The relationship (or lack thereof) between scores, prices and appellations
Summary:
Scores, prices and appellations of wines from a subset of premium wines, continue to be unpredictable. No statistical modeling I know of could accurately predict one variable from the two others.
Sub-appellations produce higher scores and prices than do the California, Carneros or county-based appellations.
The highest average points per dollar for 2011 Pinot Noir goes to the Santa Lucia Highlands, followed by the Central Coast.
The worst average points per dollar goes to the Sonoma Coast.
2011 CALIFORNIA PINOT NOIR SURVEY
Right now, I am impatiently waiting for the Crush Report to come out on February 10th. In the meantime, I thought I would do a study of the relationship between wine prices, scores and appellations. In the past, I have done research for clients that attempted to find correlations between price, scores and production volumes. All of these were for wines in the >$20 range and none showed any rhyme or reason to wine scores or prices. A chart of the results looked like a piece of graph paper that had been used for target practice with an air gun.
For my current effort, I decided to restrict my focus to Pinot Noir from California that was grown during the 2011 vintage. I chose Pinot, because I love Pinot; ditto for California. I chose 2011 because pretty much every 2011 Pinot Noir has been released by now. The sample was pulled from all Pinot Noirs reviewed by the Wine Spectator that met these parameters. I looked only at regions with sample sizes of over 30 wines. These include (sample sizes included in parentheses):
California (as in the whole sample – I did not look at wines with no appellation smaller than the statewide one, as the sample size was only four wines of the total 580.)
Central Coast (145)
Santa Lucia Highlands (62)
Santa Rita Hills (58)
Sonoma County (301)
Russian River Valley (171)
Carneros (32)
Sonoma Coast (117)
Mendocino County (48)
Anderson Valley (37)
Prior to looking at the results of this survey, here are a few caveats:
As you can imagine by what I just wrote, a high risk of sample bias exists. Ideally, I would look across reviewers and vintages, but that will have to wait until I can crowdfund an intern.
Wine scores are not a strong proxy for quality, in my opinion, once you are already looking at the subset of “premium” wines. I am not necessarily measuring quality here. I have always said that choosing a wine based on wine scores is like choosing a wife based on wet t-shirt contest results.
The wines that the Spectator tastes are drawn from the best our state has to offer. If I were to have added in every supermarket Pinot Noir and all the other nonsense being bought on the spot market, the results would likely differ.
OK, so what did I learn?
Nothing Makes (Statistical) Sense
Price does not predict scores at a statistically significant level. Same thing vice-versa. No combination of appellations, when added in, can raise correlations above 40%. Most appellations are unreliable variables and even those that tested high in reliability account for only a few percent of the price of the finished bottle.
In a way this makes sense. Yes, Anderson Valley produces much better Pinot Noir than the rest of Mendocino County. Yes, Anderson Valley Pinot sells for much more than inland Mendocino fruit. But still, even at, say, $1000 more per ton you’re looking at about a $1.55 price difference for a $50 bottle. The rest of the cost is production, packaging, sales, marketing and overhead. None of those input costs varies with grape provenance. The premium paid, of course, does make all the difference in the world for the grower’s income and the value of his asset, but not as much for the cost structure of the producer.
Remember, though, that this lack of effect on bottle price comes with two enormous caveats. First, this is only within our narrow sample of wines that all cluster near each other in perceived quality. Only 4 of the nearly 600 bottles in the sample had the appellation “California.” Most were from the most well-regarded appellations anyways. Of course a bottle of Anderson Valley Pinot Noir will sell for far more than one labelled “California.” That is not what is being studied here. I am looking at whether some AVA stands out as a real driver of price when compared to the other coveted AVAs. The second caveat is that I have been talking only about statistically supportable conclusions up to this point. But let’s depart from that path…
Wines from “Sub-Appellations” Receive Higher Scores and Prices
So, from this point on, no claim I make would stand up to the same level of statistical rigor that I apply to anything I do for work, but some of it might fly in a room full of marketing professionals. Take it with a grain of tartrate. Also keep in mind that I am using as a standard whether or not one could predict one variable (price, score or appellation) based on the other two. Clearly, the variables are related, the correlation just isn’t strong enough to provide predictive capabilities.
Anyways, wines from sub-appellations received higher scores and were priced higher than those from broader appellations. Which areas qualified as sub-appellations? Any that were not countywide, statewide or Carneros. The latter I left out since it is such a large AVA that straddles two counties, instead of being sub-set of countywide appellation. So, while Sonoma County would not be a sub-appellation, Sonoma Valley would be. Wines from sub-appellations averaged a score of 88.13 and a price of $54.11, while those without averaged 86.80 and $44.00, respectively.
