Investing in Wine Through Platforms and Funds

At the end of November 2025 the offices of Oenofuture Limited, the investment arm of London’s Oeno Group, quietly closed. The group had a wine bar at the Royal Exchange, a trade desk that promised to place clients’ bottles with restaurants, and, since June 18, 2025, a regulated fund in Portugal with a €50,000 minimum and a €20M target. The company stopped trading just before Christmas.

In January 2026 City of London Trading Standards put out a statement for the roughly 2,600 people who had paid Oenofuture money: about 80% of the client wine was believed to be held by the company itself, with “perhaps” 20% in individual customer accounts that the customers would be able to reach. A provisional liquidator was appointed in February, the High Court made a winding-up order on March 4, and the judge warned that the process would be “very lengthy.”

Nothing about the wine had changed. What had changed was the answer to a question most clients had never asked: whose name is on the warehouse account? For the fifth whose cases were booked to them personally, the failure of the firm was an inconvenience. For the rest, it was the difference between owning wine and being an unsecured creditor of a company that sold it.

That question, and its cousin, what a platform charges you every year for standing between you and the wine, is what this guide is about. The hub’s flagship, Investing in Fine Wine, covers the asset: the regions, the vintages, the long-run return, the 2022–25 drawdown. This one stays on the intermediaries: how the trade prices a case, what it costs to store one, what the platforms and funds charge and what happened to the ones that failed, what the law says when a platform goes under, and a worked ten-year comparison in dollars between doing it yourself and paying someone to do it for you.

Fine wine’s gross return is a mid-single-digit number and a managed platform’s fee is a low-single-digit number, so the fee takes a large fraction of the return before anything else is considered. That relationship, not the quality of anyone’s stock picking, is the first thing to understand about this market.

The gross return is well documented. Dimson, Rousseau and Spaenjers, in “The Price of Wine” (Journal of Financial Economics, 2015), put five Bordeaux First Growths at 4.1% a year in real terms from 1900 to 2012, net of storage and insurance, against 5.2% for equities. On our own tape the Liv-ex Fine Wine 100 closed August 31, 2026 at 322.92 on a base of 100 at the start of 2004, roughly 5.3% a year compounded over 22.7 years before any costs, and that window includes the 2004–2011 China boom that nobody expects to repeat. The honest number to plan around is 5%, give or take.

The fee is also documented, because the platforms publish it. Cult Wines’ entry plan charges 2.95% a year, Vinovest’s 2.85%, and both tier down for accounts in the hundreds of thousands: Cult Wines to 2.5% from $150,000, Vinovest to 2.25% from $250,000. The fee is charged on the whole portfolio every year, whether the wine rose or fell, and it includes storage and insurance, which would otherwise cost about a fifth of a percent. A platform investor therefore compounds at roughly 2.3% a year in a 5.3% market, before tax.

Compounding turns a modest-sounding fee into most of the outcome. $100 growing at 5.3% for ten years becomes about $168. Take 1% a year out for storage, insurance and a selling commission spread across the decade and it becomes about $152. Take 2.85% out and it becomes about $127. Take 4%, which is what fund structures and the dearer brokers have charged once spreads are counted, and it becomes about $114. Over ten years the gap between the self-directed route and the platform route is larger than the platform route’s entire gain.

A platform can earn its fee in three ways: by buying at better prices than you would, by picking wines that beat the index by three points a year, or by getting you an allocation you could not get. The short answer of the sections that follow is that buying well is real but small, selection alpha is asserted rather than demonstrated, and allocation is the one thing worth paying for.

Every price a platform quotes you descends from one place, the London International Vintners Exchange, and the gap between that price and your invoice is the intermediary’s margin. Knowing how the exchange works is what lets you measure the gap.

Liv-ex, founded in London in 2000, is a members-only exchange for the wine trade: more than 620 members in 47 countries in 2026, whose activity it estimates at about 95% of global fine-wine turnover. Members place bids and offers on specific wines in specific formats; when they meet, the exchange matches the trade, stands behind it and settles it.

Its Standard In Bond contract, which makes a case of Lafite 2005 fungible with any other, requires the wine to be in bond and in good condition, with fill at the base of the neck or better, delivered to a Liv-ex warehouse within two weeks. In early 2026 the order book held about £155M of live bids and offers across some 20,000 products. The Liv-ex Mid Price, the midpoint between the highest live bid and the lowest live offer, is the number every index in this market is built on.

The exchange charges its members, not you: a seller pays 2% to 3% by membership tier (2% gold, 2.5% silver, 3% bronze) plus a fixed settlement fee. A retail investor cannot join.

That is the structural fact of this market: you never touch the wholesale price. You buy from a merchant, broker or platform that can, that pays the 2–3% and the logistics, and that resells to you at a retail price higher again. Merchant margins are not published; the only number you can verify is the Mid Price, through a merchant who shows it or through Wine-Searcher’s aggregated retail prices, and the difference between that and your invoice is what the introduction cost.

Two intermediaries shorten the chain. Bordeaux Index’s LiveTrade makes firm two-way prices, one at which the merchant will buy and one at which it will sell, on around 1,000 of the most actively traded wines, with no joining or membership fee and its margin built into the quoted prices rather than charged separately, so the spread between its bid and its offer is visible if you look. Berry Bros. & Rudd’s BBX lets private customers with wine in BBR’s bond list it to other customers; the seller pays a flat 10% and the buyer nothing on top. Neither gets you to the exchange, but both show you a bid, which a managed platform’s statement never does.

