Mind you, the prices don't even remotely challenge the limited-production wines that make up the top 100 most expensive wines currently available in the world. After all, the current Grange release is only $A850 ($US650) per bottle, whereas these other wines cost thousands (see Top 50 Most Expensive Wines in the World). Mind you, a complete set of Grange vintages (1951-2013) recently sold for $A294,320 ($US224K).
To compete with these more expensive wines, Penfolds also occasionally releases limited-production wines, at thousands of dollars per bottle. Recently, they announced the upcoming release of a wine called G3, which is a mix of three Grange vintages, first blended and then further aged together in barrels (see Introducing Penfolds G3, a new wine born from Grange DNA). The release price will be $A3000 per bottle, with 1200 bottles available by "expressions of interest" only. This is a wine for investment, not drinking.
Needless to say, the media has had a field day, especially in Australia. Philip White has provided a summary on his blog of some of the comments: Penfolds G3 reviews reviewed. For my purposes here, the most pertinent comment has been by Campbell Mattinson, at The Wine Front:
For $3000 you could buy three or four vintages of Grange and, if you really wished, make up a blend yourself. Not only would this be more economical, it would be a more interesting experience / exercise for a wine lover ... Penfolds G3 would have been more interesting, even compelling, had it been released at $1000 per bottle rather than at $3000.In the tradition of this blog, I cannot let these numbers go by without taking a closer look at them. Penfolds has set the G3 at 3.5 times the cost of the three Grange wines being blended. Here, I show that this makes the G3 wine nearly twice as expensive as might be expected, even given that it is a Grange blend.
The expected price of Grange
For my analysis, I first need to work out whether Grange is being released at a reasonable price in the first place. Surprisingly, one can make a case that the current release is actually under-priced.
To see this, we need the prices of Australia's most expensive wines. I started with The Wine Front's most recent list (May 2017) of Australia’s Most Expensive Wines, which is comprehensive but not exhaustive. This compilation consists of current-release still wines listed at $A150 ($US115) and above (ie. no sparkling wines, no fortifieds; and excluding special & limited releases). I then updated a couple of the prices, and added a few more candidates, making a total of 117 wines for my analysis.
The idea here is to use these data to derive an "expected" price for the Grange wine based on the prices of the other wines, so that we can compare this to the actual Grange price. I will do this by relating the wine prices to the rank order of those prices. This is shown in the graph, where each point represents a single wine; only the two most expensive wines are labeled. The graph is plotted with the logarithm of both axes, which means that the Zipf "power model" can be represented by a straight line on the graph, as explained in a previous post.
As shown, the Power model fits the data extremely well (98% of the data are fitted), but only if we exclude the two most expensive wines (shown in pink). This means that neither the Penfolds Grange (the most expensive wine) nor the Henschke Hill of Grace (the second wine) has a bottle price that is in line with the other 115 wines. Indeed, both wines appear to be under-priced!
The Power model predicts that the current release of the Henschke Hill of Grace wine would be expected to have a bottle price of $A1,000, and the Penfolds Grange would have a price of $1,400 per bottle, based on the prices of the other wines. Their current retail prices are $A825 ($US635) and $A850 ($US650), respectively.
So, don't be surprised to see these two wines increase to these "expected" prices in future releases. Neither Henschke nor Penfolds has recently been shy about raising their prices to ensure their primacy in the Australian market.
The expected price of G3
We can now try to derive an "expected" price for the new G3 wine, by repeating the above analysis, but moving all of the wines down one place in the rank order (ie. Grange becomes the #2 wine instead of #1, etc). The new equation for the Power model (excluding Grange and Hill of Grace) will give us a predicted price for the new #1 most-expensive wine in the list.
This turns out to be a bottle price of $A1,550, which is fractionally more than half of Penfolds' proposed price for the G3 wine. So, even given that the wine is a Grange, the proposed G3 price is, indeed, much more than we should expect to pay. Mind you, this expected price is 50% more than Campbell Mattinson is prepared to pay for a Grange blend, so he may actually be under-valuing the wine.
