Accessible Version
Monitoring Reserve Scarcity Through Nonbank Cash Lenders, Accessible Data
Indicator 1. Share of Repo Volumes Above Interest on Reserve Balances
Plots the share of money market fund lending and GCF repo lending with interest rates above the interest rate on reserve balances over time. The horizonal axis is the timeline, running from 2013 to the end of 2024. A small yellow bar denotes the month of September 2019. The vertical axis is percent, running from zero to 100. The line for the share of money market funds is nearly always zero, with occasional small blips above zero up until 2018. Then it rises to around 20 percent, with occasional spikes downward, until the end of 2019 when it rises to roughly 80 percent, and the almost 100 percent by mid 2019. Following the end of 2019 it quickly falls in 2020 and then remains near zero from mid-2020 on, with small blips above zero beginning again after 2024. The GCF repo series begins near zero but starts spiking up to roughly 80 percent by mid-2015. By early 2018, the series is consistently between 80 percent and 100 percent. In early 2020, there is a short spike down to roughly 40 percent, after which the series comes back to nearly 100 percent and then gradually falls to 60 percent by the end of 2020, after which it rapidly falls to zero by early 2021. The GCF series then remains at zero until the end of 2022, when it begins to spike up again to between 20 and 60 percent by then end of 2024.
Note: The vertical yellow shaded area spans the month of September 2019. Gray shaded areas denote periods of quantitative tightening. All series are 30-day rolling averages. “GCF” stands for the GCF Repo service of the Fixed Income Clearing Corporation (FICC), a blind-brokered inter-dealer segment of the repo market that is centrally cleared through FICC and settled on a tri-party basis. Money market fund lending is the share of tri-party lending by MMFs above the IORB rate (or IOER rate prior to July 29, 2021). Inter-dealer is the share of GCF lending above the IORB rate.
Sources: Federal Reserve tri-party repo transactions data, DTCC Solutions LLC (an affiliate of The Depository Trust & Clearing Corporation), Federal Reserve Policy Rates release.
Indicator 2. Coefficient of 90-day rolling regression of changes in spreads on changes in TGA balances
Plots the elasticity of the spreads of MMF repo over ON RRP and EFFR over IORB over time. The horizontal axis is a timeline running from the beginning of 2015 to the end of 2024. A small yellow bar denotes the month of September 2019. The spread of MMF repo over ON RRP begins oscillating between roughly zero and roughly 0.05 from 2015 to the end of 2017. After 2017 it gradually rises over 2018 to consistently around 0.1, then jumps to 0.15 at the end of the year, staying at that level until September 2019, when it again quickly rises to roughly 0.3 by year end 2019, then gradually falls to zero by the end of 2020. The line stays near zero though the rest of the sample, though there is a slight rise to just barely above zero following mid-2024. The line for the EFFR spread begins slightly below zero and falls to roughly -0.1 by the end of 2016, then slowly rises to zero by the end of 2017. It then remains near zero until mid-2019, when it begins to rise slightly and jumps to roughly 0.05 in September 2019, remaining at that level until the end of 2019, when it falls to zero, where it remains till the end of the timeline.
Notes: The vertical yellow shaded area spans the month of September 2019. The blue and orange shaded areas, respectively, represent the 95% confidence intervals around the elasticity estimates for MMF repo spreads and EFFR spreads.
Sources: Federal Reserve tri-party repo transactions data, Federal Reserve Bank of New York Reference Rates release, Federal Reserve Policy Rates release.