The managed platforms sit at the end of the chain. They buy through the trade, then allocate cases to client portfolios at a price that embeds their sourcing cost, with the management fee on top. When a platform shows you a “valuation,” it is at the Mid Price or a merchant-list equivalent, not at the bid, and the distance between a mid valuation and what you would actually clear is the first thing a fund’s auditors argue about, as the Nobles Crus history shows.

A case of investment-grade wine in the UK lives in an HMRC-approved bonded warehouse, with duty and VAT suspended until it leaves, and that arrangement is what makes wine cheap to hold and provenance provable. The economics are small and precise.

UK duty on still wine is charged per litre of pure alcohol: £30.64 from February 1, 2026 for wine between 8.5% and 22% ABV, uprated 3.66% with RPI from the February 2025 rate of £29.54 (Château Bauduc; HMRC), so a 75cl bottle at 12.5% carries £2.87 and at 13.5% £3.10. VAT at 20% is then charged on the full price including duty. Neither is due while the wine stays in bond, and a case moving from one bonded account to another, which is how the whole secondary market trades, never triggers them.

On a £10,000 case the duty is trivial but the VAT is £2,000, which is why investment wine is bought, held and sold in bond and why a “duty paid” case is worth less to the next buyer: the tax is sunk and the custody chain has a gap.

Storage is priced per case per year, with insurance at full replacement value bundled at almost every serious bond. Farr Vintners’ tariff at London City Bond is £14.40 per 12-bottle case a year from June 2025, excluding VAT; Marlo Wine charges £13.50 with insurance included; Private Cellar’s 2025–26 rate is £15.50; Octavian, the Wiltshire bond the top of the trade prefers, charges £16.57 above a six-case minimum.

With VAT on the service a case costs about £17–20 a year to keep, which at an assumed $1.30 to the pound is $22–26: on a $12,500 case, 0.2% a year. Spot was about $1.35 on September 9, 2026; this guide, the flagship and the sister guides convert at $1.30 for comparability, which understates sterling costs in dollars by about 4%, and every sterling figure in the charts and worked example below uses that rate.

Cult Wines stores client wine at Coterie Vaults, the Suffolk bond of its exclusive storage partner Coterie Holdings under an agreement announced in June 2024, and says it books specific cases to individual customers; Vinovest stores through its Vine International arm and issues ownership certificates.

The paperwork changed in 2025 in a way that helps foreign owners. Until March 3, 2025 the Warehousekeepers and Owners of Warehoused Goods Regulations required certain owners of goods in bond to register with HMRC and non-UK businesses to appoint a duty representative; owner registration was repealed on that date, and the regime since then covers only the warehouses, which must still record who owns what.

A bonded account in a US investor’s own name is therefore straightforward: the warehouse records you as owner, gives each case a rotation number, its own identifier in the warehouse’s records, and sends a statement. That statement is the single document separating owning wine from owning a claim on a company, and section eleven explains why. Taking wine out of bond loses the replacement-value cover and the in-bond premium on resale and triggers the duty and VAT; almost no investment case should leave bond until it is drunk.

Four platforms and one merchant scheme account for most of the money a retail investor can put into managed wine in 2026, and their fee schedules are the clearest way to compare them. Two operators that would have been on this list in 2024, Vint and Oenofuture, no longer exist as going concerns and are treated where they failed: Vint in the next section, Oenofuture in the sections on funds and on title.

Cult Wines

Cult Wines, London-based, founded in 2007 and by its own account the largest managed platform, publishes four US tiers: Cru Classe from $10,000 at 2.95% a year, Premier Cru from $35,000 at 2.75%, Grand Cru from $150,000 at 2.5%, and Cult Cru from $700,000 at a rate we could not confirm. The fee covers storage at Coterie Vaults, insurance, authentication and condition checks; there is no performance fee, trading fee or published exit charge, and its CultX marketplace lets clients trade among themselves for 2.5% plus VAT a side (£10 minimum).

The company’s health matters to anyone whose wine it holds. Its 2024 accounts, reported by The Drinks Business in April 2026, show a £5M pre-tax loss, revenue down 43% to £33.5M, net liabilities of £21.6M (from £17.1M) and headcount cut from about 65 to 35, and the auditors’ report carried a “material uncertainty” paragraph on the group’s ability to continue as a going concern without further funding.

Chief executive Tom Gearing told the Financial Times the paragraph was “fairly boilerplate language” and said the first quarter of 2026 was the strongest on record, up 97% on the quarter before. The company stated about £250M under management for clients in 83 countries when it announced the Coterie agreement in June 2024 and has since cited more than £265M; both figures are company-stated.

Vinovest

Vinovest, the US-native platform, was acquired by the crowdfunding platform StartEngine in a deal announced on March 24, 2026. Its pricing page in September 2026 shows four tiers by account size: Starter (under $10,000) at 2.85% a year, Plus ($10,000 to $49,999) at 2.70%, Premier ($50,000 to $249,999) at 2.50% and Grand Cru ($250,000 and up) at 2.25%, prorated and charged on invested capital, with storage, insurance and authentication included.

