Investing in Burgundy and Champagne

On the afternoon of October 13, 2018, Sotheby’s in New York offered two bottles of 1945 Romanée-Conti from the cellar of Robert Drouhin, whose family firm had distributed the domaine’s wines in France for decades. Each was estimated at $22,000 to $32,000. The first sold for $558,000; a few minutes later the second sold for $496,000. Both beat the standing record for any bottle of wine, a three-litre 1945 Mouton Rothschild that had made $310,700 at the same house in 2007.

The 1945 vintage at Domaine de la Romanée-Conti was about 600 bottles in total (the figure Sotheby’s gave in its catalogue and Guinness World Records repeats), made from old vines that were pulled out soon afterwards, so the supply of this wine has been falling since before most of the bidders were born. On March 18, 2026 the same bottle, re-offered at Acker’s La Paulée sale, sold for $812,500.

Two weeks after the 2018 sale, none of that mattered to the people who actually buy Burgundy. Their concern was the letter. Every year, Corney & Barrow in London, the domaine’s exclusive UK agent, writes to a list of customers offering a fixed allocation of the new vintage at a price the domaine sets, on the condition that the wine stays in a UK bonded warehouse and that the merchant gets first refusal if the buyer ever sells.

For the 2023 vintage, released in 2026, that price was £4,485 a bottle for Romanée-Conti. A decade earlier the same letter had shown what it was worth: the 2013 vintage left Corney & Barrow in January 2016 at just over $2,500 a bottle in bond, and by February 2017 Wine-Searcher found it commanding more than four times that. The return was made on the day the letter arrived, by whoever was on the list, and everyone who was not on the list paid the market price and hoped.

That is the shape of both markets in this guide: the producer, not the exchange, decides who gets the asset at the price that makes money. The flagship guide on this hub, Investing in Fine Wine, covers the €30B market as a whole, its century-long record against equities and the platforms; this one stays on the five domaines and four houses whose names carry the prices, the allocation system that rations them, the index history through the 2022 peak and the drawdown, the fraud that was built on Burgundy, and a worked example in dollars.

Burgundy’s price index has beaten every other fine-wine index since Liv-ex began publishing regional series in 2003, and it has also fallen further than any of them since 2022; both facts are true at once, and any reading of the region that keeps only one of them is a sales pitch.

Liv-ex’s own extended report on the region, published in the first quarter of 2020 under the title “Burgundy: after the peak,” put the Burgundy 150 up 445% over the sixteen years to that point, against 202% for the benchmark Liv-ex 100 (Liv-ex via Decanter, January 31, 2020). Our earlier reading of the same report put the Liv-ex 50 of Bordeaux first growths at 235% and the broad Liv-ex 1000 at 248%, two figures we have not been able to re-check against the report itself. Compounded, 445% over sixteen years is roughly 11% a year in sterling, before storage, insurance, commission and tax, none of which the index carries.

The peak in the report’s title was 2018. Burgundy had already doubled between early 2006 and mid-2008, the first wave of a new generation of wealthy buyers moving from Bordeaux to a region with a fraction of the bottles, and it rose again through the 2010s as Bordeaux stalled after its own 2011 bust. Then it stumbled: US tariffs on French wine, Brexit and a run of trade headwinds took the Burgundy 150 down 8.8% in 2019, and the first quarter of 2020 cost it a further 1.5% (Liv-ex, Q1 2020).

What followed was the largest two-year rise in the index’s history. The Burgundy 150 rose 31.0% in 2021, against 19.1% for the Liv-ex 1000, and another 26.7% in 2022, double the 13.1% the broad index managed that year (Liv-ex via Decanter; our tape’s copy of the Liv-ex 1000 went from 424.46 to 480.12 over 2022). Vinetur’s February 2026 review of Liv-ex data puts the rise from the start of 2021 to the peak at about 75% and calls it “the 2021 bull run”; the annual figures only reconcile with that number if the run is read as the whole of 2021 and 2022, which is how we read it here. By September 2022 Burgundy was, on the exchange’s own measure, the most expensive it had ever been.

From that month to August 2025 the Burgundy 150 fell 34% (Vinetur, February 2026); The Drinks Business, reviewing the downturn in January 2026, called it the worst-performing sub-index of the Liv-ex 1000 at −33.7%. The broad market fell less: on our tape the Liv-ex 100 dropped 27.1% and the Liv-ex 1000 28.8% between their autumn 2022 peaks and the same August 2025 low. The wines that had run hardest fell hardest, which is what concentration does in both directions. Even after the fall, the index remained above its 2020 level; the whole of the pandemic-era gain was not given back, only most of it.

Set the long record next to equities and the gloss comes off. The S&P 500 returned about +313% with dividends over the ten years to August 31, 2026, roughly 15.2% a year (S&P Dow Jones Indices data, as used across this hub), with no storage charge, a bid at any second of the trading day, and a 20% top federal rate rather than 28%. Burgundy’s decade beat that only if you owned the right five producers and sold in 2022, and the index says nothing about whether you could have bought them at all.