Indicator 3. Fed funds volumes and spreads paid by domestic borrowers
There are two panels, the top panel shows volumes in the market in billions of dollars, with the horizontal axis running from 2016 to the end of 2024 and the vertical axis running between -10 and 40. There are three stacked areas labelled LCR Borrower - FHLB Lender, Non-LCR Borrower – FHLB Lender, and Non-LCR Borrower – Non-FHLB Lender. Volumes for LCR Borrower – FHLB Lender begin around $2.5 billion, rise to roughly $5 billion between the beginning and middle of 2019, remain at this level through mid-2020 and then gradually decline to $2.5 billion by mid-2021. They then rise again to roughly $5 billion by the end of 2023, and gradually fall again to $2.5 billion by the end of 2024. The stacked line for LCR Borrower - FHLB Lender and Non-LCR Borrower – FHLB Lender follows a similar patter, beginning at roughly $5 billion, rising to roughly $10 billion by mid-2019, then declining to roughly $2.5 billion by mid 2021 as the Non-LCR Borrower – FHLB Lender practically disappears. It then rises again to roughly $8 billion by the end of 2022, and declines again to roughly $2.5 billion by the end of 2024. Finally, the full stacked area including Non-LCR Borrower – Non-FHLB Lender begins at roughly $10-15 billion, rising to around $20 billion by early 2018, remains at this level until early 2020 when it falls to around $5 billion, and has another hump between 2022 and 2023 were it rises to roughl $20 billion before declining again by the end of 2024.
The bottom panel shows spreads over IORB for LCR Borrower - FHLB Lender, Non-LCR Borrower – FHLB Lender, and Non-LCR Borrower – Non-FHLB Lender as a time series from 2016 to the end of 2024, with a vertical axis running from -20 to 50 basis points. Non-LCR Borrower – FHLB Lender and LCR Borrower – FHLB Lender both begin near -15 basis points and gradually rise to roughly 5 basis points by mid-2019, then gradually fall to roughly -8 basis points till the end of the sample apart from a brief spike to 20 basis points in March of 2020. Non-LCR borrower – Non-FHLB lender begins a little above zero, and remains there until early 2018. Then it rises to roughly 15 basis points by the end of 2018, before gradually falling down to roughly -10 basis points by end of 2021 (with another brief spike to roughly 20 basis points in March 2020). The series then rises to 30 basis points by the beginning of 2023, after which it gradually falls back to roughly 0 basis points by the end of 2024.
Note: For volumes in the top panel, key identifies in order from top to bottom. The vertical yellow shaded area spans the month of September 2019. FHLB series may include negligible amount of non-FHLB GSEs. The “LCR-Borrower – Non-FHLB Lender” category had negligible volume and was excluded. For confidentiality reasons, data was omitted on days when there were few observations in the underlying data.
Sources: FedWire, FR 2420.
Indicator 4. FHLBs’ fed funds lending below repo rate
There are two panels, each with two axes and four time series that run from 2016 through the end of 2024. The two panels are labelled FHLBs’ Fed Funds Lending Below Repo Rate and Persistent FHLB’s Fed Funds Lending Below Repo Rate. The first two series in each panel are labelled Non-LCR Borrower spread and LCR Borrower Spread, plotted on a left axis that runs from -100 to 0 basis points for the top panel and -50 to 0 basis points for the bottom panel. The two series in each panel display a similar pattern, running near zero or missing up until early 2018, then running between 0 and -5 between 2018 and mid-2020, with brief spikes down to -10 in mid-2018 for both series, and a spike to -50 in early 2020 for Non-LCR borrow spread in the top panel. Then the series gradually increase to zero or missing by late 2021, and remain either zero or missing till the end of the sample. The second two series in both panels are Non-LCR borrower volume and LCR borrower volume and they are shown on the right vertical axis which runs from $0 to $30 billion. In both panels these series are essentially zero until early 2018, when they both spike up to roughly 20 billion, and then fall to roughly 10 billion by mid-2018, after which they gradually decline to between 0 and $2 billion by early 2020, where they remain until late 2021. After late 2021 these two series are mostly zero in both panels.
Note: The vertical yellow shaded area spans the month of September 2019. For confidentiality reasons, data was omitted on days when there were few observations in the underlying data.
Sources: Federal Reserve tri-party repo transactions data, FedWire, FR 2420.
Indicator A1. Share of Institutions in Total Private Tri-Party Repo
Shows three series: the share of money market funds (MMFs), banks and other lenders in total private repo between early 2014 and late 2024. The vertical axis ranges from 0 to 80 percent. MMFs and banks move in roughly opposite directions, both begin at roughly 30 percent and remain there through the beginning of 2017, at which point banks drop down to roughly 20 percent and MMFs rise up to roughly 55 percent. They remain at roughly these levels through early 2020, at which point between early 20202 and mid-2022 MMFs fall to 20 percent while banks rise to 45 percent. After this point, MMFs gradually rise again to 50 percent by the end of 2024 while banks fall to 20 percent. Meanwhile, other lenders have several oscillations but gradually trend down from roughly 40 percent in the beginning of 2014 to roughly 20 percent by the end of 2024.
