Our Approach to Art Market Analysis
Art Market Data Collection
Our art market database consists currently of around 1'500'000 art market transactions from auction houses and private sales. The vast majority of data is collected from approximately 1'000 auction houses around the world – some of them are listed in the Knowledge section. The private art market transactions originate from galleries, private museums, insurance companies, and art investment funds. Our staff checks each transaction and adds additional information to cover all price determinants of the art objects, before they pass them on to the next step of data coding.
If an artist can not be found in the artist database, we count on our network to get the necessary transactions to calculate individual art market index for our members.
Art Market Data Coding
The art market transactions which we have collected from our network can not be used in their initial form. The sales records must contain all information on the art price determinants of our specific hedonic regression models. Thus, all transactions have to be transformed into a code which can subsequently be used in statistic tools, to calculate the various art market indices for art categories and artists.
Some variables from our hedonic models are published in the Knowledge section. Our findings show for example, that art techniques and painting supports can be important painting price determinants for some artists. For other artists however, they are not significant at all. Data coding requires a lot of expertise and experience in the fields of art pricing and painting characteristics. As a result, our staff working in this department is also involved in art appraisals and art valuations.
Art Market Index Modeling
The fundamental idea of hedonic regression is to measure the influence that the various characteristics of an art object (e.g. size, technique, condition of a painting, etc.) have on its price – the art price determinants. Initially, our basic model has been developed by one of AFP's founding partners in order to calculate the financial performance of art investment fund portfolios.
In the meantime, many different models are used, since the variables to be used depend on the individual artists' art price determinants included in the art market category indices. Thus, all models applied are tailored to the respective artists to provide the best explanatory power possible. Furthermore, various statistical methods are used to find the best performance estimate for each art market index.
Art Investment Advisory
There is broad consensus in the financial and art advisory community that hedonic art market indices provide the most accurate estimates of art investment returns, risks, and correalations. As a result, our art indices and individual art market research reports can be used for supporting financial decisions on art investments in general, for wealth tax fillings, or art insurance purposes. Moreover, we provide art appraisals and art valuations based on the same logic.
AFP is the first provider of hedonic art market indices for professional art investment research. Based on the explanations on our approach, it becomes clear that we offer high quality art index, which will ultimately increase transparency in art markets. However, art index remain best estimates, and the true monetary value of an art object is determined only, when it is sold in the art market.