Talk:List of most expensive paintings

Peer review?
It would never have occurred to me that a peer review was needed. This is an excellent list, well-defined and organised, and extremely popular as a references source, 2000+ views per day. A credit to its lead contributors. A good rule in life is to leave well alone, unless some problem is raised. What would be the problem on this page? Perhaps the inability to use illustrations for some of the paintings, but this is something which can't be fixed (because of the way copyright legislation is interpreted). Macdonald-ross (talk) 10:23, 10 September 2013 (UTC)
 * We publish images of these pictures on their own pages without worrying about copyright legislation - why not here? (or is the question why there?) Bagunceiro (talk) 00:25, 12 February 2015 (UTC)
 * My question exactly... If it's fair use to show them on their respective pages, why not put the same images here?--Pedrojpinto (talk) 09:50, 16 November 2017 (UTC)

$100M for Cézanne painting in 2013
Thought this needed to be added. It was sold in 2013 but not announced until recently. Dagko (talk) 04:18, 28 December 2014 (UTC)

http://www.freep.com/story/entertainment/arts/2014/12/19/ford-cezanne-dia-sale-victoire/20621413/
 * Added. I will try to find and upload picture of painting as well. --Jklamo (talk) 23:03, 28 December 2014 (UTC)

inflation adjustment
Just a question: Is there an accepted wiki policy regarding the adjustment of prices for inflation? I see that it has been done here by calculating the price based on the U.S. Department Of Labor's September 2014 historical consumer price index table. Fair enough, but doesn't this conflict somewhat with WP:Original Research policy? Such a calculation certainly would go against a strict interpretation of OR policy. However, maybe there's a consensus that such a relatively mechanistic calculation if spelled out is acceptable? Curious because I'd like to use something similar at List of transport megaprojects. However, there are a number of factors to consider when updating for inflation, and I'm sure you could reach fairly different conclusions based on different indexes or methods. Any thoughts on this? Peregrine981 (talk) 19:24, 15 April 2015 (UTC)
 * There is by now an inflation template, which calculates inflations for different currencies using annual averages. Using this template the results are practically the same as using the annual averages provided in the US Dept of Labor's Table 24 of March 2015; the latter estimates are on average 0.06% (±0.13% SD) higher, the main difference probably coming from the adjustment for 2015. The monthly averages are a bit fairer, especially for paintings sold early or late in years with higher inflation. At the extremes, Van Gogh's Portrait of Roulin (sold in May 1990), will "lose" 1.9 million current dollars, while his Irises (sold in November 1987) will "gain" $1.6 million when using the template. This will cause the order to change a little bit, but given the increasing number of private sales without an exact price or date and, especially, the ease of getting the inflation adjusted numbers in place, I'll convert the table to using the template. OR worries should dissipate as well. Afasmit (talk) 11:13, 18 May 2015 (UTC)
 * Thanks for the tip! Although in my case I'm not sure we can use the template on capital costs, but it's a good place to start a discussion anyway. thanks again, Peregrine981 (talk) 20:32, 21 May 2015 (UTC)

Yes but it is reported that mona lisa insurance price would be of 620$ milions in 2016 when if you use any other calculator the transformation will be of 770$ milions in 2016 and 820$ milions on 2018 (one even allowed to set the specif month of both the initial price and current one). 150$ milions are quite a big difference don’t you agree? They should change the equation because this is not accurate Eugeniocazzo (talk) 01:41, 27 February 2018 (UTC)

Let's list the joined price for the pair of Rembrandt paintings
Shouldn't we list the pair of Rembrandts as one item, sold for €160m? They are considered to be “like a pair of shoes, never to be separated”, to quote the director of the Rijksmuseum, no less so then, say, either of Bacon's Triptyches. In fact, Bacon's Three Studies of Lucian Freud had been three separated parts until the Italian collector Francesco De Simone Niquesa reassembled them, while Maerten and Oopjen (got to love that name; it's a diminutive of the already rare and archaic Obrecht) never were separated. Afasmit (talk) 23:21, 3 November 2015 (UTC)

Two new paintings
As shown in this article, http://www.bloomberg.com/news/articles/2016-02-18/billionaire-griffin-said-to-pay-500-million-for-two-paintings, Kenneth C. Griffin bought two paintings from David Geffen for a combined $500 million. According to the article, "Griffin bought de Kooning’s 1955 oil on canvas titled “Interchanged” -- also known as “Interchange” -- for about $300 million and Pollock’s 1948 “Number 17A" canvas for about $200 million." The paintings were bought in fall of 2015. I think the de Kooning should be the top above the Gauguin because the Bloomberg article specifically notes that the deal is larger than the Gauguin deal. I would add them myself, but editing that graph is too arcane for me right now. --Dagko (talk) 18:27, 23 February 2016 (UTC)
 * Thanks for the heads up. I'll leave it to Yurik to add it to the graph. The De Kooning painting is now first on the list, though both it and Gauguin's were sold for "about $300M" and the deal was mentioned to be larger because the combined sale was obviously bigger. I don't think the precise amounts invested are ever made public, so we may never know. Afasmit (talk) 20:49, 23 February 2016 (UTC)