The entry minimum, $1,000 for years, is reported by 2026 reviews to have risen to $5,000 after the acquisition, while the pricing page itself still labels the Starter tier from $0, so confirm the minimum on the day. Listing wine for sale before Vinovest’s target maturity costs a 1.5% early-selling fee (help pages, 2026; some reviews cite 3% inside three years). Vinovest also runs a self-directed marketplace, which charges 2.5% on purchase (including three months of storage), 1% on sale and 1.5% a year for storage, about $375 a year on $25,000 for what a bond charges $45. Vinovest issues ownership certificates and says that if it failed the warehouse would contact owners directly.

WineCap

WineCap, London-based, charges no management fee. The minimum is £5,000; the client pays about £15 per case a year for storage and 5% on sales, and the company earns on the spread between its buying price and yours, which is the merchant model with an advisory layer. Over ten years the all-in cost is well under 1% a year plus the exit commission, so it prices as a broker in the comparison below.

WineFi

WineFi, a London startup that closed a £1.5M seed round in April 2025, led by Coterie Holdings (Cult Wines’ storage partner) and including a Crowdcube crowdfund, sells co-investment in “syndicates” from £3,000. The charge is 12.5% up front, presented as 2.5% a year over a five-year hold and covering sourcing, brokerage, insurance and five years of storage; after that, storage is taken at cost from sale proceeds. The structure is designed to sit outside FCA authorisation, and the company says so.

Berry Bros. & Rudd Cellar Plan

The oldest route is a merchant standing order. BBR’s Cellar Plan takes a minimum of £250 a month, which an adviser spends into your own bonded account at the firm; you own specific cases, pay BBR’s storage, and sell through BBX at 10% or take delivery. There is no management fee. What you pay is BBR’s retail margin, the price of a 300-year-old allocation book, worth paying for a First Growth en primeur allocation (wine bought while still in barrel and delivered about two years later) you could not otherwise touch and not for a case of Sassicaia you could buy anywhere.

The chart flatters the platforms in one respect: it shows the fee and not the purchase spread. Every route buys at a retail price above the exchange, so the spread is common to all, but a platform that buys en bloc and allocates cases to client portfolios at its own transfer price has a second margin the schedule does not itemise. Ask, in writing, whether cases are allocated at cost or at a marked-up transfer price. The answer is rarely volunteered.

Fractional ownership, in which a platform buys a collection, wraps it in a company and sells shares, was the 2021 answer to the minimum-investment problem, and by 2026 the largest US example had closed. The mechanism matters because it will be tried again.

Vint, founded in Richmond, Virginia, received SEC qualification for its first Regulation A offering in the spring of 2021 and launched a Champagne collection on July 8 of that year. Regulation A lets a company sell securities to non-accredited investors once the SEC has qualified the offering circular; Vint used it to securitise more than 30 collections of wine and spirits, sold shares from $25, and in 2022 reported a 21.7% internal rate of return (IRR) on the Champagne collection, a self-reported figure covering a single early exit at the top of the 2022 market. It raised about $6.9M of venture funding across three rounds (Tracxn), including a $1.7M pre-seed led by Fintech Ventures in November 2021.

The retail door closed on January 1, 2024, when Vint went accredited-only. In June 2026, still loss-making after five years, with a 2025 net loss of about $890K (from $85K in 2024) and a “substantial doubt” going-concern paragraph from its auditor in its SEC filings, it told customers it was winding down, engaging G2 Capital to sell the remaining collections and SimpleClosure to run the closure (Richmond BizSense, June 22, 2026). Shareholders in each collection receive what its liquidation fetches, less costs, on the liquidator’s timetable, in a market where the Liv-ex 100 sat about 25% below its 2022 peak on our tape.

Three lessons carry beyond Vint. The wrapper made the investment a security, which brought SEC disclosure and audited financials, the only reason the going-concern warning was public before the shutdown; no managed platform gives you that. The wrapper also made exit dependent on the platform: shares in a Reg A wine LLC have no market once the sponsor’s bulletin board is gone. And the sourcing was at the platform’s price, which the shareholder could not check against the exchange. Any fractional platform that survives or is launched is venture-stage counterparty risk stacked on wine risk, with the collection’s liquidation value as the floor.

Three funds defined the wine-fund era, from 2003 to about 2013. Two of them, The Wine Investment Fund and Nobles Crus, ended in a regulator’s hands for structural reasons that a wine fund launched in 2026 would face unchanged.

The Wine Investment Fund

The Wine Investment Fund was set up in 2003 by Andrew della Casa, Rodney Birrell, Chris Smith and William Grey, domiciled in Bermuda and advised from London, targeting 15% a year from physical Bordeaux. Its first five-year tranche closed in August 2008 and paid investors 108.6%, a genuine doubling delivered into the financial crisis and the best advertisement the wine-fund industry ever had; its website was still claiming in 2022 that net asset value had risen 139.37% from March 2004 to August 2022.

In December 2022 an investor’s complaint in Bermuda alleged the group was insolvent by at least £7.8M. The Bermuda Monetary Authority applied to the Supreme Court on January 31, 2023, provisional liquidators from Teneo were appointed on February 9, and the court ordered the fund wound up on March 3, 2023; the regulator’s notice recorded a loan to a related party, Lilliput Holdings Limited, that the offering documents did not permit and that had not been repaid in full. We found no published account of what investors have recovered since. Nothing in a wine fund’s structure stops its directors from lending the money out.

