We show that the infrastructure of decentralized exchanges subjects liquidity providers (LPs) to a “tragedy of the commons.” Liquidity providers lack incentives to exit liquidity pools to prevent arbitrage losses, as these are collectively shared, while withdrawal costs are borne individually. Arbitrage rents primarily flow to validators as infrastructure fees, with the median arbitrageur transferring 96% of profits. Proposed solutions—speed technology, alternative sequencing rules, and flexible pricing curves—fail to reduce these rents. Leveraging the Silicon Valley Bank crisis as a natural experiment, we show that LPs defensively adopt more convex pricing curves to mitigate adverse selection risks.

Decentralized exchanges (DEXs) allow traders to express their willingness to pay for quick execution through a public priority fee bidding mechanism. We provide evidence that high-fee DEX trades are more informative and contribute more to price discovery. Using address-level blockchain transaction data, we show that informed traders persistently bid higher fees to secure early execution, revealing a strong willingness to pay for execution priority. Further, analysis of Ethereum mempool data demonstrates that informed traders employ a “jump bidding” strategy, placing high initial bids to deter potential competitors.

The blockchain settlement layer facilitates systematic frontrunning, resulting in inefficient block-space allocation. Private transaction pools can reduce these inefficiencies and enhance welfare. However, full adoption is limited by misaligned incentives between users and validators. Validators are reluctant to forgo rents they earn from frontrunning – referred to as maximal extractable value – leading to a partial adoption equilibrium in which frontrunning persists. Our empirical analysis of Ethereum’s Flashbots private pool supports these findings: validators earn higher revenues, users facing greater frontrunning risk are more likely to use the private pool, and attackers’ cost-to-revenue ratios in private pools converge to one.

We study trading and risk management decisions of banks in over-the-counter markets, accounting for 2 types of risk: payoff risk from loans and counterparty risk from trading activities. Our model provides empirically supported predictions on the structure of the interbank credit default swap (CDS) market: i) banks with high default probabilities either buy or sell CDS contracts; ii) because of the counterparty risk friction, payoff risk is only partially shared; and iii) safe banks act as intermediaries and help diversify counterparty risk. Banks manage their default probabilities to become creditworthy counterparties, but they do so in a socially inefficient way.

This paper endogenizes intervention in financial crises as the strategic negotiation between a regulator and creditors of distressed banks. Incentives for banks to contribute to a voluntary bail-in arise from their exposure to financial contagion. In equilibrium, a bail-in is possible only if the regulator’s threat to not bail out insolvent banks is credible. Contrary to models without intervention or with government bailouts only, sparse networks enhance welfare for two main reasons: they improve the credibility of the regulator’s no-bailout threat for large shocks, and they reduce free-riding incentives among bail-in contributors when the threat is credible.