What is the WineFi Investment Score (WIS)?

TL;DR

The WineFi Investment Score (WIS) is WineFi’s proprietary quantitative methodology used to evaluate and rank investment-grade wines. The framework assigns each asset a score from 0 to 100 based on its potential to outperform average market returns over the coming investment period. By applying machine learning models to a dataset of approximately 300,000 investment-grade wines, the WIS methodology has historically identified price patterns that outperformed standard fine wine benchmarks by an average of 6.73% annualised during backtesting from 2009 to 2025.

Why does the fine wine market need a quantitative approach?

Fine wine price discovery is slow. Wine is a relatively illiquid asset class compared to equities, meaning prices do not update continuously on a trading screen. Instead, value is formed through discrete transactions, which can create significant information asymmetry between buyers and sellers. A data-driven approach removes emotion from the selection process, allowing investors to identify mispriced assets before the broader market reacts.

The traditional approach to wine investing relied heavily on narrative, critic scores, and personal relationships with merchants. While these elements still hold value, they are no longer sufficient for managing a modern alternative asset portfolio. In fact, they can be detrimental when used alone, over-speculation on high-scoring vintages has been common place in recent years, leading to lower-than-average returns of highly regarded wines from legendary vintages.

The investable universe of fine wine is estimated at roughly $5.5 billion globally. Navigating a market of this size requires systematic analysis. The size of the market is a critical factor, as only a fraction of a percent of global wine production consistently meets the criteria required for investment.

Because fine wine is an illiquid, real asset, its market behaviour is closer to residential property than to equities. Returns depend heavily on the entry price, asset selection, and the specific holding period rather than short-term sentiment. This structure means that relying on reputation alone is a flawed strategy. Investors require a methodology that objectively calculates fair value and forecasts future appreciation potential. Data provides the scale and precision needed to uncover these opportunities consistently, processing market movements across thousands of variables simultaneously.

How does WineFi pick wines? The step-by-step investment process

WineFi selects assets by combining rigorous quantitative analysis with the deep domain expertise of our veteran Investment Committee. The role of data is not to replace human judgement, but to scale it efficiently across a massive global market. This process systematically identifies opportunities the market has mispriced before any capital is deployed.

The objective of this methodology is to generate alpha relative to defined benchmarks, specifically the Liv-ex indices. To achieve this, we follow a strict sequence for every potential acquisition:

1. Define Scope: The process starts with a global universe of approximately 300,000 wines, from all investment-grade producers. We apply structured filters regarding region, pricing, age, and liquidity constraints to eliminate wines that are unlikely to be investable in practice.

2. Apply the WineFi Investment Score (WIS): Each wine remaining in scope is then filtered based on its numeric WIS rating, removing those at the lower end of the 0-100 scale. The score is designed and backtested to rank wines based on future price appreciation potential.

3. Human Review: The quantitative models serve as an initial filter. Wines that pass the algorithmic threshold are subsequently reviewed by our expert Investment Committee, led by Matthew Small. During this stage, assets may be accepted or rejected based on qualitative factors and market intelligence that quantitative models alone cannot capture.

4. Investor Approval: Once a compelling collection of wines has been identified, the opportunity is presented to investors. This happens on either a discretionary basis for private portfolios or a non-discretionary basis for syndicates.

5. Sourcing: We enter the market to acquire the underlying wines, aiming to transact at or below our target prices. Acquired wines are immediately transferred into secure bonded storage, with any pricing discounts achieved during negotiation passed directly to our investors.

6. Exits: WineFi continuously monitors both broader market conditions and individual wine performance. Exit recommendations are made on a case-by-case basis when an asset has reached its target return.

What is the dual-model framework behind WIS?

The WineFi quantitative framework does not rely on a single algorithm. Instead, the WineFi Investment Score is built around several closely linked models that fall under two categories and continually reinforce one another. This dual-layered approach ensures that we evaluate both current value and future trajectory simultaneously.

The first component is the Efficient Market Price Model. This model evaluates the current market landscape to estimate an "efficient market price" for each individual wine. By comparing this calculated fair value against actual market offers, the model systematically identifies wines that are currently underpriced or overpriced. In an opaque market where pricing can vary wildly between merchants, knowing the true fair value of an asset before acquiring it provides a critical structural advantage.