Nobles Crus

Nobles Crus, run by Elite Advisers in Luxembourg, was the largest European wine fund, with net assets of €114.4M on September 30, 2012, and a record too smooth to be true: it reported a gain every month from early 2011 through 2012, a period in which the Liv-ex Fine Wine 100 fell 23%. In October 2012 the Financial Times asked Liv-ex to value 50 of the fund’s wines. Elite valued each wine at the average of two auction-house prices, without deducting commission, and two merchant list prices; Liv-ex, valuing at the midpoint of its own live bids and offers, came out 36.7% lower (Decanter, October 2012). Elite commissioned Ernst & Young to review the method.

Then the redemptions came. On May 27, 2013 the CSSF suspended the fund after it could not pay institutional investors who wanted out, many of them unit trusts that an EU rule change would bar from holding wine from December 31, 2013 (Decanter, June 2013); the managers told investors it “does not have the necessary liquidity to honour these requests in the very short term.” A Luxembourg court ordered reimbursement on February 7, 2014 with a stay to April 30, a Chinese investor put in about €37M in kind at the end of that March, and in 2016 the fund provisioned nearly all of its €13.7M of en primeur purchases as unlikely to be delivered (Paperjam, 2014–2016), amid questions about counterfeit bottles in its stock.

Vintage Wine Fund

The Vintage Wine Fund was the third of the trio: launched in 2003 by OWC Asset Management, a London firm with FSA approval whose directors included Bordeaux Index founder Gary Boom and Andrew Davison, through a Cayman-domiciled vehicle with a €50M initial target and a €100,000 minimum (Harpers and Decanter, 2002–03; The Globe and Mail, 2012). Our understanding, which we could not confirm from a published source, is that it was wound down around 2013 after the 2011–13 Bordeaux slump; treat that as held knowledge rather than record.

A smaller UK vehicle shows a third ending, in which the wine was never there. The Wine Enterprise Investment Scheme Ltd raised about £4.24M from investors between 2012 and 2016 as an Enterprise Investment Scheme fund trading Bordeaux; its two directors ran it as a Ponzi scheme. It went into voluntary liquidation in 2020 with a declaration of solvency claiming more than £4.5M in cash; the liquidators found £6.50 in the bank and could not locate the assets.

On March 27, 2026 the High Court found serious audit failings by the company’s former auditor, Crowe UK, but held the directors’ “concerted fabrications” to be overwhelmingly the cause and awarded just over £101,000 against losses claimed at up to £8M, an outcome The Drinks Business called “Pyrrhic” (The Wine Enterprise Investment Scheme Ltd v Crowe UK LLP [2026] EWHC 692 (Ch)).

The structural problem is the mismatch between an open-ended fund and an asset that trades by the case, slowly, at a bid usually well below the mark. A fund promising monthly liquidity at NAV is promising what the market cannot deliver in a downturn, so it will gate, meaning suspend redemptions (Nobles Crus), mark generously to avoid the loss that would trigger redemptions (Nobles Crus again), or do something with the cash the offering document did not allow (TWIF).

The 2025 successor makes the point again: Oeno Group’s regulated fund, a Portuguese collective investment undertaking under the CMVM, launched on June 18, 2025, five months before the group’s UK trading company closed its doors, and regulation of the vehicle did nothing for the clients whose wine sat in the unregulated company.

The long-run record of the asset is positive and mid-single-digit, the record of the last five years is negative, and platform marketing shows you the first with a start date chosen to flatter it. All three statements can be checked.

Our tape holds two series that matter here. The Liv-ex Fine Wine 100 stood at 346.57 on July 31, 2021, peaked at 424.35 on September 30, 2022, troughed at 309.27 on August 31, 2025 and closed at 322.92 on August 31, 2026: down 6.8% over five years and 24% below the peak. The Liv-ex 1000 went from 380.88 to 352.31, down 7.5%. A client who joined a platform in the summer of 2021, the high point of platform marketing and of the “+272% since 2004” headline, has since paid five years of fees at 2.85% on a portfolio that fell, which puts the account roughly 20% underwater before tax.

The second series is a platform’s own index. We store the CultX Global index, published by Cult Wines’ trading arm, from its base of 100 on January 1, 2014; it read 155.62 on July 1, 2026, a compound rate of about 3.6% a year over twelve and a half years on the platform’s own choice of constituents, before the platform’s fee. Subtract 2.95% and a Cru Classe client tracking that index would have compounded at well under 1% a year before tax. We present it as our copy of a vendor index, not a market-wide figure; the point is that even on the platform’s own yardstick the fee consumed most of the return.

Platforms present three kinds of number, and each needs a label. The first is an index return over a long window, usually the Liv-ex 100 or 1000 from 2004: real but gross, survivorship-weighted toward the wines that kept trading, and not what any client portfolio did. The second is the platform’s own composite of client portfolios, self-reported, unaudited against the exchange and marked at mid; Cult Wines publishes a performance page of this kind and we could not verify any figure on it. The third is a single collection’s realised IRR, such as Vint’s 21.7% on its 2021 Champagne collection, true of one exit and silent on the rest. When a deck says “wine returned X% a year,” ask which of the three, over which window, gross or net of the fee you would pay.

Our Fine Wine sub-index, built on the Liv-ex 100, stood at 103.27 on September 8, 2026, up 3.27% over twelve months and 0.66% over thirty days: a floor with a slight upward lean, the same reading the flagship gives, and the environment in which the fee arithmetic below is run.

