Tan, YongLiu, Ye2025-08-012025-08-012025Liu_washington_0250E_28165.pdfhttps://hdl.handle.net/1773/53433Thesis (Ph.D.)--University of Washington, 2025Networks are increasingly central to how individuals make investment decisions, how prices are formed, and how platforms design user experiences in the context of modern financial technologies (FinTech). Rather than viewing market participants in isolation, this dissertation examines how surrounding social structures shape economic outcomes across three FinTech applications: social trading, cryptocurrency, and non-fungible tokens (NFTs). Each essay highlights how network position, peer behavior, and relational dynamics can influence attention, adoption, and valuation — often in ways that rival or exceed the impact of individual intrinsic characteristics. First, rather than focusing solely on individual attributes such as financial performance, I examine how strategy similarity among traders affects their popularity. I construct dynamic similarity networks and use a coevolution model to analyze how traders’ positions within these networks shape follower growth. I find that as more peers adopt similar strategies to a focal trader, that trader tends to attract fewer followers over time. This substitution effect is particularly strong for high-performing traders. Counterfactual simulations suggest that promoting a more even distribution of followers and reducing network density can mitigate substitution and increase overall platform engagement. This suggests that platform design should consider not only individual performance but also the network positioning of users to foster diversity and sustain long-term engagement. Second, I investigate how social learning drives cryptocurrency adoption. Using a large dataset from a social trading platform, I analyze how investment opinions and behaviors of network neighbors influence an individual’s decision to trade Bitcoin. The results show that people are more likely to adopt Bitcoin when their peers’ actions align with their stated views. Under higher uncertainty, individuals rely more on optimistic opinions, even if behaviors are not consistent. This suggests that social media plays a dual role — as a source of behavioral signals and emotional reinforcement — particularly when individuals face ambiguity. It also highlights the importance of credibility and alignment between opinion and action in peer influence. Third, I study the role of transactional networks in shaping NFT value. Drawing on data from a leading NFT platform, I construct monthly networks linking NFTs based on shared traders and utilize a permutation-invariance-based method to circumvent the dimensionality issue of network data. The results show that incorporating transactional network information substantially improves NFT price prediction, beyond what image or metadata features can explain. This chapter highlights the importance of social dynamics in digital asset valuation and offers implications for both pricing models and platform design. Together, these chapters demonstrate that financial decision-making and asset valuation cannot be fully understood without accounting for the social context in which they occur. Whether through imitation and substitution among traders, peer-driven learning under uncertainty, or value formation via transactional ties, networks serve not just as conduits of information but as fundamental structures shaping market outcomes. By integrating structural modeling, behavioral analysis, and machine learning on large-scale digital traces, this dissertation offers new insights into how financial technologies operate when embedded in social systems. These findings contribute to emerging conversations at the intersection of finance, platforms, and computational social science, and underscore the need for theories and tools that reflect the relational nature of modern digital markets.application/pdfen-USnoneBusiness administrationBusiness administrationThesis