The second component is the Returns Ranking Model. Identifying a cheap wine is not enough; the asset must also possess a clear path to growth. This model forecasts which specific wines are most and least likely to appreciate over the next 5 years. It analyses historical trajectories and current market dynamics to project future demand.

Together, these two models form the foundation of the WineFi Investment Score. The combination allows us to target wines that represent excellent value upon acquisition while exhibiting strong forward-looking appreciation potential. Historical backtesting from 2009 to 2025 demonstrates that this dual-model framework improves asset selection, outperforming benchmark returns by an average of 6.73% annualised.

What data variables power the WineFi Investment Score?

Our quantitative models are trained on a proprietary dataset comprising pricing, trade, bid, listing and offer data, representing all investment-grade wines currently in existence. This core market data is combined with over one million critic reviews and qualitative attributes to forecast investment performance with precision.

A model is only as effective as the data it processes. The fine wine market is highly fragmented, making granular data collection essential for systematic asset selection. Our methodology tracked hundreds of variables and includes more than 40 in the current model version, which are optimally weighted through extensive historical backtesting.

These core variables include:

• Price metrics and long-term trend analysis: Understanding historical price movements is essential for identifying cyclical patterns and establishing a baseline for future growth.

• Critic scores, tasting notes, and drinking windows: Quality assessments from recognised critics heavily influence global demand, while drinking windows indicate when a wine will reach its peak desirability and optimal liquidity.

• Brand power and producer track records: Strong brand equity provides downside protection and ensures consistent secondary market interest, particularly during economic downturns.

• Risk and volatility across vintages and cycles: Assessing historical volatility allows us to construct portfolios with asymmetric risk profiles, aiming for limited drawdowns and steady appreciation.

• Market liquidity and supply-demand dynamics: We evaluate transaction frequency and bid-to-offer ratios to ensure that acquired assets can be exited efficiently when the time comes.

• Vintage quality and climatic factors: Weather conditions during the growing season fundamentally dictate the quality and scarcity of the resulting wine.

• Regional and appellation characteristics: Strict geographic and regulatory frameworks, such as the DOCG system in Italy, constrain production and enforce structural scarcity.

• Formal classifications and rankings: Historic hierarchies, such as the 1855 Classification in Bordeaux, continue to anchor global pricing and demand.

• Relative value metrics: How does this wine’s price compare to similar wines and to our estimated efficient market price? And, what does that suggest about its future growth potential?

Why is machine learning necessary for fine wine investment?

Traditional financial modelling struggles with fine wine because the data is highly categorical and the relationships between variables are often non-linear. Machine learning techniques are required to identify the complex price patterns that persist across market cycles and varied regional dynamics.

For example, a high critic score might trigger a massive price increase for a mid-tier Burgundy, but have a negligible impact on a globally recognised First Growth Bordeaux where premium quality is already priced in by the market. These non-linear relationships mean that simple spreadsheet analysis is entirely inadequate for forecasting returns accurately. Machine learning models capture these nuances, adapting to how different factors interact across varying price tiers.

Furthermore, fine wine involves highly categorical data. The market is divided into thousands of distinct producers, appellations, vineyards, and vintages, each with its own unique characteristics and market behaviour. Certain machine learning models excel at parsing this categorical complexity, allowing for nuanced distinctions between a Grand Cru and a Premier Cru site, or determining how specific region-vintage combinations age more favourably than others over a ten-year holding period.

Academic research on quantitative asset selection frequently highlights that machine learning algorithms consistently outperform traditional linear regression models in alternative asset markets. Studies examining the application of artificial intelligence in illiquid markets show that algorithmic evaluation reduces selection bias and improves risk-adjusted returns by systematically identifying hidden pricing inefficiencies.

The impact of holding periods on quantitative models

Fine wine does not behave like a traded financial equity; value is realised over time rather than day-to-day. Our models are backtested primarily on five-year holding periods, as historic data demonstrates that returns improve and outcome volatility declines as the holding period increases.

Because wines generally improve in quality as they age toward their peak drinking window, market interest and price discovery naturally increase over time. Scarcity accelerates as bottles from a specific vintage are inevitably consumed or damaged globally. Our algorithmic asset selection weightings are deliberately optimised around five-year horizons. This strategy is designed to capture the most productive phase of a wine's lifecycle, avoiding the capital inefficiency of holding an asset indefinitely in a hope of capturing its fastest growth period.

