Best Wines to Invest In, Screened by Real Price Data

The best wines to invest in are the ones that clear four screens at once, and only one of those screens is about how the wine tastes.

Screen one is trade depth. Liv-ex will not rank a wine in its classification unless that wine traded at least 12 times across at least five vintages in a single year, because a label that sold once at a huge price has proved nothing you can exit into. Screen two is a quality signal that survives losing the critic's name: a Williams College working paper on CellarTracker reviews found Robert Parker moved prices mainly through publicity rather than through quality information. Screen three is scarcity fixed by hectares planted rather than by a marketing decision. Screen four is a drink window still open, because a wine past its peak loses the buyers who were going to drink it. Liv-ex's 2025 classification, which ranks by traded price alone, held 332 wines. That is the size of the investable universe.

What makes a wine investment grade?

Sustained trade at a high price. Not a score, not a region, not a story about the winemaker's grandfather.

Liv-ex has been running its classification since 2009 and states the aim plainly: "using price to determine a hierarchy of the leading labels in the secondary market". Its inspiration is the 1855 Classification in Bordeaux, "which ordered the wines from top estates from fifth to first growths using their market price". No terroir committee, no tasting panel, no critic. Traded price, and enough trades to believe it.

The anti-lottery-ticket rule is the part most investment lists ignore. Liv-ex writes that "one vintage of one very rare label, trading just once for a very great price alone is not enough to secure a place in the classification", and that the labels which qualify "are those that show the most sustained market interest as well". For the 2025 edition, reported by Harpers and Vino Joy News, a wine had to have traded on Liv-ex between 1 July 2024 and 30 June 2025, across five or more vintages, at least 12 times. Once a wine qualifies, Liv-ex works out its average trade price per twelve-bottle case by dividing total value traded by the number of nine-litre cases. The 2025 first tier starts at £2,839 a case. The fifth tier runs from £284 to £354.

Here is the same logic as a screen you can apply yourself.

Screen	The question	What a public source will tell you	What it will not
Trade depth	Does it sell, repeatedly, across vintages?	Liv-ex classification and index membership	Depth in your bottle size and condition
Quality signal	Does the score hold up without the byline?	Critic scores, community scores	How the two disagree on your vintage
Scarcity	Is supply fixed by land?	Hectares under vine, declared production	Whether scarcity is already in the price
Drink window	Is there drinking life left?	Vintage reports, critic windows	Where your bottle sits in that window today

The right-hand column is why lists go stale. The four screens are public. The answers move weekly.

Which wines pass the liquidity screen right now?

Start with the wines Liv-ex has already built indices around, because index membership is a liquidity statement rather than a taste judgement.

The Liv-ex Fine Wine 50 "tracks the daily price movements of the Bordeaux First Growths", covering the ten most recent vintages of Lafite Rothschild, Margaux, Mouton Rothschild, Haut-Brion and Latour. The California 50 covers the ten most recent physical vintages of "the five most actively traded Californian wines: Screaming Eagle, Opus One, Dominus, Harlan Estate and Ridge Monte Bello". The Port 50 does the same for Dow, Fonseca, Graham, Taylor and Warre, according to Liv-ex. The Champagne 50 tracks "the 16 most actively traded champagnes", the Burgundy 150 covers 15 white and red Burgundies including six Domaine de la Romanée-Conti labels, the Italy 100 covers five Super Tuscans and five other leading Italian producers, and the Rest of the World 60 covers six wines from Spain, Chile, the USA and Australia. Above all of them sits the Fine Wine 100, which Liv-ex calls "the industry leading benchmark", representing "100 of the most sought-after fine wines on the secondary market".

Liv-ex's Power 100 answers a different question: which brands are strongest across price and trade at once. WineNews sets out the four measures behind it: year-on-year price performance, taken from a case's market price on 1 October 2024 against its price on 30 September 2025; trading performance on the exchange by value and by volume; the number of commercial wines and vintages a brand offers; and the brand's average price. WineNews lists the 2025 top ten as Cheval Blanc, Tenuta San Guido, Domaine Leflaive, Joseph Drouhin, Rayas and Domaine des Tours, Krug, Haut-Brion, Château d'Yquem, Opus One and Mouton Rothschild.

Note what that list is not. It is not a ranking of the most expensive wines, and it is not a ranking of the best-scoring ones. Because the Power 100 blends price move, trade volume, brand breadth and average price, Château d'Yquem and Krug rank on the combination rather than on price alone. At the price end, Vino Joy News reports Romanée-Conti heading the 2025 classification at an average trade price of £172,461 a case, with Pétrus and Screaming Eagle also in the first tier.

Why is a critic score a weak buy signal on its own?

Because a large part of what a famous score does to price has nothing to do with the wine.

Albright, Pedroni and Sheppard built a panel of CellarTracker reviews covering "355 distinct wines on a quarterly basis from 2004 through 2017", paired it with auction prices for the same wines, and added the reviews of three leading critics. Their reduced-form results found "Robert Parker is unique in his influence on prices when compared alongside his taste-making peers Jancis Robinson and Antonio Galloni". They then separated the quality information in a review from the publicity attached to the reviewer, and reported evidence that "Robert Parker influences prices primarily through the publicity channel, rather than the quality information channel".

Read that as an investor rather than as a drinker. If a score moves price through fame, the move can unwind when the fame moves on, and you are holding an asset repriced by a byline rather than by the liquid.

The market has always behaved as though scores were structural. wine rated above 90 points usually sells well, while wines in the 85 to 89 band, rated "good to very good", are "often ignored by consumers", and some retailers refuse to order anything below 85 points. That is a cliff edge at a number, not a smooth quality curve.

