sensible pricing of left bank red Bordeaux

[rthomaspaull]

Is anybody interested in the above topic? rthomaspaull

[Racer Chris]

Hi Richard,
It's good to see you over here.

[rthomaspaull]

Thanks, Chris. The other forum seems more active. rthomaspaull

[jal]

Is anybody interested in the above topic?

Everyone here is.
But your question is very vague. Can you be more specific? Give examples?

[William P]

Welcome Richard. We are a little slower in responding. We are also a small board. What is your context of "sensible pricing"? Do you mean what is the worth of left bankers as compared to the rest of the red world? Maybe you mean, what are the left bankers that are great qpr's? Maybe you want to discuss the economic costs, so is the price of LBers sensible given the cost basis of production? If you are talking about sensible pricing the question always arises sensible to whom? A young family person has a much different criteria of sensibility as opposed to a professional who is ready to retire.

Don't mind me, I am easily confused.

Bill

[rthomaspaull]

To jal and William P: you make good points but I am trying to find what would be a reasonable maximum price for me for a wine of given quality.
A few years ago I would have paid before taxes and shipping $160 for a left bank red Bordeaux that I gave 100 points (that only happened once, with a 1982 Leoville Las Cases bought for $34.97 in 1986 or 1987 and enjoyed with family on the last day of 1999) but today I would pay $170.
For each point below that I would divide the price to pay by 1.108, and for each half-point lower I would divide the price by 1.0524.
THe reason for the 1.108 is that a QPR article giving prices of early 2000s left and right bank red Bordeaux gave $25 as the average price for 85 point wines {bottom of fifth growths per Grundeken (see later)} and gave $34 as the average price for 88 point wines (bottom of third growths per Grundeken).
(34/25) to the power of 1/3 equals 1.108. "Grundeken" refers to a mid-2004 (I think) article called "Reclassifying Bordeaux using Wine Advocate Ratings") or something like that. All ratings were by Robert Parker (RP) and the period covered was 1982 through 2003 excluding 4 poor years (the 2002 and 2003 ratings were presumably from barrel rather than bottle). The bottom scores for descending orders of classified growths were 92.5, 90.0,88.0, 86.5 and 85.0.
It is true that right bank as well as left bank wines were considered, but I sometimes drink right bank wines and equal scores should mean equal quality. Wines with higher ratings became more and more disproportionally out of line as the wine ratings rose, so these were ignored as being unreasonable. An 85 point wine is worth a maximum of $36.52 but could probably be bought for considerably less. The picture at the high end is much worse.

[Racer Chris]

I would not pay $25 for a wine I scored 85 points.

[rthomaspaull]

To Chris F. : I completely agree with you and I am sure you could find an 85 point wine for much less. Perhaps I should have chosen say $20 maximum for an 85 point wine and done the process in reverse, but many of the prices would be way lower than I would pay for really good wines. All our Bordeaux wines are scored between 90.5 and 96.5 by me. None cost as much as $85 and all were well within my price limits. Beyond 96.5 points things get really difficult but that is presumably because really good wines are very often overpriced. rthomaspaull

[rthomaspaull]

In his book "Wines of the World" Andre Simon discussing I believe the early 1820s wrote something like " a second growth cost about about a quarter less than a first growth". Other classes cost about a fifth less than the next higher growth. There were no fifth growths as such, only "deuxiemes quatriemes" ("second fourths") so I used a simple average of fourths and fifths in checking to see how well Grundeken figures matched up. the fit was extremely good for 1.108, within 1 percent in each case.
In his book "Bordeaux" Clive Coates writing about the unofficial 1845 classification stated something like "so a lowly fifth would bring 40% of a Chateau Lafite". At the time Lafite was the second ranked first growth. In Grundeken the second ranked first growth averaged 94.1 and the lowliest fifth averaged 85.1 . The difference is 9 points. 0.40 to the power of 1/9 equals 1/1.107 .
I calculated average scores of many left bank red Bordeaux from 2002-2014 (except the poor year 2013), all ratings being bottle ratings and made by Robert Parker (RP) through 2012 and 2014 by RP's then designated successor Neal Martin. One point was added to each classified growth rating due to average Wine Advocate Vintage Ratings being higher by a rounded 2 than for Grundeken. Paired results were closer to 1.109 per point than 1.108.
Thus !.108 works extremely well. It is very possible that the DIFFERENCE in points between properly ranked classes has stayed fairly constant over the years. rthomaspaull

[DavidG]

I was told there would be no math.

