Do photographs burn?

Although art prices and art returns have been long studied, there is still work to be done in gaining more insights in the factors that might play a role in art price formation. One of the factors that is commonly believed to have a negative effect on art prices, is the event of a failure in auction, also known as the burning effect. Little is known about the burning effect or its scope, as it has only been explored by a handful of researchers who mainly focused on the painting medium. This shortfall in knowledge on whether items burn after having failed, is being reflected in the precarious approach of the topic. Its very existence has been doubted, or is even thought of as a myth. Nevertheless, the relevance of gaining knowledge on this topic is of growing importance, as a new group of investment-focused art collectors is emerging. In addition, the magnitude of the burning effect has direct implications for sellers and auction houses in setting reserve prices. Therefore, this research aimed to explore the existence and magnitude of the burning effect. The following research question was the main focus: ‘To what extent does the failure to find a buyer and thus the failure to sell affect the returns for photographs in public art auctions?’. This question was answered using a data set of repeat auction sales containing photographs of the Dusseldorf School of Photography, which provided the opportunity to expand the knowledge on the burning effect to another medium than paintings. The data set contained 529 “sold, sold” observations, of which 237 included one or two fails in between the sold events. For the statistical analysis, a pooled OLS repeat sales regression was carried out. Photographs that failed once or twice in between two successful auction sales were found to return significantly less than photographs that did not fail. Thus, the burning effect had a magnitude of -11 percent (p<.05). Photographs that did not fail in between two successful sales were found to return 11 percent more (p<.05). These results are in line with earlier findings of researches on similar topics. Based on the results, the conclusion suggests that the burning effect is existent for photographs and could thus be taken into account in predicting art prices and returns on art.