NOBODY KNOWS – BUT THE AUCTION HOUSE?

In this thesis we are concerned with estimate bias of online auctioneer Auctionata AG and investigate components influencing bias in our data set of eighteen auctions. Pre-sale estimates are published by most auctioneers and provide the audience with relevant price information in a market where standardized valuation systems are missing. We find that estimates at Auctionata AG are upward biased, but that art of Asian origin and the starting price have a significant positive influence in reducing bias. We further find significant evidence that known provenance and art of deceased artists reduce upward bias in our sample of paintings. Moreover, the order in the catalogue has a significant effect on bias. Contradictory to earlier studies, we do not find evidence for the significance of physical properties and only slight improvement in the accuracy of estimates over a one year period.