Note: The vertical yellow shaded area spans the month of September 2019. “Other” series includes non-MMF asset managers and other lenders.
Source: Federal Reserve tri-party transactions data.
Indicator A2. MMF Repo Rates vs. Bank Reserves (left) and MMF Private Repo Allocation (right)
There are two scatter plots labelled “Bank reserves” and “MMF private repo”. The two scatter plots share one vertical axis showing the spread of MMF tri-party to ON RRP in percentage points which ranges from -0.2 to 0.8. The “bank reserves” panel has a horizontal axis labelled “ratio of bank reserves to Treasuries outstanding” which ranges from 12.5 percent to 27.5 percent. This panel shows a generally negative relationship between bank reserves and spreads, with a decreasing slope that begins nearly vertical near 12.5 percent on the horizontal axis, and ends nearly horizontal after 20 on the horizontal axis. The “MMF private repo panel” has a horizontal axis labelled “MMF private repo allocation (share)” that ranges from 0.1 to roughly 0.6. It shows a generally positive relationship between private repo allocation and spreads, with a slope that begins slightly positive around 0.1 on the horizontal axis, then increases quickly after 0.4 on the horizontal axis, becoming near vertical around 0.45-0.5.
Note: Ratio of bank reserves to treasuries outstanding are calculated as 10-day rolling averages. MMF private repo allocation is defined as the quantity of MMF private repo investments divided by the sum of MMF private repo, MMF investments in the Federal Reserve overnight reverse-repo facility, and estimated MMF holdings of Treasury bills.
Sources: Federal Reserve tri-party transaction data, Reserve Central-Reserve Account Administration data, Treasury Direct.
Indicator A3. Fed funds’ volume by borrower-lender types
Shows volumes in billions of dollars, with the horizontal axis running from 2016 to the end of 2024 and the vertical axis running between -50 and 200. There are four stacked areas labelled FBO Borrower – FHLB Lender, LCR Borrower - FHLB Lender, Non-LCR Borrower – FHLB Lender, and Non-LCR Borrower – Non-FHLB Lender. Volumes for LCR Borrower – FHLB Lender begin around $2.5 billion, rise to roughly $5 billion between the beginning and middle of 2019, remain at this level through mid-2020 and then gradually decline to $2.5 billion by mid-2021. They then rise again to roughly $5 billion by the end of 2023, and gradually fall again to $2.5 billion by the end of 2024. The stacked line for LCR Borrower - FHLB Lender and Non-LCR Borrower – FHLB Lender follows a similar patter, beginning at roughly $5 billion, rising to roughly $10 billion by mid-2019, then declining to roughly $2.5 billion by mid 2021 as the Non-LCR Borrower – FHLB Lender practically disappears. It then rises again to roughly $8 billion by the end of 2022, and declines again to roughly $2.5 billion by the end of 2024. The stacked area including Non-LCR Borrower – Non-FHLB Lender begins at roughly $10-15 billion, rising to around $20 billion by early 2018, remains at this level until early 2020 when it falls to around $5 billion, and has another hump between 2022 and 2023 were it rises to roughly $20 billion before declining again by the end of 2024. The full stacked area including FBO Borrower – FHLB Lender varies around $50 billion to $75 billion until mid-2020 before gradually rising to almost $150 billion by mid-2023. It then gradually falls to roughly $100 billion by the end of 2024.
Note: Key identifies in order from top to bottom. The vertical yellow shaded area spans the month of September 2019. FHLB series may include negligible amount of non-FHLB GSEs. The “FBO-Borrower – Non-FHLB Lender” and “LCR Borrower – Non-FHLB Lender” categories had negligible volume and were excluded. For confidentiality reasons, data was omitted on days when there were few observations in the underlying data.
Sources: FedWire, FR 2420.