I’m out...

[stefan]

Let's use letters instead, David. Richard is saying that if f(y) is the average price of a wine rated y, then f(y+1)/f(y) is approximately the same no matter what y is (except maybe when y is close to 100). In other words, to get a wine A that is one point better than a wine B, you expect the price increase to get A to be a fixed percentage of the price of B.

Without seeing the data, would you have thought that to be the case? If so, would you have guessed that the fixed percentage is about 10.8%?

[jal]

I think my head just exploded

[DavidG]

I'm actually ok at math, though not in the same universe as Stefan. I was just being cheeky.

Give me a taste and a price. That's all I really need or want. Past experience with the producer and knowing if they've changed style recently helps too.

If I can't have a taste, I want to know what my BWE buddies who I've drunk from the same bottle with think. I'll read the critics but they play a smaller role in my decision making. That wasn't always the case. I used to put a lot more faith in some of the critics but that's another story.

Points just aren't that reliable at predicting how much I'll like a wine. Maybe they are more reliable with respect to future price appreciation but that's not my thing.

[rthomaspaull]

The calculations are just a rough guide. I go by my own taste buds, but price is also a consideration for most of us.
I think it was an Italian professor of oenology who said something like "There are three types of wine: good, bad, would drink it if someone else paid for it."

[marcs]

Let's use letters instead, David. Richard is saying that if f(y) is the average price of a wine rated y, then f(y+1)/f(y) is approximately the same no matter what y is (except maybe when y is close to 100). In other words, to get a wine A that is one point better than a wine B, you expect the price increase to get A to be a fixed percentage of the price of B.

Without seeing the data, would you have thought that to be the case? If so, would you have guessed that the fixed percentage is about 10.8%?

Yes, there is no reason at all to expect f(y) to be a linear function in the manner described. In fact quite the opposite, given how unequally wealth and income are distributed, f(y) might turn exponential as you get to the extreme high end of quality. That's probably the assumption in luxury goods markets where supply is limited. In Burgundy, which is a much more limited supply market, prices really do go exponential at the high end. (Noting that another issue in this pricing model is that there appears to be no role for supply or product differentiation, just a uniform single-dimensional "quality").

Richard also seems to assume that the wine price-quality relationship f(y) stays constant between time periods like some kind of "law of nature". For example, he quotes Clive Coates in "Bordeaux" saying that in the original 1845 classification "so a lowly fifth would bring 40% of a Chateau Lafite". Obviously this has changed a lot as these days a low-end fifth growth would be more like 5 to 10% of the price of Chateau Lafite! Given the huge expansion in the market for wine that makes sense.

[rthomaspaull]

To marcs and stefan: I do not assume that the price-quality relationship {(f/y)} stays constant. The problem at the high end is that it does not: things seem way out of line at the high end. I do assume that the price-quality relationship was much better years ago, with something like a "value-quality" relationship staying fairly constant.

[rthomaspaull]

Further to my last post, if someone had read my post of Thu Mar 4, 2021 at 1:31 pm carefully (not the easiest thing to do, I must admit) they would have noticed that I said "wines with higher ratings became more and more disproportionately out of line as the wine ratings rose....." i certainly did not say that the price-quality relationship {(f/y)} stayed constant. rthomaspaull

[rthomaspaull]

Does an X symbol always show up beside the latest post on a subject?

[DavidG]

Richard:

No doubt, the increase in price for each additional critic point increases substantially once you get above 95 points. Maybe not if they’re Suckling points. As I noted before, they’re flabby and lack lifting power.

I have no clue abut I he “x”...