You Should Buy Pinot Noir from…
I took the average scores and prices of the appellations mentioned above and put them into ranked charts.
Region by score:
Score: Region:
88.94 Santa Lucia Highlands
88.31 Central Coast
88.13 Russian River Valley
87.97 Anderson Valley
87.95 Sonoma County (Any)
87.81 Santa Rita Hills
87.79 Sonoma Coast
87.54 Mendocino County (Any)
86.72 Carneros
Region by price:
Price: Region:
$55.63 Sonoma Coast
$54.43 Anderson Valley
$54.32 Russian River Valley
$53.98 Sonoma County (Any)
$53.00 Mendocino County (Any)
$52.97 Santa Rita Hills
$51.08 Central Coast
$50.26 Santa Lucia Highlands
$47.97 Carneros
Region by Score, with Price Rank Included:
Score Rank: AVA: Price Rank:
1 Santa Lucia Highlands 8
2 Central Coast 7
3 Russian River Valley 3
4 Anderson Valley 2
5 Sonoma County (Any) 4
6 Santa Rita Hills 6
7 Sonoma Coast 1
8 Mendocino (Any) 5
9 Carneros 9
Now, I myself will not use this chart to determine what wines I drink – and not just because the conclusions do not stand up to my high levels of statistical significance. I go by my own palate and not by the palate of tasters whose salary is paid by producers’ advertising. Furthermore, I do not like to give too much weight to aggregate data. That is to say, just like not all Zinfandel is good, not all 2011 Pinot Noir from the Santa Lucia Highlands will be as great as a deal as the region is overall, were we to accept the scores as valid measurements of quality. Oh, there I go again, all doom and gloom and no fun. I know, with such a lack of suspension of disbelief, I must be tons of fun to watch romantic comedies with, huh?
So, let’s forget about my cantankerous views for a bit. What do the above rankings say about each AVA for buyers and for AVA values (see my last blog post to understand why this is important).
The Santa Lucia Highlands ranked tops in scores, but bottom in price. OK, even I’m intrigued enough by this that I’ll probably buy a bottle or two from there next time I get a chance. According to the data, 2011 Pinot Noir from the Santa Lucia Highlands offers the most points per dollar of all other ’11 Pinot. The bad news here for growers is that, until the market corrects this, their AVA is not working that hard for them. They are, according to the scores, in an area that produces fruit as high quality as anywhere else, yet they’re not getting top dollar for the wines their grapes are made of. I’m not too familiar with the grape market down there, but I would guess these relatively lower prices trickle up the supply chain. The good news is that this is only in comparison to the best Pinot AVAs in the country. Their AVA should still offer plenty of value for amortization as a tax shield. Looking at land prices, this may be a good place to invest, based on the assumption that prices will rise over time.
The larger Central Coast appellation, of which the Santa Lucia Highlands is a part, is just behind the SLH on value, at 2nd in scores and 7th in quality. Speaking of… how much more land is plantable on the Central Coast? We’re nearly out of good Pinot land up here.
The Russian River Valley looks to be properly priced, coming in 3rd place for both scores and prices.
The Anderson Valley is outpriced only by the Sonoma Coast. Yet it comes in fourth place in terms of scores. The AVA is doing its work, pushing prices above those of the Russian River Valley, despite the fact that the reviewers see RRV wine as being better. If this sample is valid, then Anderson Valley producers should be making healthy profits. They are paying almost $500 per ton less than purchasers of Russian River Valley fruit, but charging more for their wine. Even if they were farming their own land, they would be paying less for the capitalized expenditure of obtaining the land, which is selling for less than Russian River Pinot land.
Sonoma County as a whole looks to be pretty fairly priced and the Santa Rita Hills is spot on.
Mendocino and Carneros round out the bottom of the list on scores, but only Carneros is priced as such, with Mendocino County being priced higher than it should be (I concur on that, by and large.)
Finally, the Sonoma Coast, according to these measures, is a huge rip-off for 2011 vintages. Considering the weather that year, that may be valid. It is, however, a strong competitor for my favorite AVA for Pinot Noir. If you are a grower out on the Sonoma Coast who has purchased your land after the summer of 1993, you better be amortizing you AVA, though, because the value there looks to be enormous.
Anyways, I’ve still got a month before the Crush Report comes out. Anything that any of you would like to see me take on? Contact me>>>