Since the low, the index has done what floors do: +2.2% between September 2025 and February 2026 (Vinetur), +1.9% over the year to June 2026 (Liv-ex via WineNews). Trade has come back faster than price. In 2026 Burgundy accounted for 25% of everything traded on Liv-ex by value, the highest share ever recorded (Liv-ex via wein.plus), up from a 19.7% record in 2019 and 21.4% in 2021. Buyers and sellers are meeting again, at a lower price.

Burgundy’s entire investment case is the arithmetic of a hierarchy in which the top rung produces about one bottle in a hundred. The Bourgogne wine board (BIVB) divides the region’s output into four tiers: regional appellations are 52% of production, the 44 village appellations 37%, premier cru vineyards 10%, and grand cru vineyards 1%. The 33 grand cru appellations (the conventional count, which treats Chablis Grand Cru as one; about 550 hectares in all, roughly 2% of the region’s vines) run in a strip along the Côte d’Or, and most of the money in this guide concerns perhaps a dozen of them: Romanée-Conti, La Tâche, Richebourg, Musigny, Bonnes-Mares, Chambertin, Clos de Bèze, Clos de la Roche, Montrachet, Corton-Charlemagne.

The second half of the arithmetic is ownership. Bordeaux’s great estates are single properties of dozens of hectares under one owner, which is why a first growth makes hundreds of thousands of bottles and why the Bordeaux 500 is a deep, liquid index (the Investing in Bordeaux guide on this hub covers that market). Burgundy’s vineyards were broken up after the Revolution and divided again by two centuries of French inheritance law, which splits land equally among heirs; the result is that a famous grand cru is not an estate but a field with many owners, each bottling a few rows under their own name.

Clos de Vougeot’s 50.6 hectares are shared by about 80 owners across some 100 parcels (Decanter; Wine-Searcher). What trades, therefore, is not “Musigny” but “Roumier Musigny,” a specific grower’s parcel, and the parcel can be very small indeed.

The bottle counts below are the market. Domaine de la Romanée-Conti’s monopole Romanée-Conti is 1.81 hectares and yields between 5,000 and 6,000 bottles in a normal year; the domaine’s larger monopole, La Tâche, is 6.06 hectares and makes about 20,000 (Sotheby’s producer guide).

Georges Roumier’s holding in Musigny is 0.10 hectares, about 300 bottles a vintage (Decanter; Burgundy-Report), out of an estate of just under 12 hectares whose total output Vins et Millésimes puts at roughly 40,000 bottles (a figure we have not re-verified). Armand Rousseau, the reference for Chambertin, holds 2.55 hectares there and 1.48 in Clos de la Roche and averages 65,000 bottles across everything it makes. For comparison, Krug’s walled Clos du Mesnil in Champagne is 1.84 hectares and roughly 12,000 bottles, and Salon’s 2008, the smallest release in that house’s history, was the equivalent of about 16,000 bottles, sold only in magnums. A single Bordeaux first growth’s grand vin dwarfs all of these together.

Weather then decides how much of even that small number exists in a given year. Successive nights of hard frost in the first week of April 2021 (April 5 to 8) cost the region 30–50% of its crop, with Côte de Beaune whites losing up to 80% (BIVB president François Labet; Jancis Robinson; The Drinks Business); the growers’ confederation said it would be the smallest vintage in memory. A short vintage tightens supply for a decade, because the wine is drunk on a schedule that does not care how much was made. The asset, then, is a specific parcel from a specific grower in a specific year, its count fixed the day the bottles are filled; the only variable you control is whether you are standing in the right place when it is released.

Five domaines carry the Burgundy market the way four houses carry Champagne, and the reason is not only quality but the length of the record: each has been making the same grand crus, from the same rows, under the same family, for long enough that a fifty-year vertical exists to be priced. Domaine de la Romanée-Conti is the reference. It bottles eight grand crus: six reds in the Côte de Nuits, Corton in the Côte de Beaune since the 2009 vintage, and a tiny Montrachet; two of them, Romanée-Conti and La Tâche, are monopoles, vineyards it owns outright.

Its wines are the most traded Burgundy on Liv-ex and the bottles that set auction records; the 1945 at $558,000 in 2018 and $812,500 in 2026, and the Methuselah of 1999 Romanée-Conti that made $275,000 at Christie’s sale of Bill Koch’s cellar in June 2025, are all DRC.

Domaine Leroy is the price ceiling. Lalou Bize-Leroy, once co-director of DRC, built her own biodynamic estate from 1988 on yields far below her neighbours’, and her Musigny became the first wine ever to average more than $50,000 a bottle across all vintages on Wine-Searcher: its global average peaked just short of $54,000 in April 2024 and its US listings averaged $55,880 in mid-2025 (Wine-Searcher, June 2025; Vinetur, July 3, 2025).

