The regression model of the art market in Poland

Aim: The aim of the article was to study the impact of the functional form of the regression model, i.e. the set of explanatory variables, on the rates of return. The assessment of the investment effectiveness in the Polish art market required building a regression model to determine price indices and rates of return. The knowledge of price-forming factors can be used to build a hedonic regression model that maps the dependence of painting prices and their characteristics. Methodology: The purpose of the article was achieved by analysing the sales records from Polish auction houses, for the most sold painters in the period 2007–2022. On this sample of 7,442 records, the parameters of nine hedonic regression models were estimated with the LSM using many combinations of explanatory variables. Results: The research results indicate that the form of the regression model is crucial for the value of indices and affects the rates of return with differences in examined models ranging from 3.71 p.p. to 13.71 p.p. Although the selection of explanatory variables affects the value of the indices and the rates of return, the variability of the average index values remained within the limits of a strong and positive correlation between the individual studies. Implications and recommendations: The article expands the knowledge on the assessment of investment effectiveness in the Polish art market. Recommendations for future research include searching for new price-forming factors and incorporating them into the regression model. Originality/value: The study of hedonic regression on the Polish market is unique due to the lack of access to a database on the sales of paintings on the Polish auction market.