellation was at before the collapse. In other words, the model does not assume the 2025 fall is reversed and quickly forgotten. It assumes eight years to get back to roughly where the category was, plus a rounding error.

That is a more honest forecast than it is usually given credit for. It is also an extremely demanding one, because between here and there sit four things that can each take 1.57 per cent away.

One: the stock has to clear without a discount

The appellation is carrying 1.28 billion bottles — EUR 27.42 billion of wine at current ex-cellar values against EUR 5.7 billion of annual revenue. That is 4.81 years of shipments against a stated target of 4.2: roughly 162.3 million bottles, or EUR 3,476 million of wine, more than the appellation wants to hold.

The forecast requires that overhang to be worked down without discounting. Nothing in the 2025 record supports that. The blended price fell 10.56 per cent in the year the ratio reached 4.81, and discounting is the one mechanism that moves 162 million bottles quickly. If the surplus clears through price rather than through time, the 1.57 per cent required each year becomes a number that has to be earned back first — a year of −5 per cent needs four years of +1.57 to get over, and the forecast window is eight years long.

Two: a 15 per cent tariff on the market that pays the most

The United States takes 26.5 million bottles at EUR 26.00 — 9.96 per cent of the appellation, 12.1 per cent of its revenue, and the second-highest price in the published series. The 15 per cent measure in force from 1 July 2026 is EUR 3.90 a bottle, with a sunset of 31 December 2029 and no sector exemption.

The timing is the problem rather than the size. A price recovery that has to compound for eight years is being asked to happen underneath a regime with a three-year sunset, a threatened 100 per cent measure raised in June 2026, and a live Section 301 review with hearings in July 2026. An importer cannot underwrite a five-year capacity plan against a tariff schedule that expires inside it, so the rational response is to set the plan at the pessimistic end and keep the option to reprice.

Three: two-thirds of the revenue is in someone else’s currency

Sixty-four per cent of Champagne’s revenue is earned outside France and therefore outside the euro. At the EUR/USD 1.08 rate this report uses, ten per cent of the growth case is a currency convention: a ten per cent move in the pair moves the dollar headline ten per cent with no change in a single shipment.

This is not a reason to distrust the model so much as a reminder of what the USD 7.13 billion is. It is a euro number converted at a stated rate. The euro number has to grow; the dollar number can be made to grow.

Four: the ceiling above Champagne got lower, not higher

For two decades Champagne’s pricing rested on being the top of a ladder: Crémant and Prosecco at EUR 5 to 9 below it, and nothing above. That is no longer the shape of the market. A price umbrella that shortens at the top — revenue per bottle down 10.56 per cent — does not lift the tier beneath it; it removes the reason to trade up.

And above Champagne the substitute is not a still wine at the same occasion. It is a spirits serve at EUR 12 to 15. Meanwhile the fastest-growing product class in beverages is closed to the appellation by definition: European labelling rules adopted in February 2026 define alcohol-free, 0.0 per cent and alcohol-reduced wine, and a de-alcoholised wine cannot be Champagne because the method is the product.

What the forecast gets right, and why that is awkward

The volume line is the strong part of this model. Champagne has shipped between roughly 266 and 271 million bottles for a decade, and a quarter of a per cent a year is a description of that fact rather than a prediction about it. The volume forecast will very probably land.

Which creates the trap. A number that lands gets quoted, and a landed volume forecast is easily mistaken for validation of the whole thing. The line with the information in it is EUR 21.42 heading for EUR 24.26, and that line depends on a stock of 162 million surplus bottles clearing without a discount, on a tariff that expires inside the forecast window, and on a premium that is Japanese and American rather than universal — EUR 32.03 and EUR 26.00, against EUR 21.50 in the United Kingdom, EUR 19.87 in Australia and EUR 17.97 at home.

What would falsify it, in order

The inventory ratio. If inventory divided by shipments is not clearly below 4.4 by 2028, the discount is still in the system and the price path fails. This is the single most diagnostic number in the category and it is published.

The top tier. Prestige cuvées fell 17 per cent to 7 million bottles while carrying 15.7 per cent of export value. If that tier stabilises, mix stops working against the model. If it keeps falling, price has to be earned entirely in the base, where the substitutes are.

The US realisation. Watch the EUR per bottle on the American series rather than the bottles. If 26.5 million bottles clear at EUR 26.00 through a 15 per cent measure, the tariff was absorbed somewhere in the chain. If the bottle count holds and the euro value falls, it was absorbed by the producer.

The settlement beyond 2029. A tariff with a three-year sunset cannot support an eight-year forecast. Anything that converts it into a known, permanent rate — in either direction — changes the model more than a year of shipments will.

None of this makes the forecast wrong. It makes it conditional, and the condition is worth stating in plain language: Champagne does not have to sell more wine to be worth USD 7.13 billion in 2033. It has to stop selling it for less — in a year when the most protected appellation in the world sold it for 10.56 per cent less, out of a cellar holding 4.81 years of it.

Companion analysis: Global Champagne Market Report, Size & Forecast 2026–2033.