Title to reflect exclusion of masterworks?
Can we try to find a title that reflects that paintings like the Mona Lisa are not listed? How about "List of most expensive paintings sold"? --Bod (talk) 19:49, 16 September 2016 (UTC)

No pics?
In the table, some of the entries have -- Please don't insert a picture here; artist died less than 70 years ago and work postdates 1923 (see WP:NFC  and specifically WP:NFLISTS) --. Why? The main entry for the work has an image? Why wouldn't it also be in thumbnail? Bangabandhu (talk) 06:54, 10 December 2016 (UTC)


 * See Image use policy and Public domain. Also, some images have a fair use rationale that under some circumstances allows an otherwise-disallowed image to be used to illustrate an article on the work illustrated, but not in other articles such as this. For example, see File:Photo of Interchanged by Willem de Kooning.jpg. Verbcatcher (talk) 07:20, 10 December 2016 (UTC)


 * I see, thanks. Its not intuitive that thumbnails would not be acceptable but full pics are. Bangabandhu (talk) 22:16, 10 December 2016 (UTC)
 * Ok, it's frustrating but that makes sense. Is there a way to circumvent this by getting other images? Especially for older paintings, apparently no longer subject to copyright?--Pedrojpinto (talk) 10:00, 16 November 2017 (UTC)

List Sorting?
The list has a sorting function that doesn't work due to poorly formatted data. For example, it cannot be sorted by price or date of sale. Someone clever could fix this. 24.236.70.18 (talk) 15:05, 14 April 2017 (UTC)
 * The sorting appears to be working fine for me. I could improve the seller and buyer columns (by surnames; empty cells should sort last), but I doubt there is much interest in sorting those columns. Afasmit (talk) 18:52, 14 April 2017 (UTC)

The list has a sorting function which is not rightly defined. It can be easily detected when anyone wants to find all the Andy Warhol's silkscreens one after the other. Before the sale of the "hot Sage Blue Marilyn" during the auctiom by Christie's in New York City on May 9, 2022, if anyone pressed the ARTIST sort ascending, the list was led by Salvator Mundi, by Da Vince. OK, but when one presses now the sort ascending, instead of taking the reader to Bbacon, because they are placed alphabetiocally, the one which appears is Andy Warhol's Marilyn, when it shoukd lead all Wahols, which are still there, but almost at the end of the list because alphabetocally the W is before the Y and the Z. Many thanbks — Preceding unsigned comment added by 186.77.197.209 (talk) 18:45, 29 September 2022 (UTC)

Missed auction
The reference to the Portrait of Adele Bloch-Bauer II is now out of date. It was sold recently for $150 million. I've not edited a table, so I don't want to break the format. If someone who knows how they work could effect the changes that would be wonderful. All the best - The Bounder (talk) 15:07, 27 April 2017 (UTC)
 * Updated. Thanks for notification. Jklamo (talk) 17:32, 27 April 2017 (UTC)

Inflation calculation (CPI vs GDP)
I note that the text indicates the adjusted values are inflated (implemented with the inflation template) using CPI values instead of GDP. From the notification on the template, CPI should only be used for small values transacted for items like those in the CPI basket. Ought we be using GDP instead? If so, UK or US? Cmprince (talk) 18:59, 16 November 2017 (UTC)
 * Sounds good. The Inflation/US-NGDPPC template postdates this table by some 5 years, otherwise we would probably have used it. The US version is preferable, since nearly all paintings have been sold in dollars. I thought of replacing it quickly to see the difference, but its usage is a bit more involved through a parser function. If someone wants to give it a try and see how it differs, feel free, perhaps initially on a personal sandbox page. Afasmit (talk)
 * NGDPPC series was ill-conceived as an inflation adjustment. So these have by now been replaced by other indexes (based on GDP deflators). I've made the changes to the article. cherkash (talk) 04:41, 13 January 2018 (UTC)
 * I made the long-needed changes to the rankings, since nobody else had done it since the change was made. --Blemby (talk) 03:11, 11 May 2018 (UTC)