Community scores are the correction, and they are not perfect either. Kopsacheilis and colleagues, in a paper for the American Association of Wine Economists, collected more than 200,000 Vivino reviews of red Bordeaux from vintages 2001 to 2017 and compared them with professional critics. Crowd ratings correlated most closely with Jeff Leve's scores at a Spearman's rho of 0.58, and the authors note that while the overlap in tastes between expert and non-expert opinion is significant, "the relation is far from perfect". They also flag why the two diverge in a predictable direction: experts' scores "capture the wine's potential rather than immediate value". The same paper cites the Oczkowski and Doucouliagos meta-analysis putting the price-to-quality correlation at 0.30, which is why paying more is a poor proxy for buying better.

That is the case for weighting both. Our score is 40% critic and 60% CellarTracker, so the professional read of a wine's ceiling counts, and the accumulated verdict of people who actually pulled the cork counts for more. The site had passed 288,000 registered users and nearly 3.8 million reviews by January 2014, a deeper sample than any single critic will ever publish. Every wine page carries that score next to the liquidity screen, so you can see a wine that scores well and never trades, which is the most common trap on any published list of investment wines.

How scarce does a wine have to be?

Scarce enough that no commercial decision can undo it. The test is hectares, not intent.

Romanée-Conti is the extreme case. The appellation, created in 1936, had 1.88 hectares in production as of 2008, and in that year produced 26 hectolitres, just under 3,500 bottles. The five-year average across 2003 to 2007 was 42 hectolitres, around 5,600 bottles. There is no version of the future in which that number doubles. You can walk the parcel boundaries on the Vosne-Romanée producer atlas and see the constraint as land rather than as a claim.

The estate sits on 48.21 acres, 19.51 hectares, and annual production "ranges from 400 to 750 cases (between 5,000 and 9,000 750ml bottles)". Same order of scarcity, reached a different way: through selection rather than through a tiny appellation.

Scarcity on its own is worth nothing. A rare wine nobody trades is an illiquid wine, which is why Liv-ex pairs its price ranking with the five-vintage, twelve-trade floor. Scarcity plus depth is the combination that holds a price. Scarcity alone is a bottle you cannot sell without finding the one buyer who wants it this month.

Does the region matter more than the label?

It has been mattering more, and 2025 was the year the shift showed up in every published measure at once.

Liv-ex's 2025 classification counted 106 Bordeaux wines and 67 from Burgundy, with Italy up to 86, according to Harpers and Vino Joy News. The Power 100 tilted further: Burgundy took 29 places, Bordeaux 27, Italy 20 and Champagne nine. Bordeaux still supplies the largest single block of classified labels while losing the brand-strength race.

The trade data says the same thing. Liv-ex's November market report records Italy at "19.4% of Liv-ex trade in October", its highest share since August 2020, and notes that the Bordeaux 500 index "saw its first increase since March 2023". That is a two-and-a-half year decline ending, not a recovery you can bank. The same report tracks UK buyers falling from 74% of traded volume in 2005 to around 30% now, which is the internationalisation of the buyer base showing up as regional rotation.

Prices overall went nowhere pleasant. Knight Frank's 2026 Luxury Investment Index reports the Liv-ex Fine Wine 100 down 2.5% in 2025, with total losses of almost 25% since the 2022 peak, against the whole luxury index down 0.4%. Knight Frank's read on where value now sits: "Many top Tuscan wines boast quality scores on par with Bordeaux and Burgundy vintages, yet trade at around half the price." Liquidity came back before prices did. Liv-ex's December report puts bids on the exchange at £31m, the highest since April 2023, on an exchange Liv-ex says has over 500 members across 42 countries.

The honest summary: region has been the bigger driver of outcome than label selection within a region, and the market index is where you watch that rather than in any annual list.

When does the drink window start costing you money?

The moment the wine stops having drinkers ahead of it. Price follows the window, and the window is the part of the screen that changes without anyone announcing it.

This is the mechanism behind the critic-versus-crowd gap. Kopsacheilis and colleagues point out that expert scores describe potential, and that wines judged early "would feel astringent and potentially unpleasant if judged for their hedonic value prematurely". A critic score is a forecast of a wine at maturity. A community score is a running report on what the wine is doing now. The moment those two diverge on a specific vintage, you are looking at either a wine that has not arrived yet or a wine that has gone past.

For an investor the second case is the expensive one. A wine still climbing has buyers who intend to hold and buyers who intend to drink. A wine past its window has only the second group, shrinking.

If the bottles are already in your cellar, the drink-now list answers the same question from the other end: which of your wines are inside their window today.

What the screen will not price for you

Costs, and they decide whether a correct call still makes you money.

A wine that clears all four screens can still lose against a tracker fund once you have paid the buyer's premium, the VAT on that premium, the shipping, the storage and the selling commission at the far end. Run the entry number through the landed-cost calculator before you decide a price looks cheap, and read the wine investment risks guide for the round-trip arithmetic and the illiquidity that no index level shows. The fine wine investment guide covers the tax treatment and the holding costs in full.

The screens tell you which wines are worth owning. The cost stack tells you at what price. Confusing the two is how people end up owning excellent wine at a bad entry.

Watch the screen, not the list

A list of the best wines to invest in is out of date the week it is published. A screen is not.

Put the wines you care about on a watchlist and you get the 40/60 score on every one of them, moving as new critic reviews and new community notes land, sitting next to the trade-depth numbers that tell you whether you could sell if you wanted to. When a wine's score holds while its price falls, you hear about it. When the community half of the score starts sliding away from the critic half, which is usually the first sign a wine has passed its window, you hear about that too.

That is the whole trade. Most buyers screen on scores and discover liquidity at the point of sale. Watch both from the start and the exit is priced before you buy.