A thesis on digital art market

Abstract

Digital art has been increasingly popular with the development of blockchain technology. Blockchain technology provides a form of protection for digital art sellers and buyers as it enables verifiable and tamper-proof records of ownership of digital artworks. In addition, it also records the information and transaction history of the artworks and allows public access to the information with exceptional transparency. Hence, it not only offers digital artists a new format to sell and promote their works but also allows researchers to observe the market activities. Thus, in this thesis, I focus on the subset of digital arts with information and transaction histories recorded on the blockchain. A digital artwork/collectible (e.g., image, video, music) recorded on the blockchain is also called a non-fungible token (NFT). NFTs are becoming an increasingly popular form of art investment, with rocketed market growth in 2021. However, both consumers and sellers face uncertainty in the market activities. Consumers encounter uncertainty when making purchase decisions as there are no established valuation guidelines for NFTs. Sellers confront uncertainty when creating artwork owing to the lack of knowledge about consumer demand. Thus, this thesis investigates two different topics that surround uncertainty in the NFT market. Chapter 1 examines herding behavior in NFT auctions and the moderating role of visual complexity and familiarity. Auction is one of the most popular selling formats in the NFT market. Given the valuation uncertainty for NFTs, consumers’ bidding tendencies would be influenced by other collectors’ bidding decisions (herding behavior). I build a model on bid continuation probability to capture the herding behavior in the primary auctions. In addition, the visual cues are essential in evaluating the artwork. In this regard, I examine how visual complexity and familiarity affect herding behavior, as these two visual characteristics usually affect both the ease of understanding and recognizing the visual stimulus. I use a computer vision technique to quantify the visual complexity of an NFT, and apply transfer learning to measure its visual familiarity. The findings show that there is herding behavior in NFT auctions, and visual complexity and familiarity both exhibit a U-shape moderating effect on herding tendency. This research further tests the subsequent impact of herding behavior in the primary auctions on the auction winners’ resale decisions. Results suggest that auction winners are more willing to resale the NFT and set a higher price premium when there is stronger herding behavior in the primary auctions. This research contributes to herding, NFT, and creative content production literature, and provides managerial implications for content creators and market platforms to promote sales. Chapter 2 investigates how sellers can enhance revenue and benefit from introducing consumer uncertainty in product assignment (i.e., probabilistic selling). Probabilistic selling has become an increasingly popular and relevant phenomenon across industries, from traditional online travel agencies (e.g., probabilistic hotel rooms), retailing (e.g., clothing, and cosmetics subscription boxes) to digital gaming (e.g., in-game products), non-fungible tokens (NFTs), etc. However, despite the market prevalence, there is no empirical investigation into the actual benefit of marketers employing probabilistic selling to the best of our knowledge. Leveraging a feature change in an NFT marketplace, this research quantifies the effect of adopting probabilistic selling on sellers’ monthly revenue using three estimators (i.e., two-way fixed effect DiD, staggered DiD, and aggregate synthetic DiD). Results consistently show that probabilistic selling increases adopters’ monthly revenue within the range of $500.455 to $665.869. In addition, this research finds three underlying mechanisms that drive the revenue lift: protection provision on the transparent market, increased market thickness and matches, and enhanced market matching efficiency. At the same time, alternative mechanisms are discussed and ruled out. These findings add to the probabilistic selling, uncertain marketing and NFT literature. This research also provides managerial implications for sellers to increase sales and for market platforms to promote market participation and enhance market matching efficiency.