Why start in 1987?
Yes, this is (arguably) the beginning of the Big Money Era in the art market. But it would help to put into perspective today's nine-figure prices. In 1961, for example, it was stunning, front-page news when New York's Metropolitan Museum of Art bought a Rembrandt for $2.3 million. — Preceding unsigned comment added by 2601:18E:C400:66F2:4CDF:1B13:CEB2:1F02 (talk) 18:09, 11 December 2017 (UTC)
 * It starts with 1987 because before that time no paintings were sold for more than the necessary cutoff price (which I initially set in 2006 to have a top 20 list). However, you could start a second table with the highest prices paid before 1987. It'll take some research. Afasmit (talk) 23:06, 11 December 2017 (UTC)
 * Alternatively, a list of the single most expensive painting up to that point in time may be more interesting. I seem to remember making one, but can't find it in the history.Afasmit (talk) 23:09, 11 December 2017 (UTC)

Inflation of Mona Lisa
The automatic inflation adjuster says that the current price of the Mona lisa assurance is 620$ milions but any other inflation calculator says that the actual difference from 100$ milions in december 1962 would be of 770$ milions in 2016 and 820$ milion in January 2018. 150$ milions of difference are quite a lot, can anyone correct the error? Eugeniocazzo (talk) 01:47, 27 February 2018 (UTC)
 * This has to do with the switch from CPI- to GDP-based inflation, which was suggested by user:Cmprince, very briefly discussed above and then implemented by user:Cherkash. In general, the latter has been slower than the first, which is the default inflation used nearly everywhere else. The explosion in high-end spending over the last 40 years appears exacerbated with the GDP-correction. Unfortunately, the list is now partially out of order, and the "rank at sale" column conflicts as well. Blissfully ignorant of economic matters, I don't know if the switch really was the right move, e.g. in light of the CPI inflation numbers used nearly everywhere else. Can anyone provide a good source to defend one or the other? Some expert must have discussed the appropriate inflation rate to use for art sales somewhere on the web. Afasmit (talk) 18:35, 27 February 2018 (UTC)
 * The CPI-based adjustments only make sense for smaller-size expenditures, something akin to the consumer-level spending (on which the CPI basket-creation methodology is based). For anything large, GDP-based adjustment is generally more appropriate. cherkash (talk) 14:00, 28 February 2018 (UTC)
 * Since this has such an impact and CPI-based inflation seems to be the default at the rare occasions inflation is discussed in articles on art sales, I think a relevant reference, preferably a few, is called for. Somewhere, someone must have discussed this in the professional literature. Afasmit (talk) 17:42, 11 May 2018 (UTC)

cut-off
Thanks User:Blemby for re-arranging the list. But during that process cut-off was also made and I think that needs prior discussion. Current $68 line is arbitrary, I prefer $65 line. I do not share the view that the list "is known to be incomplete at the lower ranges".--Jklamo (talk) 09:45, 11 May 2018 (UTC)
 * Great idea. I think the $65 line is actually better too. I am just busy now, i think in about 12 hours I will be able to add the rest up to $65 million. --Blemby (talk) 12:09, 11 May 2018 (UTC)
 * There is no "round number" to do a cut off, as next year, the round number won't be round anymore. We've been keeping this list complete to anything that is, with the CPI adjustment, more expensive than the Pontormo since 2006, with later small adjustments to be able to add the Suprematist Composition in 2008, the Zhichuan Resettlement in 2011 and the Manet in 2014. With the new scheme (which I still like to see some referenced support for), the painting originally forming the cutoff, has fallen off the list (a mere $60.9 M; you and I can afford it). I think you can assume that anything over $67 M was already on the list. It would be nice to keep the Pontormo as a cutoff, as it was the only Old Master on the list sold before 2002, but that would take some work. Even for a cutoff as low as $65 you might have to do some serious digging on line (and calculating) to be sure that the list is complete. Also, it doesn't look like you've redone the "rank at sale" column. I'm sure those numbers have changed as well. Afasmit (talk) 18:49, 11 May 2018 (UTC)
 * Great idea. I will do a softer cutoff that keeps the other paintings in the source as before. Also, I will update the rankings at the very end. --Blemby (talk) 23:29, 11 May 2018 (UTC)
 * I am not sure if the rankings are necessary. They don't really account for the private sales we don't know about, or if a painting was sold at a lower price, then a higher one. --Blemby (talk) 23:58, 17 May 2018 (UTC)
 * The ranking seems interesting enough; for example, it points out that a painting like the Pontormo on the bottom of the list once was the 5th most expensive painting sold. Calculating the rank may be a bit tricky at times indeed. If a private sale come to light at a later time though, the ranking at the time of sale can be adjusted; if they remain a secret, they might as ell not be acknowledged. Sometimes the dates of private sales remain a secret, and a guess needs to be made. Paintings like La Rêve and Suprematist Composition that sold more than once for a price making it in the list do cause problems as well. To keep track, it'll be good to mention such earlier (or later, if it so happens) sale in a footnote.Afasmit (talk) 04:59, 18 May 2018 (UTC)

