Wine Market Analysis
Given that there is no cashflow for fine wine and it pays no dividends, the fundamentals to assess its ‘value’ are rather elusive. But what it is possible to assess – and therefore, in our opinion, an essential aspect of wine investment analysis – is relative value. But the vast diversity of the market makes this highly challenging. How do you compare two different wines from different estates, produced in different countries in different years with dramatically different prices? For this, we have developed our MARRVIN (Mathematically Automated Risk and Relative-Value Investment Number-Cruncher) system.
MARRVIN is essentially a sophisticated algorithm which allows us to compare the relative value, risk and financial potential of fine wines. MARRVIN processes market data (critic’s scores, price history and volatility, brand power, production levels, regional vintage quality, optimum drinking age etc.) and calculates various outputs, which when combined help us assess the potential risk and reward for a given wine.
If a wine falls below the relevant benchmark figure we rule it out of our stock selection, allowing us to focus on a reduced universe of wines, and thereby reducing the risk of our recommendations depreciating in value.
The Amphora effect – the benefits of applying quantitative analysis
Wine pricing depends on several different variables – Amphora have identified 9 such factors and weighted them according to their importance in an algorithm – clearly a First Growth will be more expensive than a 2nd Growth, a great vintage more expensive than a poorer vintage etc. The trick is to know how to calculate the correct differentials
The chart right shows the prices of Mouton Rothschild over the most recent 15 physical vintages. “On” vintages are marked as red columns and are as you would expect they’re more expensive that the “off” vintages. Price alone therefore gives you no clue as to which the best investment is.
The blue dots represent the absolute outputs from the algorithm. Again, this absolute figure alone does not guide you to the best investment.
You only get a clear picture of relative value by dividing each of the variables by the price of the wine giving you a Points-per-Pound score.
The results of this process can be seen in second chart, which looks very different. The red bar, 2012 -an “off” vintage, is the very best investment, closely followed by the ’06 (off), ’09 & ’15 (both “on”) on a relative value basis.
Weighted Variables within algorithm
1. Cru Classé status (or equivalent)
2. Brand Power
3. Regional Vintage Quality
4. Robert Parker score
5. Averaged Other Critics’ scores
6. Wine-Searcher rank / Google reach
7. Weighted production levels
8. Liquidity as evidenced by Liv-ex bid/offer spread
9. Supply to market over time.
The example considering various vintages of Mouton Rothschild is a “vertical” analysis – one producer over several years. A “horizontal” analysis considers many producers in a single vintage. The chart below considers the relative value of multiple producers in vintage 2009.
The Amphora proprietary algorithm considers some 167 producers over 23 vintages (so 3841 possible investment opportunities) both vertically and horizontally on a daily basis, thereby modelling the market on a relative value basis.
This process highlights anomalous pricing and pockets of value, as well as reducing the universe of wines considered on an individual basis by the Amphora team. It must be stressed that the Algorithm alone is not a tool for making investment decisions, but rather an invaluable tool in leading the team to promising prospects for the essential qualitative overlay.
However, on a purely quantitative basis, Amphora ran a model portfolio containing the top 34 wines as highlighted by the algorithm since Jan 1 2017. The 2 year results as compared to the Liv-ex 100 Fine Wine Index may be seen to the right.
All charts are for illustrative purposes only, and no purchasing decision should be based upon them.