Two investors put $25,000 into the same two cases on the same day, one through a merchant into a bonded account in their own name, the other through a managed platform, and ten years later the merchant client has about $3,800 more after tax. The dollar-by-dollar version is in the next section; the assumptions and the ranking are here.

The assumptions are the same for both: 5.3% a year gross, our tape’s Liv-ex 100 rate since 2004, generous for two cases because nobody can buy the index; $1.30 to the pound; the same retail purchase price, so the buying spread cancels, and a platform allocating at a marked-up transfer price would do worse than shown; nothing leaves bond; both sell at the end of year ten and pay the 28% federal collectibles rate, with no state tax or net investment income tax, which would reduce both equally.

The self-directed investor pays London City Bond’s £14.40 a case plus VAT, about $45 a year for two cases, and sells at BBX’s 10%. The platform investor pays 2.85% of start-of-year value each year with no exit fee, which matches Cult Wines’ terms and Vinovest’s once its wines are held to target maturity; at 2.95% the result is about $220 worse.

Read the chart as a ranking of cost structures rather than of companies, because the purchase spread is missing and it is not the same everywhere. The broker model edges the self-directed route only because a 5% exit is cheaper than BBX’s 10%; a merchant selling your cases into Liv-ex at cost plus a 2–3% member commission would beat both. The up-front syndicate lands close to self-directed because 12.5% taken once is cheaper over ten years than 2.85% taken ten times, though it is worse over five.

The managed platform is last by a wide margin, and the gap, about $3,800 against self-directed and $5,300 against the cheapest broker model, is the fee: $7,964 collected over ten years, more than the $3,559 of federal tax the self-directed investor paid on a larger gain.

Nothing in that arithmetic depends on the market going up. If the wine compounds at 2% instead of 5.3%, the self-directed investor finishes slightly ahead after tax and the platform investor finishes with a loss, because the fee is charged on the value whether it rose or not. A percentage fee on a mid-single-digit asset converts a flat market into a losing one.

Two cases of a classified Bordeaux at $12,500 a case, bought through a UK merchant in September 2026, held ten years in bond and sold in September 2036: this is the self-directed route, dollar by dollar, for a US investor, and then the same cases on a platform.

Buying the cases

The merchant quotes £9,615 a case in bond, $12,500 at $1.30. No duty or VAT is due because the cases stay in bond. The merchant opens a bonded account in your name at London City Bond and the warehouse issues a statement with two rotation numbers. Your tax basis is $25,000 plus any transfer charge, usually nil.

Holding in bond

Storage and insurance at £14.40 a case plus 20% VAT is £17.28 a case a year, $22.46, or $44.93 a year for two cases: $449 over ten years, assuming the tariff does not rise (it rose in June 2025, so it will, modestly). Nothing is deductible: storage is an investment expense, and §67(g), which suspended that deduction from 2018 and was made permanent by the 2025 tax legislation, gives no relief for it; section thirteen has the detail.

Selling at year ten

At 5.3% a year the two cases are worth $41,901 in September 2036. You list them on a merchant exchange at that price; the exchange keeps 10% ($4,190) and you receive $37,711. A member selling for you into Liv-ex would typically charge less, but 10% is the published, verifiable retail rate.

The tax

The gain is proceeds less basis: $37,711 minus $25,000 is $12,711. Wine is a collectible under IRC §1(h)(4)–(5), via §408(m)(2)’s list (“any alcoholic beverage”), so the gain on a holding of more than one year is taxed at a maximum federal rate of 28%: $3,559. A lower ordinary bracket applies a lower rate; above $200,000 of modified adjusted gross income single or $250,000 married, the 3.8% net investment income tax adds $483; state tax is on top. We show the federal 28% only.

What you keep

$37,711 less $3,559 of tax less $449 of storage is $33,703, about 3.0% a year after everything, from a gross of 5.3%. Sold at a member’s 2–3% commission instead of 10%, the figure would be near $36,000 and about 3.7% a year.

The same cases on a platform

The platform investor’s $25,000 is charged 2.85% on the start-of-year value each year: $713 in year one, rising with the portfolio, $7,964 in total. The portfolio is worth $31,846 at the end of year ten; with no exit fee the gain is $6,846, the tax $1,917, and the investor keeps $29,929, about 1.8% a year. On Cult Wines’ 2.95% entry rate the fees are $8,206 and $29,707 is kept. Every dollar of the $3,774 difference is fee, because every other assumption is identical.

The example is an illustration on stated assumptions, not a forecast, and two of them cut against the self-directed investor: a few hours a year of work, and the chance of buying less well than a professional. Whether a professional buys $380 a year better, which is what the fee gap works out to on this account, is the question to put to any platform that wants the business.

In both the UK and the US, a company that sells you cases of wine and stores them is not a financial firm and no financial regulator supervises it; a fund that pools money to buy wine usually is one; and that distinction has decided who got their money back in every failure in this guide.

In the United Kingdom fine wine is not a specified investment under the Financial Services and Markets Act, so a firm selling cases to private clients, whether merchant, broker or managed platform, needs no FCA authorisation, and its clients have no access to the Financial Services Compensation Scheme or the Financial Ombudsman. Decanter, reporting an FCA warning aimed at over-55s, put it plainly: the vast majority of wine-investment frauds involve companies selling cases of wine, and those companies are not regulated. What is regulated is a collective investment scheme, which is why WineFi says openly that its syndicates are designed to fall outside authorisation.