Participate in the deletion discussion at the. Community Tech bot (talk) 18:51, 12 July 2018 (UTC)-->

Inflation again
I still have strong doubts about using the GDP deflator for the adjustment of the prices. MeasuringWorth is given as a reference for what the GDP deflator is. If you plug in a value and years (say Van Gogh's $39.7 M in 1987 and 2016) on their page a large number of answers is returned. By far the weakest inflation is given by the GDP deflator ($73.7M, used on the page since January). The site recommends to use this to calculate the the "real cost" of a project (amongst which they count things like constructing a canal, financing a war, net worth of a company). Three other cost calculations for a project are all much higher in 2016 dollars. It appears to me that buying/selling a painting is not really a project, but more an exchange of a commodity or, even better, wealth (in which they include e.g. bank deposit, stock portfolio, or real estate). The lowest value given for both of these ($83.9) is the one using CPI ("real wealth"), which is also the method used by any article I know that bothers to calculate the inflation-adjusted price of art. Other adjustments (e.g. household purchasing power or relative wealth) give even higher numbers ($93.2 and $115M, respectively). Check out their tutorial for further help. My guess is that the CPI calculation is more appropriate (and much more standard, updated faster, and easier available) and that it still probably underestimates the proper inflation. With the GDP deflator method recent auction prices appear too impressive. Unless someone shows convincing references to keep the GDP deflator, let's turn back to our old method. Afasmit (talk) 17:55, 4 September 2018 (UTC)
 * I agree with switching to "old method", I share the view that in the case of this list CPI is more appropriate than GDP deflator.--Jklamo (talk) 07:47, 5 September 2018 (UTC)
 * I agree too. Even if that means that all that laborious re-ordering I did was in vain :) --Blemby (talk) 22:15, 5 September 2018 (UTC)
 * I just went ahead and done it. Please check for mistakes, including in the intro. Afasmit (talk) 01:51, 4 October 2018 (UTC)

Progressive records
As per this article;

MBG02 (talk) 20:30, 2 January 2019 (UTC)

Lichtenstein's "Masterpiece"?
Don't want to upset any boats, but I just learned about Roy Lichtenstein's "Masterpiece" selling for $165M and was wondering whether it should be included on this page. &mdash;  X   S   G   02:26, 28 October 2020 (UTC)


 * On the topic of Lichtenstein, Nurse also sold at a high enough price to warrant inclusion on this list.

No mention of tax avoision or money laundering?
https://www.natlawreview.com/article/art-and-money-laundering

Both have been a significant driver of prices for many years.96.240.128.124 (talk) 03:49, 30 March 2021 (UTC)

De Vaandeldrager Rembrandt
The Vaandeldrager should be added to the list since it was sold for 175 million euros. 84.104.0.52 (talk) 17:30, 6 May 2022 (UTC)

Original research in use of inflation index
I came here from the featured list nomination of List of most expensive books and manuscripts, where I raised concerns regarding original research and the use of the template to adjust prices for ranking purposes. I can see there's already been discussion here over use of the template, but unfortunately none of the indexes available with the inflation template are appropriate for art prices. Paintings are not a consumer goods, nor can they be indexed in the way that an asset (capital good) like property can be, the values are notoriously fluid and require specific asset price indexing against other similar works across similar periods. To avoid the original research problem, the list would be better structured around sales above a certain point (eg US$50 million) as List of most expensive photographs does. Regards, Goldsztajn (talk) 02:04, 10 August 2022 (UTC)


 * Pinging users from last conversation on this: @Afasmit, @Jklamo, @Blemby. Regards, Goldsztajn (talk) 02:47, 10 August 2022 (UTC)
 * I do not consider the use of CPI to be problematic, as the use of an inflation template is intended to show the impact of "general" inflation. Sorting the list by price at the time of sale would be much more misleading (comparing 1987 and 2022 prices for example). But feel free to extend the lead paragraph of the list about the drawbacks of the chosen CPI solution.--Jklamo (talk) 08:17, 10 August 2022 (UTC)
 * Hi @Jklamo thank you for replying. "as the use of an inflation template is intended to show the impact of "general" inflation." At the very least, we need reliable sources that validate such an approach to asset price inflation in rare art.  The only other approach is sourcing which gives estimates of the contemporary values of each of these items. Otherwise, this is original research. Regards, Goldsztajn (talk) 23:24, 10 August 2022 (UTC)