Pay, Stay or Delay? How to Settle a Run

Abstract

We study how the trade-off between liquidity provision and value preservation shapes redemption rules for financial intermediaries. In the literature, a bank prioritizes liquidity provision in a run by selling all assets. In reality, default followed by a mandatory stay on payments is triggered once a bank runs out of liquidity, removing withdrawal queue priority and reducing run incentives. Orderly resolution avoids fire sales, but reduces liquidity provision so strict sequential service dominates when assets are highly liquid (e.g., Treasury MMFs). We show that run frequency under sequential service may be nonmonotonic in asset liquidity. For high enough asset liquidity and sufficiently small liquidity benefits, a payment stay with immediate asset sale and payout is best, rationalizing new rules on gates on Prime MMFs. Our results explain the asset allocation and investor sorting across commercial, narrow, and shadow banks.