When a UK wine company fails, the bodies that act are Trading Standards, the Insolvency Service and the courts: the Insolvency Service had Global Wine Exchange put into compulsory liquidation in March 2022 and Bordeaux Fine Wines Ltd wound up in the public interest by the High Court on February 26, 2014 after it took at least £9.3M for wine it never bought, and City of London Trading Standards, not the FCA, handled Oenofuture.

In the United States the line runs through securities law. Vint’s collections were securities qualified by the SEC under Regulation A, which brought offering circulars, annual reports and the going-concern disclosure that preceded its shutdown. A managed account at Vinovest or Cult Wines, in which you own specific cases the platform stores and advises on, is not on its face a securities offering.

Vinovest’s managed portfolios are run by Vinovest Capital Management LLC, and its 2026 acquirer StartEngine owns an SEC- and FINRA-registered broker-dealer, StartEngine Primary LLC; whether Vinovest’s own advisory entity or Cult Wines’ US arm is registered as an investment adviser is something we could not verify from published sources, and readers should check the SEC’s IAPD database and FINRA BrokerCheck before assuming either is. Your protection is title to the wine and the warehouse’s records, not a regulator. The SEC does act where the wrapper is a security or the pitch is a fraud, as with Windsor Jones LLC, and the Bordeaux Cellars prosecution in section twelve was brought by federal prosecutors in Brooklyn.

Offshore fund regulators have a mixed record. Luxembourg’s CSSF suspended Nobles Crus, but only after years of self-marked NAVs; the Bermuda Monetary Authority petitioned to wind up TWIF two decades after launch; Portugal’s CMVM authorised Oeno’s fund in June 2025. A regulator’s stamp on a fund vehicle tells you the vehicle files reports. It does not tell you how the wine is valued or who holds it, and it is the wine, not the vehicle, that you are buying.

Whether you get your wine back when a platform fails is decided by one English case from 1986 and by whether your cases were booked to you by name, and the Oenofuture split of 20% to 80% is what that decision looks like in practice. The same rule governs the adjacent whisky-cask trade, where the Official Receiver winding up one broker in October 2024 found the company owned none of the whisky its customers held certificates for; Investing in Whisky Casks sets out that record.

The case is Re London Wine Co (Shippers) Ltd [1986] PCC 121. London Wine Co sold wine to customers and stored it for them, but did not segregate each customer’s cases; it kept all the wine of a given producer and vintage together and recorded in its stock book how many cases belonged to whom. When it went into receivership the customers argued the wine was theirs, held on trust, ahead of the company’s creditors. The court disagreed: because no specific cases had ever been appropriated to any specific customer, no property had passed and no trust existed, and the customers were unsecured creditors with a claim for money, behind the bank.

The principle has governed every wine insolvency since. If the warehouse records show a case with a rotation number in your name, it is yours and the insolvency practitioner has no claim on it. If the records show the platform as owner and you as a line in its ledger, you have a claim against the platform, paid from what its estate can pay, when it can pay it.

The platforms know this, and the good ones say so. Cult Wines states that it assigns specific cases to individual customers rather than pooling stock and that Coterie Vaults, the partner bond that holds its clients’ stock, keeps records of every owner. Vinovest issues ownership certificates, states that the wine is owned by the client and not by its storage arm, and says that if it went out of business the warehouse would contact clients with the options of continued storage, delivery or sale. Those are the right claims. The test is whether you can confirm them independently: a statement from the warehouse, not the platform, showing the cases and your name, and a right in the client agreement to instruct the warehouse directly.

What fails in practice is rarely the warehouse; London City Bond, Octavian and their peers are regulated warehousekeepers whose business is custody. What fails is the company between you and the bond, in one of three ways: it never bought the wine (Bordeaux Fine Wines, Global Wine Exchange), it bought in its own name and used the stock as working capital (the 80% at Oenofuture, the pooled stock at London Wine Co), or it lent the money elsewhere (TWIF). Segregation in your name defeats the second; only verification defeats the first; only a fund’s governance defeats the third, which is why funds rank last.

The largest wine frauds of the past decade did not involve fake bottles; they involved real-sounding platforms selling wine, or loans secured on wine, that did not exist, and the sums are large enough to end any sense that this is a boutique problem.

Bordeaux Cellars, a London-registered business run by Stephen Burton with James Wellesley, told investors from June 2017 to February 2019 that it made short-term loans to wealthy collectors secured on their cellars of rare Bordeaux and Burgundy and passed the interest through. The indictment, filed in the Eastern District of New York on February 28, 2022, says the pair raised $99.4M from more than 140 investors, 71 of them in the United States, that the collectors did not exist and that Bordeaux Cellars did not hold much of the wine it listed as collateral. The indictment put investors’ losses at about $25M net of the interest paid to them from later investors’ money; by sentencing, prosecutors said about $97M had been raised and only about $14M returned, leaving losses near $83M.

Wellesley pleaded guilty to wire-fraud conspiracy in October 2025 and on April 20, 2026 was sentenced in Brooklyn federal court to ten years by Judge Pamela K. Chen, with $1M forfeited; Burton, extradited in July 2025, pleaded guilty that month to wire-fraud and money-laundering conspiracy and on September 3, 2026 was sentenced to six years with $26M forfeited (Department of Justice; The Drinks Business, April 2026; Bloomberg, September 3, 2026). The pitch was a platform pitch: professional documents, a London address, quarterly interest, collateral described by château and vintage.