Wine-Searcher’s September 2025 list of the world’s 50 most expensive wines had 44 Burgundies on it, nine of them in the top ten. The parcel is a fraction of a hectare, the production is counted in hundreds of bottles, and a large share of what appears on the market has been through several hands. Leroy is where the “trophy” tier of this guide begins: bottles priced like paintings, traded by a handful of people, and largely detached from the index.

Armand Rousseau in Gevrey-Chambertin, Georges Roumier in Chambolle-Musigny and Dujac in Morey-Saint-Denis are the investable blue chips beneath those two, and they are where an outsider’s money is more likely to go. Rousseau’s Chambertin and Clos de Bèze, Roumier’s Bonnes-Mares and Musigny, and Dujac’s Clos de la Roche (the domaine was founded by Jacques Seysses in 1967, with a first release in 1969) are each made in the low thousands of bottles or fewer, sold through allocation, and priced on the secondary market at multiples of their release.

They also show what happens on the way down: over the two years to early 2025, the average DRC wine fell 23.6% and Roumier’s Bonnes-Mares fell 44% (Liv-ex data via Decanter, 2025, as verified for the flagship guide). Small numbers of bottles move prices in both directions, and the thinner the wine, the wider the swing.

Two names belong in the white column, Coche-Dury’s Corton-Charlemagne and Domaine Leflaive’s Montrachet, but white Burgundy carries a risk the reds do not: premature oxidation, a fault that hit vintages from the mid-1990s onwards unpredictably and that no seller will warrant against. It is one reason the investment market is overwhelmingly red.

The producer sets the price at which Burgundy makes money, and the producer decides who pays it; everything that happens on Liv-ex or at Sotheby’s afterwards is a secondary market in bottles that were already sold once, cheaply, to somebody else. The mechanism is the allocation.

A domaine like DRC sells its wine through one importer or agent per country, at an ex-cellar price it sets each year, in fixed quantities that reflect how much the agent bought in previous years. The agent sells to its own customers on the same basis: the same names, the same quantities, year after year, at a price that has almost nothing to do with what the last vintage fetches on the open market. In the UK that agent is Corney & Barrow, which has held the DRC agency for decades; in the US, DRC’s wines have been imported by Wilson Daniels since 1979 (Wilson Daniels; Forbes, March 2023).

The numbers tell you why the list is worth more than the wine. Corney & Barrow’s release of the 2019 Romanée-Conti was £3,600 a bottle in bond (Wine-Searcher, February 2022); the 2023, released in 2026, was £4,485, about 6% above the 2022 release (Decanter). The step from that price to the market is where the money is made. The 2013 vintage left the merchant in January 2016 at just over $2,500 a bottle in bond, and by February 2017 Wine-Searcher found it commanding more than four times that. A buyer on the list therefore paid perhaps a quarter of the market price on release, and the difference, the whole of the return most people imagine when they hear “Burgundy investment,” accrued to whoever had been buying from that merchant for long enough to be offered a bottle.

The conditions attached are equally revealing: the wine must be stored and delivered in the UK, and the buyer is expected to offer the merchant first refusal on any resale, so that the merchant can control who gets the secondary supply and at what price.

Every serious domaine runs some version of this. Allocations come in mixed cases, so a buyer who wants the Romanée-Conti takes the Échézeaux alongside it in a ratio the domaine decides. The step from release to market is what makes the merchants’ lists the most valuable asset in the trade, and it is why they are policed: sell your allocation on the open market the week you receive it and you will not be offered the next vintage. The rule is enforced quietly, without contracts, and it holds because the allocation is worth more than any single year’s flip.

The grey market is what leaks. Bottles that leave the allocation chain, through a restaurant that sells its cellar, a collector who breaks the first-refusal convention, a distributor in a market with a weak agent, arrive on the secondary market at the full price and with a provenance gap where the merchant’s storage record should be. This is the supply an outsider can actually buy, and it is also, not by coincidence, where the counterfeits live. The 2018 Sotheby’s sale pulled money toward that grey market; the rise in Burgundy’s share of Liv-ex trade from 19.7% in 2019 to 25% in 2026 is, in large part, the growth of a secondary market in bottles that were never meant to trade.

The two 1945 bottles at Sotheby’s in October 2018 made roughly twenty times their estimates, and the whole Drouhin consignment made $7.3M with all 100 lots sold, more than five times its high estimate as Decanter and CNBC reported at the time. A price like that does not stay in the room. Within a year Burgundy’s share of Liv-ex trade set a record; within four years the Burgundy 150 was at an all-time high.

The mechanism is worth stating plainly because it recurs in every collectible market on this hub: a public record price for the rarest example re-rates every lesser example, because owners of ordinary bottles from the same domaine mark their own holdings to a new anchor. The 1945 sold for $558,000; the domaine’s releases then on the market, its 2010s vintages and even its lesser grand crus were re-priced in its light, not because anything about them had changed but because the ceiling had moved.

