---
title: "Bailing Out (Firms') Uninsured Deposits: A Quantitative Analysis"
authors: "N. Aaron Pancost; Roberto Robatto"
source: "Federal Reserve Bank of Atlanta (Working Paper 2026-16)"
source_url: "https://www.atlantafed.org/research-and-data/publications/working-papers/2026/09/28/16-bailing-out-firms-uninsured-deposits-a-quantitative-analysis"
paper_id: "Atlanta Fed WP 2026-16"
harvested: 2026-09-28
year: 2026
keywords: [uninsured deposits, deposit insurance, bank runs, bailouts, bank failures, financial stability, general equilibrium]
pdf_url: "https://www.atlantafed.org/-/media/Project/Atlanta/FRBA/Documents/research-and-data/publications/working-papers/2026/09/28/16-bailing-out-firms-uninsured-deposits-a-quantitative-analysis.pdf"
---
## Abstract

We document two novel stylized facts about uninsured deposits and bank failures: (i) firms hold more uninsured deposits than households, and (ii) uninsured depositors do not suffer losses in 94 percent of bank failures. We construct a quantitative general equilibrium model designed around these facts. We characterize the marginal benefits of increased insurance for firms' and households' deposits separately, finding higher benefits for household insurance. Quantitatively, the welfare gains from increasing insurance arise primarily from the reduction in bank runs driven by the higher limit on households' deposits. Reducing bailouts increases bank failures due to runs and generates severe macroeconomic effects.

JEL classification: E20, G21, G28, G32. Key words: uninsured deposits, deposit insurance, bailouts, firms' deposits, households' deposits. DOI: 10.29338/wp2026-16

## Why it's in the StratMac library

- **Quantifies the de facto bailout guarantee, which is the assumption behind SM's banking cards.** The second stylized fact — uninsured depositors take no losses in 94% of bank failures — is the empirical reason uninsured deposits have not been priced as unsecured funding. A banking-liquidity card that treats uninsured deposits as runnable is assuming a regime this paper shows has not held in practice.
- **Separates the household and firm margins.** Firms hold *more* uninsured deposits than households, yet the welfare gain from raising the insurance limit comes mainly through households (the run-reduction channel). The two margins are not interchangeable, which matters for any reading of deposit-flow stress as a macro signal.
- **Counterfactual discipline on "let them fail".** The model says reducing bailouts raises failures via runs and produces severe macroeconomic effects — a bound on the cost side of any post-2023 regime change, and a reason the deposit-guarantee assumption is load-bearing rather than incidental.
- Supports the `banking` section; the paper's own related links (FDIC deposit-insurance-reform options, FDIC deposit-flow study) are the primary-source companions.