Retroactive Blending

You don’t often get the chance to reconsider the blend ten years on, but this is what happened when I visited Château Léoville Lascases in St. Julien. We started with a tasting of the individual varieties from 1999. That year the Grand Vin was 62% Cabernet Sauvignon, 18% Cabernet Franc, and 18% Merlot (there is no Petit Verdot because they believe it is too rustic.) Samples of the individual varieties were bottled separately (starting in new oak and then transferring to one year oak, to give an overall exposure close to the grand vin’s 60% new oak).

The Merlot showed surprisingly fresh red fruits, with just a touch of tertiary development. The Cabernet Franc was evidently more refined, more elegant, than the other varieties and showed a faint herbaceous touch with an impression of tobacco. It was less developed than the other two varieties.  The Cabernet Sauvignon was quite stern, and gave the most complete impression of any of the single varieties, showing as black fruits with a herbal edge and a touch of herbaceousness showing only on the aftertaste. It’s the most closely related (not surprising since it’s dominant component) to the Grand Vin.

The Grand Vin showed more development than was evident with any of the individual varieties, bringing greater complexity. This has certainly taken its superficial softness and roundness from the Merlot, but you can see the Spartan structure of the Cabernet Sauvignon coming through the fruits; in fact, in some ways it seems more evident here than it did in the sample of Cabernet Sauvignon alone (perhaps because the combination of fruits has less weight than the Cabernet Sauvignon alone), but the overall balance is rescued by the freshness of the finish. There is no doubt that the blend is more complex than its components. In terms of overall assessment, this is a fairly tight wine, with the fruits showing just enough roundness to counteract the leanness of this difficult year.

The most fascinating moment came when technical director Michael Georges made some new blends to see what the effect would be of increasing each variety by another 10%. I liked the blends with more Cabernet Sauvignon or Cabernet Franc; they seemed to me to have at least as good a balance of fruit to structure as the Grand Vin. I could believe that either of them might be Léoville Lascases. But the blend with additional Merlot seemed to be unbalanced, to have a rusticity that had lost the character of St. Julien: I would not believe in this as a Léoville Lascases. The trick seems to be to add just enough Merlot to flesh out the wine, but not enough to go over the edge into rusticity. Further experimentation suggested that the ideal blend might have just 5% more of each Cabernet; it seemed to me that this showed just a touch more finesse than the Grand Vin. “Perhaps we should wait ten years to do the assemblage,” said Michael Georges, but then we agreed that this might have some adverse financial consequences.

For me this tasting also cast an interesting light on the question of whether assemblage should be done early or late. Some people believe that the sooner the cépages are blended, the better they marry together, and the better the final wine. The earliest practical moment is after malolactic fermentation is finished. Others hold the contrary position, that you are in a better position to judge the quality of each lot if you keep the individual cépages separate until the last moment. I felt that the retroactive blend with 5% more of each Cabernet had more youthful liveliness than the Grand Vin, but then it might of course have developed differently had this been the blend from the beginning. Based on this limited experience, I’m inclined to the view that it might be best to mature each lot separately, allowing for significant adjustment of oak and variety, as long as possible, and I think it would be very interesting to see what the châteaux would do if they weren’t under pressure from the en primeur system to blend before the April tastings.

Bordeaux and Robert Parker

There is no doubt about Robert Parker’s influence on Bordeaux. Ever since he famously got the 82 vintage right whereas many others got it wrong, he has been by far the most influential commentator on Bordeaux’s annual release. His influence, together with that of other critics, is much increased by fact that most Bordeaux wines are sold en primeur, pushing consumers to make their purchasing decisions long before they have a chance to taste the wines. Unless you want to pay prices that in good vintages can be much higher by the time the wines are available, you have to rely on the opinions of critics who taste en primeur.

In most vintages, chateau owners wait with bated press for the April issue of the Wine Advocate containing Parker’s ratings of the vintage. When Parker did not make his visit to assess the 2002 vintage (because of the Gulf war), Jancis Robinson commented that the Bordelais would now have to relearn the whole art of selling their wine en primeur – without Mr. Parker. But in every other vintage for the past three decades,the Wine Advocate rating has been an important factor in determining the level of interest in the vintage – and presumably in individual chateaux.