Auction records are the visible top of a market whose real volume is elsewhere. Sotheby’s global wine and spirits sales were $127.5M in 2025, up 12%, and Burgundy’s share of them rose five points, from 34% in 2024, to 39%, with DRC alone 17% of the house’s sales (Sotheby’s 2025 Wine & Spirits Market Report, February 2026; 2024 report); Acker, which sold the 1945 for $812,500 on March 18, 2026, reports more than $200M for 2025 and Zachys $93.9M, both self-reported.

Christie’s sale of the Koch cellar over June 12–14, 2025 made roughly $28.8M with every lot sold. The Hospices de Beaune charity auction made €18.75M ($21.5M) before buyer’s premium on November 16, 2025, its third-highest total behind 2022 and 2023, with 539 lots of wine (428 barrels of red, 111 of white) averaging €33,930 a barrel, up 4.6% on 2024 (BIVB; Decanter), after a weaker €14.4M in 2024.

The trophy top and the index can move in opposite directions, and 2025 proved it: while the Burgundy 150 was finding its floor in August, Koch’s cellar sold out in June, Sotheby’s had its fourth-best year and Hospices de Beaune rose 30% on 2024. The rarest, best-documented bottles behave like trophy art, largely indifferent to the cycle; everything beneath them is the cyclical market you would actually be investing in, and the two should never be confused when someone quotes you a record.

Fine wine’s 2020–2022 rise was a liquidity event, the same tide that carried crypto and cards, and Burgundy was where it ran deepest because there was the least of it to buy. The Liv-ex 100 peaked at 424.35 on September 30, 2022 and the broader Liv-ex 1000 at 484.99 a month later (our tape’s month-end copies of the indices). The Burgundy 150 peaked in September 2022, in a year it still ended up 26.7%. Rates rose, the tide went out everywhere, and wine fell for 35 months, to August 2025; Liv-ex called it the longest and deepest downturn in the 25 years of its data. Before 2022 the Fine Wine 1000 had never fallen more than 13.2% or for more than 13 consecutive months. Both records broke.

The order in which the regions fell is the order in which they had risen. In calendar 2024 the Burgundy 150 lost 14.4% against 11.3% for the Bordeaux 500 and 6.0% for the Italy 100 (Liv-ex, The Fine Wine Market in 2024), and over the full drawdown Burgundy’s −34% and Champagne’s −33.1% were the two deepest falls of any Liv-ex sub-index, four to seven points worse than the broad indices. At the level of individual wines the damage was larger still: the average DRC wine fell 23.6% and Roumier’s Bonnes-Mares 44% over the two years to early 2025 (Liv-ex via Decanter). When the marginal buyer of a 300-bottle wine steps back, there is no bid until the price finds one.

Our own tape shows the shape from the inside. We keep month-end copies of the Liv-ex 100 and Liv-ex 1000 from July 2021 and a monthly CultX global wine index from January 2014, and we build our Fine Wine sub-index on the Liv-ex 100.

On those series the Liv-ex 1000 stood at 424.46 at the end of 2021, 484.99 at its October 2022 peak, 414.58 at the end of 2023, 366.21 at the end of 2024, 345.33 at the August 2025 low and 352.31 on August 31, 2026, a 2.0% recovery from the low in twelve months. The CultX series, a broader basket with more Burgundy and Champagne weight, peaked at 182.86 in January 2023 and read 155.62 on July 1, 2026, still 14.9% below its top. Our Fine Wine sub-index was 103.27 on September 8, 2026, +3.27% over twelve months and +0.66% over thirty days. These are our copies of published series and our own composite, not a market-wide measure; we quote them so the reader can see the same tape we do.

What ended the fall was a change in who was bidding. By December 2025 the value of bids on Liv-ex reached £31M, the highest since April 2023, with UK and Asian bids 135% above their 2024 average (Vinetur, December 2025); the trade’s phrase, on Liv-ex’s data, was “five-year floor” (The Drinks Business, December 2025), not recovery. In Burgundy the trade has come back before the price: a 25% share of Liv-ex trade in 2026 on an index still a third below its peak means a great deal of wine changing hands at the new, lower level. That is what a bottom looks like while it is happening, and it is not what a recovery looks like.

Champagne became an investment market in the 2010s and behaved, in its first cycle, like a smaller and faster copy of Burgundy. Ten years before 2022 the region was roughly 2% of secondary-market trade; by 2021 it was 8.8% of Liv-ex trade by value and in 2022 it reached 13.7%, the biggest increase of any region and enough to make Champagne the third most-traded category on the exchange after Bordeaux and Burgundy (Liv-ex, “The most traded wines on Liv-ex in 2022”). The number of distinct Champagnes changing hands rose 14.8% in 2022 to 381. Buyers priced out of Burgundy found famous houses making their top wines in tens of thousands of bottles, with a vintage system and an established habit of ageing, at a few hundred dollars a bottle rather than a few thousand.

The price response was steep. The Liv-ex Champagne 50 rose 93.9% from March 2020 to its 2022 peak (Liv-ex, as verified for the flagship guide); Cult Wines’ tally in October 2022 had Champagne up 76.6% since January 2021, more than twice the Liv-ex 1000’s 36.1% and ahead of the Burgundy 150’s 68.4% over the same window (via Harpers). For five consecutive months to October 2022 the Champagne 50 was the best-performing sub-index Liv-ex published (Liv-ex, November 2022); then it fell 2.5% in November.

