What Is A Painting Worth?
Many would agree that a piece of art is substantively different from most other everyday objects, such as a mug, a phone or a car. Art, unlike many consumer products, is not produced on an assembly line, in large quantities where each piece is an exact copy of another. As such, art is not bought, sold or traded at a fixed price that is established daily by the law of demand and supply.
Instead, every piece of art is, in principle, unique and may only sell once, or at best, twice in a decade. This is why excitement in the art world occurs when, after many years, a masterpiece resurfaces and is available for sale. At the same time, this quality makes it very difficult to evaluate the fair market price of an art piece, since art often has little to no intrinsic value. Typically, an expert appraiser is brought on to evaluate and estimate the value of a given artwork. However, in many cases, this can be costlier and more time consuming than necessary when we take into account the degree of precision required, or the real value of the work in question.
In such cases, it is often possible and smarter to use data and statistical methods to obtain a quick and relatively inexpensive estimate of the likely sales price. A popular method used by academics and companies in the art market is called hedonic regression, or hedonic pricing method. First introduced in 1922, this tool was originally used in the real estate market to appraise the value of a property. A hedonic regression uses data points to estimate the extent to which several factors affect the price of a product. Precisely because extensive data sets with observations on recorded sales are available in the art world, a hedonic regression can be applied. This is because, whilst all works are unique, each have common physical and non-physical characteristics which make it possible to infer the value of the artwork.
Such characteristics include the artist, material(s) used, the size, the period of production, the origin, as well as subjective features like the quality or reputation of the artist. In fact, exclusive characteristics such as composition or desirable subject matter have proven to play a particularly important role in the valuation of art.
Mathematically, a hedonic regression takes the price of all recorded sales of all artworks at a given time. It then regresses them against their individual common characteristics and subsequently compares them with the artwork in question to estimate its market value. This sounds very arcane, but can easily be explained using the following example which highlights the mechanism of the hedonic regression method:
Imagine you want to determine the price of a painting you would like to sell. Because you own the painting, you have all the information on its specific qualities such as size, provenance, material(s) used, and/or time of production. Additionally, imagine you have an empty room and an art storage unit that contains all artworks that have ever been sold, and which sales have been publicly recorded. To determine the value of your painting you sort through the storage and select all paintings that have similar characteristics to yours. You then move them from the storage unit into the empty room.
Let’s say you start with all works that have the same material as your painting. Suddenly your room is filled with paintings. You then decide to only select paintings with the same size and from the same period. The number of pictures in the room gets smaller and smaller. You continue until the only pictures left in the room are paintings that have roughly the same size as your painting, and are also similar to your painting in most other characteristics. Once you have filled the room, the only thing left to do is use the recorded sales prices of the paintings and adjust for inflation.
Thus, the logic being, if you selected the right painting from the art storage, then the value of your painting should be similar to the average adjusted sale price of the paintings that embellish the walls of your now not so empty room.
Let’s recap what we just did: The example of selecting artworks from a storage unit, moving them to a room and using the price of each to determine the fair value of your artwork represents the inherent function of the hedonic regression. The art storage represents your data. The more artworks in your imaginary art storage, the bigger the data set. The more detailed information about the characteristics of your artworks in the art storage, the better the quality of your data, which leads to a more precise estimate of the value of your artwork.
Some considerations demand attention, however. Firstly, we would not want the room to be too crowded with artwork. To obtain a good estimate, we would need to be as specific as possible when choosing the characteristics of the paintings. At the same time, we need enough paintings in the room for we might otherwise get an imprecise estimate of our painting’s value. Furthermore, we have to acknowledge that some characteristics are more relevant because they have a greater impact on the price. The artist and his/her reputation are by far the most important factors affecting the price of an artwork for example. Also, in the instance of an artist for whom many recorded sale prices exist, our hedonic regression should deliver more accurate results. Lastly, it is important to reiterate that each work of art is unique, which means that we cannot rely blindly on our hedonic regression. Besides, we have to understand how our painting compares to other paintings in the room, and adjust our estimates accordingly.
In spite of these considerations, a hedonic regression should, in most cases, provide a reasonable measure of the value of an art piece, and in many instances, offer a faster and cheaper alternative to an expert consultation. Expert appraisers often employ hedonic regressions as a solid basis for their evaluation. If you are interested in using hedonic regressions, there are several companies that provide online services in a user-friendly way.
References
Ashenfelter, O., & Graddy, K. (2003). Auctions and the Price of Art. Journal of Economic Literature, 2003, vol. 41, issue 3, 763–787.
Bakhouce, A., & Thebault, L. P. (2011). What Determines Cézanne’s Art Pricing? A Hedonic Regression Method. Analele Stiintifice ale Universitatii “Alexandru Ioan Cuza” din Iasi — Stiinte Economice, 2011, vol. 58, 515–532.
Gailbraith, J. W. (2018). Econometric Fine Art Valuation by Combining Hedonic and Repeat-Sales Information. Econometrics 2018, 6, 32.
Witkowska, D. (2014). An Application of Hedonic Regression to Evaluate Prices of Polish Paintings. International Advances in Economic Research, Volume 20, Issue 3, 281–293.