There are certainly some striking examples of the effect of the Wine Advocate’s ratings of individual wines. In the 2001 en primeur campaign, Climens was its usual one third above the price of Suduiraut and Rieussec. When it was promoted to a perfect score of  100 in the final review in June 2004, its price differential over the others increased to 200-300% worldwide. In the en primeur campaign for the 1990 vintage, Chateau Montrose’s release price was in its usual position just below Cos d’Estournel. The Wine Advocate’s review of the vintage en primeur rated it highly, but then a re-review after bottling promoted it to 100 points. Immediately its price reached double that of Cos d’Estournel at the auctions.

Both these examples represents cases where the wine was available on the after-market when the 100 point review came out. I’ve been looking for some way to assess whether critical influence has the same effect when the wines are en primeur. I’ve just spent a week in Bordeaux working on my book on Cabernet Sauvignon, but also picking up data for a new edition of my earlier book, What Price Bordeaux?, including en primeur prices from the Place de Bordeaux for recent vintages. The 2008 vintage is an interesting situation that might answer the question.

2008 was a difficult vintage in Bordeaux. Various climatic problems resulted in a rather small vintage, decent but not outstanding in quality. When the en primeur campaign started in April 2009, the world economy was in full retreat and it was not obvious to the Bordelais how they would be able to sell the vintage. The first growths took an unexpected lead. Usually they come out at the end of the campaign, but this year they acted in concert right at the beginning, dropping their prices by almost half from 2007. This was a clear signal to the market to create value. During the second half of April, chateaux came out at prices ranging from two thirds of their 2007 prices to a couple at parity.

The opinion of the vintage expressed in the Wine Advocate at the end of April was a surprise. Mr. Parker declared that the vintage was much better than had generally been appreciated. “I was worried that at best, quality would be average to above average… several French newspapers came out with stories about the deplorable quality of the 2008 Bordeaux vintage,” he said, but then found that “It did not take me long to realize that the 2008 vintage was dramatically better than I had expected… the quality of the 2008 vintage turned out to be excellent, with a number of superb wines that are close to, if not equal to the prodigious 2005 or 2000 vintages.”

Comparisons with 2005 and 2000 might be expected to give a definite lift to the reputation of the vintage. Indeed, chateau proprietors who had already committed their prices felt somewhat rueful that they might have done better to wait. But those who came on the market later did not in fact see much significant gain. Excluding the first growths, chateaux declaring during April averaged 87% of their price relative to 2007, while those coming out in May averaged 94%. The difference is mostly due to a small number of chateaux whose prices were rather low right at the beginning of the campaign, around the time the First Growths came out. If the Wine Advocate report had a direct effect, it was probably to reinforce a view that was already developing that extreme reductions (below 80% relative to 2007) were not necessary. As a scatter plot of the data shows, the majority of chateaux were in a range between 80% and 100% of their 2007 prices throughout the campaign (the First Growths are in red).

Chateau prices for 2008 relative to 2007 during the en primeur campaign

Now I am not suggesting that the Wine Advocate doesn’t have influence: far from it. But I suggest the data show that there are other important factors also, in this case the combination of a poor reputation that (rightly or wrongly) had already been established for the vintage, together with the background of economic difficulties. Generally speaking, when the Wine Advocate presents its view of the vintage, it’s against a background where expectations have already been created, and to some extent the detailed ratings refine and extend those expectations. When it’s reputed to be a great vintage, consumers are looking for assistance in making their decisions. It seems they are not so easily swayed when the news is significantly different from their expectations. In the same way, I showed in What Price Bordeaux? that the most important influence on the relative price of a wine in any year is the price it achieved the previous year. It’s not that criticism isn’t important, but that – with the notable exception of cases where a wine is give a perfect score that lifts it right up out of its usual situation –  critical opinion takes some years to affect the annual pricing.

So the naysayers who worry that Robert Parker (or any other critic) has undue influence over the entire process might take some comfort from the thought that there are many influences at work here. Some of those influences are due to informed opinion, some are due to propaganda, some are due to ignorance, and when you look back years later at the supposed vintages of the century, it’s far from evident that the outcome, however determined, is particularly accurate at the time. Certainly it’s not easy, even with experience, to be sure of judging wines after only six months in the barrel, and if the blame is to put anywhere for the rush of the lemmings, it’s on the system rather than individual critics.