The fall ran for almost three years. By March 2024 the index was 22.9% below its high (Liv-ex, March 2024), it lost 4.2% over calendar 2025, and at its August 2025 low it stood 33.1% below the peak, a fall as deep as Burgundy’s from a shorter and shallower boom (The Drinks Business, May 2026, citing Liv-ex). Liv-ex dated that peak to October 2022 in its 2024 report and to September 2022 in 2026; the charts here keep each source’s month, and the difference is immaterial to the arithmetic. The houses that had run hardest fell hardest: Cristal, Salon and Taittinger each lost more than 30% over the downturn, with Cristal 2012 and Salon 2008 falling furthest (The Drinks Business, January 2026).

Champagne differs from Burgundy in one mechanism that matters for the long run: the house, not the vineyard, controls how much of a vintage exists at any moment. A Burgundy domaine bottles a vintage once and the count is fixed. A Champagne house can hold part of a vintage on its lees for years and release it later as a late-disgorged edition, which is what Dom Pérignon’s P2 and Krug’s library releases are, so mature supply can be added to as well as drunk down.

The region also manages total volume by decree. The Comité Champagne set the 2026 harvest cap at 8,800 kg per hectare, equivalent to about 250M bottles and the fourth cut in a row, below the 9,000 kg allowed in 2025 and the lowest outside 2020, after shipments fell for a third straight year to 266M bottles in 2025 (299M in 2023, 271.4M in 2024) from a record 326M in 2022 (The Drinks Business, July 2026 and January 2026; Comité Champagne). That is a producers’ body managing a glut, and it is the reason Champagne prices have a floor that Burgundy’s do not: the houses would rather cut the harvest than cut the price.

Since the low the Champagne 50 has risen 1.7%, +1.6% in the first months of 2026 and +2.6% over twelve months (The Drinks Business, May 2026; WineNews). The trade describes the move as one from correction to consolidation: the selling has stopped and the buying has not started, which for an outsider is the useful part of the cycle, provided the entry price is judged against release rather than against 2022.

Four houses’ vintage wines carry the Champagne investment market the way five domaines carry Burgundy’s, and they are, in rising order of scarcity, Dom Pérignon, Louis Roederer’s Cristal, Krug, and Salon. Dom Pérignon is the most liquid prestige cuvée in the world: it is made only in declared years, but in those years the volume is very large.

LVMH does not publish the number; a widely repeated trade estimate puts it at about 5M bottles in a declared vintage (Vinovest, 2025; other trade guesses run from 4M upward, and the house disputes all of them), which makes a case of Dom Pérignon 2008 closer to a first growth than to a grand cru in its trading behaviour, with a deep bid and a narrow spread. Its second release, P2, is the same wine disgorged after roughly 12–15 years on lees and sells for about twice the original (Vinovest; as verified for the flagship guide).

Cristal, created for Tsar Alexander II in 1876 and not sold commercially until 1945 (Louis Roederer), is Roederer’s vintage prestige cuvée and the house whose 2012 fell furthest in the 2022–2025 correction. Krug makes a vintage Champagne and two single-vineyard wines; Clos du Mesnil is a walled 1.84-hectare Chardonnay plot in Le Mesnil-sur-Oger that yields about 12,000 bottles in a declared year (Krug; Decanter), and Clos d’Ambonnay is smaller still.

Salon is the scarcest of all: one wine, from one village, in one grape, declared only when the house judges the year right, and held at least ten years before release. Salon has released 45 vintages since its first in 1905, 37 of them in the twentieth century, on average three a decade; its maximum production is about 60,000 bottles and the 2008 was its smallest ever, the equivalent of roughly 16,000 bottles released only in magnums (The Finest Bubble; Wine Enthusiast). The 2012, a vintage the house had first said it would not declare, was launched in September 2022 as its first release in four years (The Drinks Business), at the very top of the market.

The vintage carries most of the price. 2008 is the modern benchmark (Cristal 2008 was the most-traded wine on Liv-ex by value in 2022, ahead of Cristal 2014 and Dom Pérignon 2012), with 2002 and 2012 the other reference years; a declared vintage from a great year at a great house is the unit of account. That makes Champagne easier to analyse than Burgundy, since the universe is a few houses and a handful of years, and easier to overpay for, because the same handful is what every newcomer buys. The lesson of Cristal 2012 and Salon 2008, the two wines that fell furthest, is that the most fashionable bottle at the top of a cycle is the one with the most sellers on the way down.

Burgundy and Champagne discover their prices on three venues, and each gives a different number for the same bottle. The wholesale exchange is Liv-ex, a London members-only platform on which merchants and platforms trade with each other; private investors cannot join and see its prices only through a member. Its Burgundy 150 and Champagne 50 are the indices this guide has quoted, built from the mid-point of bids and offers on the most-traded wines, and its bid-offer spreads run from the mid-single digits on the most liquid names to well over 10% on thin ones, widening in exactly the months you would want to sell.

The retail layer is the merchants, and their exchanges: Berry Bros. & Rudd’s BBX, Farr Vintners, Lay & Wheeler, Justerini & Brooks and Corney & Barrow in London, the importers’ retailers in the US. This is where allocations are issued and where most secondary Burgundy clears, a few percent above Liv-ex for the same in-bond case, with a seller’s commission of about 10% on the merchant exchanges (BBX, as verified for the flagship). The merchant’s storage record is also the provenance document the next buyer will demand, which is why wine that has never left bond commands a premium over identical bottles from a private cellar.

The auction houses are the third venue and the most expensive. Sotheby’s charges a flat 24% buyer’s premium on wine and spirits, the one category it excluded when it raised its general buyer’s premium to 28% from February 13, 2026 (Antiques Trade Gazette; The Value); the seller pays a commission on top, typically around 10% for wine and negotiable on large consignments. Acker and Zachys, the two American specialists, each publish a 25% buyer’s premium (fee schedules, September 2026), so the round trip there is wider still. A bottle that hammers at $10,000 therefore costs the buyer $12,400 and pays the seller about $9,000, a 27% round trip before anyone has stored, insured or shipped anything.

Auction is nonetheless where trophy Burgundy goes, because it is the only venue where the buyer of a single 1945 bottle will be found, and where a documented cellar like Koch’s can be sold in a weekend. For a six-bottle case of a 2015 grand cru it is the wrong door.

The public benchmark for young Burgundy is the Hospices de Beaune, which auctions the charity’s new vintage by the barrel each November; the €33,930 average per lot in November 2025 works out to roughly €118 a bottle at the 288 bottles the Hospices count to a 228-litre barrel (our arithmetic; Decanter), and the year-to-year change in that total is the trade’s first read on the price of the new vintage. Wine-Searcher, which aggregates retail listings worldwide, is the free price check available to anyone; it is a list of asks, not trades, and should be read as a ceiling.

Every return in this guide is quoted from a price index that carries none of the costs of owning the wine, so the honest return is the index return minus a stack of charges you can add up in advance. Storage in a UK bonded warehouse, where investment-grade Burgundy lives because bond defers VAT and duty and creates the provenance record, costs about £14.40 per twelve-bottle case per year plus VAT at London City Bond via Farr Vintners (from June 2025) and £16.57 at Octavian, insurance usually bundled; call it $22 a case-year at $1.30 to the pound.

Selling through a merchant exchange costs about 10%; selling at auction costs the seller a commission and the buyer 24% on top of the hammer, which comes out of what the buyer will bid. Managed platforms charge 2.85–2.95% a year on top of all of that (their schedules are in the playbook below). And the market clears in sterling and sources in euros, so a dollar holder’s return includes an FX position they did not choose.

The tax is the largest single cost for a US holder. Wine is a collectible under the Internal Revenue Code: §1(h)(4)–(5) point to the §408(m)(2) list, which includes “any alcoholic beverage,” and the gain on a bottle held more than a year is taxed at a maximum federal rate of 28% rather than the 20% top rate on shares (IRS Topic 409). The 3.8% net investment income tax under §1411 applies above $200,000 of income for a single filer and $250,000 for a couple, and state tax comes on top. Hold for a year or less and the gain is ordinary income.

The 2025 tax legislation left the collectibles rate untouched, so the 28% applies to the 2025 and 2026 tax years. Two things soften it: 28% is a ceiling, so a lower ordinary bracket pays the lower rate; and basis includes what you paid to acquire the case, so a buyer’s premium or commission on the purchase reduces the eventual gain. Keep every invoice in the provenance file. The Precious Metals guide on this hub walks the same rule through bullion.

The worked example

Take a six-bottle case of a blue-chip Rousseau or Roumier cru from a good recent vintage, bought in bond through a London merchant at $2,000 a bottle, $12,000 all in, and held seven years. Assume it appreciates at 5.3% a year, the rate the Liv-ex 100 has compounded at since 2004 on our tape (322.92 on August 31, 2026 from a base of 100 at the start of 2004), before costs; that is a broad-market rate, and Burgundy’s own record has been both higher and lower over any given seven years. After seven years the case is worth $12,000 × 1.053⁷ = $17,226. Storage at $22.46 a year, the London City Bond rate with VAT converted at $1.30, is $157 over the holding. Sold through a merchant exchange at a 10% commission, the seller receives $15,503.

The taxable gain is the net proceeds less the $12,000 basis, $3,503; we treat storage as a non-deductible carrying cost, which is the conservative reading for an individual after the 2017 suspension of miscellaneous itemised deductions. Federal tax at 28% is $981 and NIIT at 3.8% is $133, a combined $1,114. After storage and tax the holder keeps $14,232, a profit of $2,232 on $12,000 over seven years: 18.6% in total, or about 2.5% a year, from a 5.3% gross. Commission took 1.6 points a year, tax 1.1, storage 0.2.

Sell the same case at auction instead and the arithmetic turns hostile: a buyer willing to pay $17,226 all-in will hammer at about $13,892 once the 24% premium is on top, a 10% seller’s commission leaves $12,503, and after $157 of storage and $160 of tax on the $503 gain the holder is left with a profit of $186, about 0.2% a year. A case bought at the Burgundy 150’s September 2022 peak and marked at the August 2025 low is simpler still: $12,000 became $7,920, a 34% loss before any cost, deductible against other capital gains and against $3,000 a year of ordinary income.

The numbers explain the structure of the trade. An allocation holder who paid a quarter of the market price can absorb a 10% commission, a 28% tax and a 34% drawdown and still be far ahead; a secondary-market buyer paying full price is working for 2.5% a year in the good case. Everything in this guide about merchants, lists and patience follows from that gap.

The largest wine fraud ever prosecuted was a Burgundy fraud, and it succeeded because Burgundy is the one region where the bottles are scarce enough, old enough and expensive enough that nobody in the room had tasted the real thing. Rudy Kurniawan appeared in the Los Angeles collecting scene in the early 2000s, bought heavily, and by 2006 was the market’s largest consignor: two Acker auctions that year, $10.6M in January and a then-record $24.7M in October, sold about $35M of his “legendary” old Burgundy and Bordeaux.

He was caught by the domaines themselves. In 2008 he consigned Domaine Ponsot Clos Saint-Denis from vintages between 1945 and 1971, and Laurent Ponsot, who knew his family had not had access to that appellation until 1982, flew to New York to see the lots withdrawn (Wine Spectator; Decanter’s trial coverage, at which Ponsot testified that the wine “cannot exist”). When the FBI raided Kurniawan’s house in March 2012 it found a production line: empty bottles of great vintages, corks, wax, a re-corking device, recipes for blending cheaper wines into convincing fakes, and some 18,000 printed labels.

The sentence, handed down in Manhattan federal court in August 2014, was ten years in prison, $28.4M in restitution and $20M of forfeiture (US Department of Justice, Southern District of New York); Judge Berman thanked the Burgundian winemakers, Ponsot among them, who had testified. Kurniawan was released in November 2020 and deported to Indonesia (Decanter). By one widely repeated estimate as many as 10,000 of his bottles are still in circulation, and the trade’s guesses at how much old fine wine on the secondary market is fake run from 5% to 20%; nobody has measured it, and whatever the true share, it is concentrated in exactly the bottles this guide is about. The auction house that sold his consignments in 2006 is the house that sold the record 1945 Romanée-Conti in 2026.

The defence is not expertise, which Kurniawan’s victims had in abundance; it is documentation. A grand cru with a continuous in-bond record from the merchant’s allocation to the day you buy it is an asset; the identical bottle with a gap, however plausible the story, is a coin flip you will lose at the point of sale, because the next buyer’s specialist will ask the question you failed to.

Kurniawan was not the only forger working the region. In December 2012 DRC itself alerted the French police to suspect bottles of Romanée-Conti, and the Europol sweep that followed dismantled a counterfeiting ring across ten countries (Wine Spectator). The domaines have hardened the bottles since (DRC, among others, is reported to have adopted tamper-evident coded capsule seals in the 2010s, a measure we have not re-verified), but none of that helps with a 1990 bottle, which is where the money is.

Beyond forgery, the quiet killers are the same as in the flagship guide: a case stored warm for a summer is worthless and looks perfect; a merchant that fails with your wine in its bond, or never bought the wine it invoiced, leaves you a creditor rather than an owner; and cold-called “Burgundy investment” schemes selling obscure labels at three times the market are a pattern the UK regulators have prosecuted for two decades.

An outsider gets into Burgundy through a merchant, into Champagne through a merchant or an exchange, and into either through an auction house only for trophy bottles they intend to keep; that is the ranking, and the fees explain it. The doors, honestly ordered, are these.

A merchant with allocations is the only route to release prices: in London, Corney & Barrow for DRC and the Burgundy specialists (Berry Bros. & Rudd, Farr Vintners, Lay & Wheeler, Justerini & Brooks); in the US, the retailers supplied by the domaines’ importers. The costs are the merchant’s margin, storage in its bond and about 10% commission on its exchange when you sell. The price of admission is time, and the wine you buy while waiting.

A Liv-ex member or merchant exchange is the route to the secondary market near the wholesale price: a merchant will source a listed case at a few percent over the exchange price, in bond, with the provenance record attached. This is where a first case should come from if you are not yet on a list.

An auction house is the route to bottles that do not trade any other way. Sotheby’s 24% buyer’s premium and a seller’s commission of about 10% make it a 27% round trip, which no five-year return in this guide survives; it is the venue for a documented old bottle you will hold for decades, and for selling a large collection with a name attached.

