The emergence of the blockchain technology has enabled innovations such as cryptocurrencies and decentralized finance (DeFi) applications, raising new opportunities and challenges for investors. This thesis contributes to the understanding of these emerging markets by examining three of their key elements: non-fungible tokens (NFTs), decentralized exchanges (DEXs), and stablecoins.

The thesis consists of three chapters. In the first chapter, "Economic Incentives in the Digital Art Market", I study non-fungible tokens (NFTs) - assets on a blockchain that represent ownership of digital art and are traded on NFT marketplaces. The NFT market on the Ethereum blockchain was monopolistic until the end of 2022, when a new marketplace entered and captured a significant market share. I collect transactions from these marketplaces to study the effects of increased competition on the incumbent marketplace, artists, and investors. While competition had positive effects by reducing transaction costs, it decreased the profits of artists, discouraged them from creating new artwork, and thereby reduced the supply of new assets. I also study user migration, multi-homing behavior, and market segmentation, and compares the results with the predictions from theories of platform economics.

The second chapter, "Decentralized Exchanges (DEXs) with Low Price Impact", co-authored with W. Huang and Z. Song, examines the impact of DEXs on price stability and transaction costs, focusing on assets with stable values. We develop a model that demonstrates that the introduction of a DEX can reduce price variation and transaction costs, by lowering the price impact of large trades. We validate the model's predictions by collecting and analyzing trade-level data on stablecoins from both centralized and decentralized exchanges such as Binance, Curve and Uniswap.

The third chapter, "How Algorithmic Stablecoins Fail'', co-authored with G. Kurovskiy, examines one of the largest collapses in cryptocurrency markets - the collapse of the 18.7-billion USD algorithmic stablecoin TerraUSD (UST) and its 20-billion USD backing token, Luna, in May 2022. We collect and analyze transaction-level data from the Terra blockchain and cryptocurrency exchanges, and argue that several flaws in UST's design impeded its price stabilization. Using a simple model, we demonstrate that a combination of these design features helps explain the data patterns observed during the crash.