Fine wine
A market of a few thousand serious participants, priced in cases of twelve, settled in bonded warehouses, and quoted almost entirely gross of cost. It has had a boom and a twenty-four percent correction inside four years and almost nobody outside the trade noticed either.
Under all of this is a finite quantity of dirt. Chateau Lafite is 112 hectares; Romanee-Conti is 1.81 hectares and makes around 450 cases in a good year, for the entire world, forever. Scarcity in this market is not a marketing posture, it is a cadastral fact, and it is why the price of a great Burgundy behaves nothing like the price of a great Bordeaux.
The classifications are three competing answers to the same question, and knowing which one someone is quoting tells you what they are selling.
| Classification | Year | Ranked by | Still honest? |
|---|---|---|---|
| The 1855 Classification | 1855 | The price a chateau fetched on the Bordeaux market, ordered by brokers for the Exposition Universelle at the request of Napoleon III. | As a price ranking, no. As a piece of durable branding, extraordinarily. It has been amended once in 170 years. |
| Saint-Emilion Classification | 1955, revised roughly every decade | Tasting, terroir, reputation and commercial standing, reassessed by committee. | Contested. Cheval Blanc and Ausone withdrew rather than take part in the 2022 revision, which tells you what the participants think it measures. |
| Burgundy's Grand and Premier Cru | Codified 1936 | The parcel of ground itself. Not the producer, not the label, the climat. | Yes, and it is the only one that is really about the dirt. Romanee-Conti is 1.81 hectares and no committee can make more of it. |
| The Liv-ex Classification | 2007 onward, every two years | Average trade price on the exchange over the preceding period. Nothing else. | It is the honest answer to the 1855 question, restated with current prices. Which is also why it is the least romantic. |
Provenance, and the part nobody advertises
Rudy Kurniawan sold forged Burgundy through the top auction houses for years. He was caught partly because he offered Domaine Ponsot Clos Saint-Denis from vintages in the 1940s through 1960s, and Laurent Ponsot knew the domaine had not made that wine until 1982. Ponsot flew to New York and stopped the sale from the floor. Nobody knows how many of his bottles are still in circulation, and the trade's estimates of the fake share of the old and rare market are uncomfortable enough that they are rarely printed.
Fill level
Ullage tells you how the bottle was stored. For an older Bordeaux, into-neck or top-shoulder is what you want; low-shoulder on a 1982 means it has been somewhere warm.
Original wooden case
An intact OWC with matching bottle numbers is worth a real premium, and its absence on a wine that should have one is a question rather than an answer.
Unbroken chain in bond
A case that has never left a bonded warehouse has a paper trail. One that has been in somebody's garage does not, whatever they tell you about the garage.
Labels that are too good
The Kurniawan tell. A sixty-year-old label in perfect condition, on a bottle whose glass and capsule show their age, is the wrong kind of miracle.
Capsule, cork and glass
Branding on the capsule and the cork should match the era, and so should the bottle mould. Forgers get the label right long before they get the glass right.
Rare format, rare vintage
The highest-value combinations carry the highest fake risk, because that is where the money is. A jeroboam of a legendary vintage deserves more scrutiny than the wine inside it will ever get.
The long arc
1855
Bordeaux is ranked by price, and never quite escapes it
Brokers order the Medoc by what each chateau was fetching, for an exhibition. The ranking outlives the prices it was based on by a century and a half, which is the first lesson this market teaches about the difference between a label and a value.
1976
The Judgement of Paris
California beats Bordeaux in a blind tasting judged by French panellists. The immediate effect is on prices in Napa. The lasting effect is on the credibility of blind tasting as an instrument, which cuts both ways.
1982
The vintage that made a critic
Robert Parker backs the 1982 Bordeaux against an establishment that thought it overripe. He is right, and for the next thirty years a numerical score from one American palate moves the price of French farmland.
2008-2011
Lafite, and the Chinese bid
Lafite becomes the status object of a generation of Chinese buyers and runs far ahead of its peers. The chateau etches the character for eight onto the 2008 vintage for that market. The premium does not survive the decade.
2011-2014
The first modern correction
The bid withdraws, the Bordeaux-heavy indices fall hard, and the trade discovers that a market of a few thousand participants can reprice very quickly when one of them leaves.
2020-2022
The boom nobody outside the trade noticed
Burgundy and Champagne run, the Liv-ex 1000 broadens, and fine wine has its best run since the Lafite years. The Fine Wine 100 peaks around the end of 2022.
2023-2025
Down twenty-four percent
Three consecutive negative years: -14.1%, roughly -9.5%, then -2.5%. Average case prices traded fall harder than the index, down 12.3% in 2025 alone, because the mix moved down as well as the prices.
2025-2026
The bottom, probably
Bid:offer ratios climb from historic lows in mid-summer 2025. The Liv-ex 1000 rises every month from August. Chinese imports grow for the first time in nearly a decade, and the trade consensus turns to +2.1% for the year ahead.