A managed platform buys access for a fee. Cult Wines opens at $10,000 with annual fees from 2.95%; Vinovest at 2.85% falling by tier to 2.25% from $250,000, with a 1.5% early-selling fee and an entry minimum reported at $5,000 since its March 2026 sale to StartEngine; WineCap in the UK takes £5,000 minimum and charges per-bottle storage plus 5% on sales rather than a management fee (published schedules, September 2026). The fee is more than half of wine’s honest gross return, and none of them can put you on Corney & Barrow’s list; the sister guide Investing in Wine Through Platforms and Funds ranks them in detail. Fractional ownership has largely closed to US retail: Vint went accredited-only on January 1, 2024 and announced a wind-down in June 2026.

How to begin

1. Decide the sleeve first. Cap Burgundy and Champagne together at an amount whose 35% fall you can ignore; write the number down.
2. Open an account with one London Burgundy merchant that has its own bonded storage and exchange, and buy a first in-bond case of a liquid name, a Dom Pérignon 2008 or a premier cru from one of the five domaines, at a price checked against Liv-ex through the merchant and against Wine-Searcher yourself.
3. Ask, in writing, to be considered for the merchant’s allocations, and buy a modest amount of what it offers each January, when the Burgundy campaign runs, even when the offer is dull. Allocation history is the asset.
4. Keep everything in bond, in the merchant’s name-tagged account, and keep every invoice, stock record and condition report in one file; that file is what you will sell.
5. Price the position at the Liv-ex bid, not at the retail listing, every quarter, and apply the two price tests in this guide: buy Champagne within 25% of release, sell at twice release; treat a Burgundy with no Liv-ex trade in twelve months as illiquid.
6. Plan the exit before you need it: the merchant exchange for cases, auction only for a single documented old bottle, never a cold caller.
7. Sit down with a tax adviser before the first sale to confirm the 28% collectibles treatment, the NIIT threshold and your basis, and file the invoices with the return.

The readings that would change our view of both regions are specific, and every one is dated so a reader a year from now can see how far the tape has moved. As of September 10, 2026:

• The Burgundy 150 against its August 2025 low. The index fell 34% from September 2022 and had recovered about 2% by early 2026. A new low below the August 2025 level would mean the floor failed; a reading 15% above it would mean the easy recovery is done and the entry window has closed.
• The Champagne 50 against its 2022 peak. It was 33.1% below the peak at the August 2025 low and about 32% below in mid-2026. We do not buy the recovery until it is within 15% of the peak, and we treat a fall back below the 2025 low as evidence that the harvest cap is not enough.
• Corney & Barrow’s DRC release price. The 2023 Romanée-Conti was £4,485 a bottle in bond. A 2024 release below that would say the domaine has accepted the market; a step up would say it has not.
• Burgundy’s share of Liv-ex trade. 25% in 2026, a record. A share that stays above 20% while the index is flat is a bottom being built; a share that falls back toward 2019’s 19.7% is the bid drying up.
• Hospices de Beaune, November 2026. €18.75M in November 2025 and €14.4M in 2024. A total above €18.75M for the 2026 vintage says the trade is paying up for young Burgundy again; below €15M says it is not.
• Champagne shipments. 266M bottles in 2025 against a 2026 harvest cap near 250M-bottle equivalent. A fourth straight fall, below 260M, would test the houses’ willingness to hold price; a return above 271M would remove the glut argument.
• Sotheby’s Burgundy share. 39% of 2025 wine sales. A year below a third would say the trophy buyer has moved on.
• Our own tape. The Liv-ex 100 read 322.92 and the Liv-ex 1000 352.31 on August 31, 2026; our Fine Wine sub-index was 103.27 on September 8, 2026, +3.27% over twelve months. A negative twelve-month print on the sub-index would put the floor back in question.

Burgundy is the best-performing fine-wine region on record and the worst investment most outsiders can make in fine wine, and both halves of that sentence follow from the same fact: almost none of it exists, and what exists is allocated at a price the market never sees. The Burgundy 150’s 445% to 2020 was earned by people on merchants’ lists buying at release; the 34% fall from 2022 was absorbed by people who bought the same bottles at market. Champagne offers the outsider a fairer deal, because the houses make enough of the great cuvées that the secondary market is real, the release price is public and the region will cut its harvest before it lets the price collapse, and because its boom was shallower and newer, there is less legend to overpay for.

Our position, then, is a set of rules rather than a call on the level. Burgundy and Champagne together are never more than half of a wine sleeve sized so that a 35% fall is a rounding error. Burgundy is bought on release, through a merchant relationship built over years, or it is bought in bond on the secondary market only from the five domaines and their second tier, only with a continuous paper trail, and only with the knowledge that the honest return from that price is a few percent a year after commission and the 28% rate.

Champagne is bought within 25% of release and sold at twice it. Auction is for documented old bottles you will keep, not for cases you will sell. And the record price for a 1945 bottle, whatever it is next year, tells you nothing about what the 2019 in your bond is worth.

This guide is written for readers who want to understand how a market works before they decide whether to take part in it, and every figure in it should be checked against the source named beside it before any money moves.