Beyond the Median: Reserve Scarcity, Liquidity Buffers and Distributional Stress in US Repo Markets

Abstract How can banks and central banks recognise pressure in overnight funding markets when most transactions still take place at ordinary rates? This study examines the full published distribution of US secured overnight funding rates from April 2018 to October 2026. It combines 2,134 SOFR observation dates with reserve, bank-asset and overnight reverse repo information, using 438 weekly observations for the main reserve regressions. Funding costs are related to reserves throughout the rate distribution, with a larger estimated slope at the 99th percentile (− 2.00 basis points per percentage-point increase in reserves relative to bank assets) than at the median (− 1.21 basis points). Dispersion between the 75th and 25th percentiles also narrows as reserves rise. Interactions with the Federal Reserve’s overnight reverse repo balance and quarter-end reporting dates identify conditions under which these associations differ. The upper-tail reserve threshold is not consistently earlier than the median threshold. In a separate 2023-onward forecast test, upper-tail information improves Brier scores relative to a median model at one and five days, but forecast probabilities require substantial recalibration. The results support a distribution-aware approach to liquidity monitoring that combines reserve levels, liquidity buffers, market segmentation and disciplined forecast verification.

Authors

Institutions

Publication Details

Journal
Research Square
Published
2026-11-12
DOI
https://doi.org/10.21203/rs.3.rs-11444156/v1
Primary Topic
Banking stability, regulation, efficiency
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Beyond the Median: Reserve Scarcity, Liquidity Buffers and Distributional Stress in US Repo Markets

Vamsidhar Ambatipudi
Research Square
Banking stability, regulation, efficiency
preprint

Beyond the Median: Reserve Scarcity, Liquidity Buffers and Distributional Stress in US Repo Markets

Vamsidhar Ambatipudi
preprint en

Abstract

Abstract How can banks and central banks recognise pressure in overnight funding markets when most transactions still take place at ordinary rates? This study examines the full published distribution of US secured overnight funding rates from April 2018 to October 2026. It combines 2,134 SOFR observation dates with reserve, bank-asset and overnight reverse repo information, using 438 weekly observations for the main reserve regressions. Funding costs are related to reserves throughout the rate distribution, with a larger estimated slope at the 99th percentile (− 2.00 basis points per percentage-point increase in reserves relative to bank assets) than at the median (− 1.21 basis points). Dispersion between the 75th and 25th percentiles also narrows as reserves rise. Interactions with the Federal Reserve’s overnight reverse repo balance and quarter-end reporting dates identify conditions under which these associations differ. The upper-tail reserve threshold is not consistently earlier than the median threshold. In a separate 2023-onward forecast test, upper-tail information improves Brier scores relative to a median model at one and five days, but forecast probabilities require substantial recalibration. The results support a distribution-aware approach to liquidity monitoring that combines reserve levels, liquidity buffers, market segmentation and disciplined forecast verification.

Research Square
Birla Institute of Technology and Science, Pilani (IN)
Banking stability, regulation, efficiency
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon