Working Paper: NBER ID: w31160
Authors: Jiageng Liu; Igor Makarov; Antoinette Schoar
Abstract: Terra, the third largest cryptocurrency ecosystem after Bitcoin and Ethereum, collapsed in three days in May 2022 and wiped out $50 billion in valuation. At the center of the collapse was a run on a blockchain-based borrowing and lending protocol (Anchor) that promised high yields to its stablecoin (UST) depositors. Using detailed data from the Terra blockchain and trading data from exchanges, we show that the run on Terra was a complex phenomenon that happened across multiple chains and assets. It was unlikely due to concentrated market manipulation by a third party but instead was precipitated by growing concerns about the sustainability of the system. Once a few large holders of UST adjusted their positions on May 7th, 2022, other large traders followed. Blockchain technology allowed investors to monitor each other's actions and amplified the speed of the run. Wealthier and more sophisticated investors were the first to run and experienced much smaller losses. Poorer and less sophisticated investors ran later and had larger losses. The complexity of the system made it difficult even for insiders to understand the buildup of risk. Finally, we draw broader lessons about financial fragility in an environment where a regulatory safety net does not exist, pseudonymous transactions are publicly observable, and market participants are incentivized to monitor the financial health of the system.
Keywords: Terra Luna; cryptocurrency; decentralized finance; financial fragility; algorithmic stablecoin
JEL Codes: E42; E44; G21
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
Cause | Effect |
---|---|
concerns about system sustainability (Q56) | run on Terra (Q49) |
adjustment of positions by large holders of UST (F32) | subsequent withdrawals by other traders (E49) |
blockchain technology facilitates monitoring (E42) | amplifies speed of the run (Y60) |
sophistication of investors (G11) | timing and magnitude of responses to market signals (G14) |
failure in risk management (H12) | crash of Terra (Y70) |
design of Terra ecosystem (Q24) | vulnerabilities exploited during the crisis (H12) |
reliance on algorithmic mechanisms (C69) | vulnerabilities in the Terra ecosystem (D52) |