Honestly, I think Christie's has an interest in their estimates being semi-accurate, but not _too_ accurate. They obviously want to be within an order of magnitude so they don't look stupid, but overestimating might drive up the price and make them money. Also, when things sell for double or triple the estimate, it makes the news, with Christie's name right beside.
Someone else said Sotheby's bought a database of art prices for 7 figures - why wouldn't a bunch of headlines with their name be worth a few million to them?
Additionally I think they have an interest in prices being as high as possible normally, with estimates high to encourage that. So the pattern would look like:
Antique Japanese woodcarving, $10MM predicted -> $12MM public predicted
Set of documents from Roman times, $5MM predicted -> $6MM public predicted
Van Gogh paintings (aka more headline potential), $10MM predicted -> $3MM public predicted
Then you get the headlines "Van Gogh paintings fetch more than triple the estimate at Christie's auction", which is free PR for Christie's.
If you cannot, I would be shocked. Price estimates at auction houses are not their best faith effort at estimating sales prices. They are much more complex, and involve to some degree, an attempt to manipulate the price as high as possible.
It seems to me like a back-justification of something not measurable analitically.
A 5 millions US$ bracket is not in any way an approximation.
IF you had narrowed the interval (and IF the picture is at the auction sold within it), then all your data mingling exercises might have some usefulness, otherwise it is just an arbitrary estimation like the one from Christie's.
So the similarity calculated by an arbitrarily chosen software between (photos of) study and painting of improvisation #8 is 64%.
The similarity between (photos of) study and painting of improvisation #20 is a MUCH lower 57%.
It is clear how this HUGE 7% different can easily bring down the value of the estimation by more than 50% from the improvisation #8 sale price of 23 millions for a 98x70=6,860 sqcm painting, i.e. 3,350 US$/sqcm to a mere 10 millions for the improvisation #20, a 71x99=7,029 sqcm painting, i.e. 1,423 US$/sqcm, give or take a couple of millions.
Thanks for reading, I hope to improve the analysis over time. I spent the last three years building a unique database that contains the complete paintings of ~40 of the best known artists of the twentieth century. I was motivated by reading about the large percentage of works that are forged or misattributed. I had always just assumed there was a single database listing how many works artists created, average size of the works, changes in materials over time etc. (similar to sports stats). After asking around and learning no such database exists I decided to build one. Getting my blog out there has allowed me to meet with a lot of really sharp folks like yourself who are helping me refine the analysis.
>Thanks for reading, I hope to improve the analysis over time.
Good, but I don't see the analysis at all.
I mean you are betting (without any actual money which is good) on the actual auction price of something in the future.
The auction price of a work of art - by definition - is not in any way like the actual market value of anything (it will be MUCH lower or MUCH higher or JUST ABOUT right) depending on a zillion factors that have nothing to do with the actual value of the piece.
And you have so few data points that a statistical approach is mostly meaningless, let alone one based (in the case of a study) based on the differences between it and the final painting.
You could use a few sessions of flippism and get a similar value.
Or count the number of unique RGB colours in the (photos of) the paintings and give each one a value between 1,217 and 23,146 US$ depending on brightness , later multiplied by the square root of the year it was painted and arrive to a same "mathematical" estimation.
Not sure what you're exactly building here, is it for fun or profit or both?
I heard when Sotheby's bought the Mei Moses Database/Index last year, the valuation they paid for it was in the 7 figures. This type of data analysis is on people's minds more and more and potentially quite lucrative, good luck.
Thanks for reading my post. I am building a database of complete known works across the most important art and artists of the twentieth century. Currently no such database exists. Hoping to work with folks much sharper than I am to build out meaningful analytics around art and artists using this new data.
Artnet has been doing the same basic analysis (finding comparable artworks and calculating the worth of a piece of art) for over a decade. They offer theirs as an add-on to their price subscription service: https://www.artnet.com/analytics-reports/
Thanks for reading, Artnet is awesome. To my knowledge they do not have a database looking at all known works by artists, they focus on works sold at auction. This makes sense as they cover much larger group of artists and info on complete works is typically not available for the majority of artists.
Yes – I think the point of sticking just to open auctions is so that the prices and indexes they have represent what the market thinks of an artist and particular works.
This is awesome - really eye opening. This seems like a good proof point of the inefficiency of the art market (partially due to lots of missing data, partially due to the relatively small sample size of high-end art sales). Excited to see where this goes, and how it impacts the efficiency of the market.
Part of the art market's appeal is it's opacity, though - there's plenty of manipulation that would be outright illegal in regulated markets, and plenty of dirty money being washed through it - all contributing to the exploding valuations for fine art in recent decades. I wonder how much transparency and efficiency the art world actually wants, and what is the optimal amount to maximize art valuations.
What about guarantees? That is a large component of auction price estimates. A lot of these big name paintings are sold with the house guaranteeing a price to the seller. Can't have a full understanding of the industry without understanding that component.
I'd like to open source at least a portion of the data to see if it is possible to build a community driven database. Could connect works back to museums and auction records as well as articles and research and potentially build a crowd sourced provenance.
Someone else said Sotheby's bought a database of art prices for 7 figures - why wouldn't a bunch of headlines with their name be worth a few million to them?
Additionally I think they have an interest in prices being as high as possible normally, with estimates high to encourage that. So the pattern would look like:
Antique Japanese woodcarving, $10MM predicted -> $12MM public predicted
Set of documents from Roman times, $5MM predicted -> $6MM public predicted
Van Gogh paintings (aka more headline potential), $10MM predicted -> $3MM public predicted
Then you get the headlines "Van Gogh paintings fetch more than triple the estimate at Christie's auction", which is free PR for Christie's.