Why average returns do not equal your returns

The reality of the fine wine market is that broad market averages conceal a massive dispersion of individual outcomes. The WineFi Investment Score was created specifically to navigate this dispersion, isolating the small percentage of assets that actively drive market performance.

Looking at historical market performance across a ten-year window, 11.1% of investment-grade wines achieved 10% or more annualised returns, while 2.7% yielded negative returns. A passive approach to fine wine acquisition exposes an investor to this entire spectrum of outcomes, including the underperforming assets.

The WIS framework acts as a highly calibrated filter, stripping away the lower percentiles. By systematically identifying the characteristics shared by the top-performing 11.1%, the quantitative models construct portfolios designed to consistently outpace the broader market average.

How quantitative wine investment connects to your portfolio

The fundamental goal of the WineFi Investment Score is to narrow the gap between simply owning fine wine and owning the right fine wine. By replacing speculative assumptions with a rigorous data-driven framework, we provide a structured path to genuine capital appreciation in the alternative asset space.

Investors seeking to diversify their holdings with real assets require confidence that their allocations are built on solid fundamentals rather than subjective opinions. The WIS methodology provides that exact transparency. Whether you are accessing the market through a private portfolio or participating in one of our fine wine syndicates, you can be assured that every asset has been comprehensively vetted for liquidity, fair value, and appreciation potential. For more information on how to start your journey, see our piece on how to start investing in fine wine in the UK.

Explore the 2026 Fine Wine Investment Guide to learn more about market cycles and regional variance. Ready to allocate capital? Sign up on our homepage to view our current investment opportunities.

Frequently asked questions

What is the WineFi Investment Score (WIS)?

The WineFi Investment Score (WIS) is a proprietary quantitative rating system that ranks investment-grade wines on a scale of 0 to 100. It evaluates an asset's potential to outperform average market returns over the coming investment period.

How does quantitative wine investing actually work?

Quantitative wine investing applies data science and machine learning to the selection of fine wine assets. Rather than relying on subjective tasting notes, quantitative models process massive datasets involving historic prices, liquidity metrics, and weather patterns to forecast future returns and optimise portfolio construction.

Does WineFi rely entirely on algorithms to pick wines?

No. The quantitative output from the WIS models is paired with deep domain expertise from our veteran Investment Committee. Every wine that passes the algorithmic filters undergoes a rigorous human review to account for qualitative market dynamics before any capital is deployed.

What makes a wine investment-grade?

An investment-grade wine is defined by its market behaviour rather than simply its quality or taste. Key characteristics include established secondary market liquidity, a proven track record of international demand, recognised provenance, and limited production volumes constrained by strict geographical regulations.

How does the WineFi Investment Score generate alpha?

Through historical backtesting covering the period from 2009 to 2025, the WIS model has demonstrated the ability to outperform standard benchmark returns by an average of 6.73% annualised. It achieves this by systematically identifying pricing inefficiencies and patterns that persist across fine wine market cycles.

What data goes into the WineFi models?

The models process a proprietary dataset of approximately 300,000 wines and over one million critic reviews. It assesses over 40 variables, optimally weighted through backtesting, including long-term price trends, producer track records, supply-demand dynamics, and regional climate factors.

Are algorithmic wine portfolios completely automated?

While the data processing and initial asset ranking are automated, the final investment decisions are not. The models serve to scale human judgement by rapidly filtering the global universe of wines, allowing the Investment Committee to focus their expertise on a curated list of high-potential assets.

This article is provided for general information and is not personal tax or investment advice. Capital is at risk. Wine values can go down as well as up, and investments may not perform as expected. Returns may vary. You should not invest more than you can afford to lose. WineFi is not authorised by the Financial Conduct Authority. Investments are not regulated and you will have no access to the Financial Services Compensation Scheme (FSCS) or the Financial Ombudsman Service (FOS). Past performance and forecasts are not reliable indicators of future results. Investments are illiquid. Tax treatment depends on individual circumstances and may change. You are advised to obtain appropriate tax or investment advice where necessary. WineFi is a trading name of WineFi Management Limited.