This paper explores the role of bank deposits in providing liquidity and the mechanisms that prevent bank runs. It argues that traditional demand deposit contracts, which offer liquidity, have multiple equilibria, one of which is a bank run. Bank runs cause significant economic damage by disrupting production and reducing risk sharing. The paper analyzes optimal bank contracts that can prevent runs, finding that suspension of convertibility (allowing banks to stop withdrawals when they are too numerous) can achieve optimal risk sharing but is not feasible in general due to the sequential service constraint. Government deposit insurance is shown to be superior, as it allows banks to follow a desirable asset liquidation policy and prevent runs without reducing the ability to transform illiquid assets into liquid liabilities. The paper concludes with policy implications, emphasizing the importance of government intervention in preventing bank runs and ensuring stable financial systems.This paper explores the role of bank deposits in providing liquidity and the mechanisms that prevent bank runs. It argues that traditional demand deposit contracts, which offer liquidity, have multiple equilibria, one of which is a bank run. Bank runs cause significant economic damage by disrupting production and reducing risk sharing. The paper analyzes optimal bank contracts that can prevent runs, finding that suspension of convertibility (allowing banks to stop withdrawals when they are too numerous) can achieve optimal risk sharing but is not feasible in general due to the sequential service constraint. Government deposit insurance is shown to be superior, as it allows banks to follow a desirable asset liquidation policy and prevent runs without reducing the ability to transform illiquid assets into liquid liabilities. The paper concludes with policy implications, emphasizing the importance of government intervention in preventing bank runs and ensuring stable financial systems.