Premier Cru was a real wine shop in Berkeley, California, with a real reputation for low prices on Bordeaux futures. Its owner, John Fox, admitted in August 2016 to selling about $20M of phantom wine between 2010 and 2015, using new customers’ deposits to deliver old customers’ orders; the scheme ran back to 1994, some 4,500 customers lost at least $45M, restitution was ordered at $45–50M, and he was sentenced in December 2016 to six and a half years (Department of Justice, NDCA; Berkeleyside). The lesson is specific to this guide: a merchant that takes your money for wine to be delivered later is running an unsegregated pool by definition, and en primeur, with delivery two years out, is the largest such pool in the trade.

Smaller cases fill in the pattern. Windsor Jones LLC took $4.09M from about twelve investors, most of them elderly, between November 2017 and September 2021, spending no more than 32% of it on wine and storage; the SEC sued in February 2023 and later took a $5.8M default judgment. Global Wine Exchange in the UK took £1.9M from mostly elderly investors between January 2019 and March 2021, bought about £770,000 of wine, failed to deliver, and was put into compulsory liquidation in March 2022, after which the Insolvency Service had to warn its victims about “recovery room” callers offering, for a fee, to get the money back. None of these involved a counterfeit bottle; the flagship guide covers Kurniawan and Rodenstock, a different risk with different defences.

The defences are dull and effective. A loan secured on wine you have not seen a warehouse statement for is unsecured. A merchant that cannot show you the rotation number of a case it says it holds for you does not hold it. A return that arrives every quarter regardless of what the Liv-ex 1000 did is being paid by someone else’s deposit.

The route changes the tax bill, and by more than platform brochures admit, because the collectibles rate applies to the gain but not to the fee and because a fund wrapper can convert the whole thing into ordinary income.

Wine is a collectible under IRC §1(h)(4)–(5), which points to the §408(m)(2) list that includes “any alcoholic beverage” (IRS Topic 409). A gain on a case held more than a year is taxed at a maximum federal rate of 28%, against 15–20% for stocks; held a year or less, it is ordinary income. The 3.8% net investment income tax under §1411 applies above the income thresholds, and state tax on top. The 2025 tax legislation left the collectibles rate untouched; it applies for 2025 and 2026. The hub’s flagship walks the same rule for wine and the Precious Metals guide for bullion; the mechanics are identical.

Basis is where the routes diverge. Basis is what you paid to acquire the case, including a buyer’s commission or a merchant’s invoiced charge, so those reduce the eventual gain. A platform’s annual management fee does not. It is an investment expense, a miscellaneous itemised deduction of the kind §67(g) suspended from 2018, and the One Big Beautiful Bill Act of July 4, 2025 amended that section to make the suspension permanent, with a carve-out for educator expenses only.

So the platform investor in the worked example pays $7,964 of fees with no tax relief and then 28% on the smaller gain, while the self-directed investor pays $449 of storage with no relief and 28% on the larger gain. Selling commissions reduce the amount realised on both routes. Keep every invoice and warehouse statement; in this asset the provenance file is also the tax file.

A fund changes the character of the income. Shares in a Reg A wine LLC of Vint’s kind were securities, and if the issuing series was taxed as a corporation a sale of the shares was a long-term capital gain at 15–20% rather than 28%, a genuine advantage of the wrapper while it lasted; if the series was taxed as a partnership, §1(h)(5)(B) looks through and treats the gain attributable to the collectibles as collectibles gain. We could not confirm from the offering circulars which election Vint’s series made.

An offshore fund is the reverse: a Luxembourg, Cayman or Bermuda wine fund is very likely a passive foreign investment company for a US holder, because gains on property that produces no income count as passive under §1297 by way of §954(c)(1)(B)(iii), and the PFIC rules can tax gains as ordinary income with an interest charge and require an annual Form 8621. That is our reading of the Code as it stood in 2026, not confirmed with a practitioner for this piece, and a question for a tax adviser before subscribing.

The UK reader has the opposite problem: HMRC’s manual (CG76901) treats most wine as a wasting asset exempt from capital gains tax but says it would normally contend that fine wine kept for substantial periods is not one; the flagship covers the nuance.

Ranked on what a $25,000 investor keeps after ten years, on the evidence in this guide, the routes fall into four tiers, and the order does not change with the direction of the market.

1. A bonded account at a merchant, in your own name. Berry Bros. & Rudd, Farr Vintners, Lay & Wheeler and their peers will sell you cases in bond, open a warehouse account in your name and store at £14–17 a case plus VAT with insurance included. You pay the merchant’s margin once, storage of about 0.2% a year, and 10% or less on the way out. You do the choosing, with Liv-ex Mid Prices and Wine-Searcher as the check and the flagship’s six filters as the rulebook. This is the route for anyone above $10,000 who will spend a few hours a year.
2. A broker with no management fee. WineCap’s model, storage plus 5% on sale, and Bordeaux Index’s LiveTrade, with firm two-way prices and no membership fee, charge on the spread and the exit rather than on the balance, and both show you a price you can act on. The cost is the spread, paid once. Second rather than first because you rely on one house’s pricing; the defence is to check it against the exchange.
3. A managed platform. Cult Wines and Vinovest at 2.85–2.95% for the entry tiers, and WineFi’s 12.5% up front, are the answer for money that is too small for a merchant to take seriously, too busy to be managed, or, the one good reason, for an allocation the platform can get that you cannot. The ten-year cost against tier one is about $3,800 on $25,000. If you use one, have the cases booked to your name at the warehouse with the statement in hand, and take the lowest tier you can reach: the difference between 2.95% and 2.25% over ten years is about $1,600 on this account.
4. A wine fund. The record from 2003 to 2026 is TWIF wound up by a regulator, Nobles Crus suspended by one, Vint’s pooled collections closed by the sponsor’s own losses, and Oeno’s fund launched five months before its parent’s trading arm stopped. A fund adds a manager’s fee, a valuation policy you cannot audit, a redemption term the market cannot honour in a downturn and, for a US investor, PFIC risk. It gets none of your money unless every question in section six has a written answer.

Tier two edges tier one in the chart because of the exit commission, not the model, and a merchant who sells your cases into Liv-ex at its member rate reverses the order. The top two tiers are within a few hundred dollars of each other and the bottom two are thousands behind, and the line between them is whether anyone charges you a percentage of the balance every year.

A US investor with $25,000 and no wine-trade relationships can be holding two cases in bond, in their own name, in about three weeks, by a sequence that puts the custody check before the money.

1. Pick the merchant, not the wine. Choose a fine-wine merchant that is a Liv-ex member, sells in bond, and will open a bonded account in your name at London City Bond, Octavian or a comparable HMRC-approved warehouse. Ask in writing whether cases are booked to clients by name with rotation numbers, whether you can instruct the warehouse directly, and what the seller commission is on the merchant’s exchange or buy-back.
2. Set the price check. Get Liv-ex Mid Prices through the merchant or a data subscription and use Wine-Searcher as the retail cross-check. Buy nothing whose exchange price you cannot see on the day.
3. Buy to the flagship’s filters. Two cases, not twelve bottles of different things: a top producer, a great vintage for that region, a 95-plus score, original wooden case, continuous in-bond provenance, and a wine that trades often enough that a bid exists. The flagship names the candidates by region; the sister guides on Bordeaux and on Burgundy and Champagne go deeper.
4. Pay in bond. Confirm the invoice shows the price in bond with no duty or VAT and that the cases move bond-to-bond. Keep the invoice; it is your tax basis.
5. Get the warehouse statement. Within two weeks you should hold a statement from the warehouse itself, in your name, listing each case, its rotation number and its insured value. If the merchant cannot produce one, the wine is in the merchant’s name, and you go back to step one.
6. Set the calendar. Once a year, check insured values against the Mid Price, and check the merchant’s accounts at Companies House if it is UK-based, because the failure that costs you is the merchant’s, not the bond’s.
7. Plan the exit at purchase. Decide at purchase whether you will sell through the merchant’s exchange at 10%, through a Liv-ex member at 2–3% plus handling, or at auction for a trophy lot. Anything you cannot name a bid for should not be in the two cases.
8. If you use a platform anyway. Take the lowest-fee tier you qualify for, get the warehouse statement in your name before funding, confirm in writing that cases are allocated at cost rather than a transfer price, and set a date two or three years out to move the account to a merchant.

The view in this guide changes if any of the following crosses its threshold, and each is dated so that a reader in 2027 can see what has moved.

• Platform fees. As of September 2026 the entry tiers are Cult Wines 2.95% and Vinovest 2.85%. A published entry-tier fee at or below 2.0%, storage and insurance included, with no exit charge, would move the managed route from tier three to tier two for accounts under $25,000. A move above 3% at any platform is the signal to leave it.
• Our tape. The Liv-ex 100 read 322.92 on August 31, 2026, 24% below its September 2022 peak, and our Fine Wine sub-index 103.27 on September 8, 2026. Above roughly 370, the October 2021 level at which the “+272%” headline was written, the flagship says the easy part of the recovery is over; below 309, the August 2025 trough, the fee arithmetic gets worse still because platforms charge on the way down.
• The CultX Global index. 155.62 on July 1, 2026 on our tape, a 3.6% compound rate since 2014. If it fails to outrun the Liv-ex 100 over a full cycle, the platform’s selection has not covered its fee; if it beats it by three points a year for five years, the tier-three placement should be revisited.
• Cult Wines’ next accounts. The year to December 2024 showed a £5M pre-tax loss, net liabilities of £21.6M and an auditors’ material-uncertainty paragraph on going concern. The 2025 accounts, due at Companies House in 2026, are the most important document for any client whose wine sits in Coterie Vaults: rising net liabilities or a repeated going-concern qualification is the moment to obtain a warehouse statement and consider moving.
• The Oenofuture liquidation. The liquidator’s first report will state what fraction of the 80% of pooled client wine was actually there. If it is much less than 80%, the case joins Bordeaux Fine Wines as a never-bought fraud rather than a custody failure.
• Regulation. Wine sold by the case remains outside FCA and SEC supervision as of September 2026. A UK move to bring managed wine portfolios inside the FCA perimeter, or an SEC action treating a managed wine account as an advisory relationship, would change the counterparty analysis in the investor’s favour.
• The fractional revival. Vint wound down in June 2026. A new Reg A wine platform with audited financials, a third-party custodian and a secondary market not run by the sponsor would be worth a look at the $1,000 end; without all three it is Vint again.