Figure 1: The wiring of the tri-party repo
market
Note: The
decentralized dealer-lender OTC market is represented by the set of
dealers in purple squares and the set of lenders as blue circles at the
bottom along with the trading relationships indicated as gray lines
among the two. Dealers that borrow from the GCF market are connected
with a red line to the GCF, whereas the dealers that lend to the GCF
market are connected with a gray line. Lenders with cash held at the
Fed’s ON RRP are indicated with a line connection to the ON RRP. Source: FRBNY Tri-party repo and the authors’
construction.
A comprehensive diagram visualizing the relationships between dealers and lenders in the tri-party repo market. A node at the top labeled GCF (General Collateral Finance) is connected to dealers below it, which are represented by purple boxes, via red or grey lines that represent a dealer is a net borrower or lender, respectively. Almost every dealer has a line connecting it to the GCF node. This top third of the diagram represents the centrally-cleared market between dealers. Grey lines then drop down to connect dealers with lenders, represented by blue circles, representing the decentralized over-the-counter (OTC ) portion of the market. Not every dealer has a line indicatinig a relationship to a lender, but every lender has at least one line indicating a relationship with a dealer. There are labeled arrows indicating that lenders provide cash to dealers while dealers put up collateral for lenders. The bottom third of the chart connects lenders to a node representing the Federal Reserve's overnight reverse repo (ON RRP) facility. Not all lenders are connected to the ON RRP.
Figure 2: Duration of trading relationships as
fraction of the sample period
Note: The x-axis is the duration of
dealer-lender relationships as a fraction of the sample period from July
2014 to October 2022 or 2005 trading days. Source: FRBNY Tri-party repo and authors’
calculations.
A histogram showing durations of dealer-lender trading relationships as percentages of the sample period on the x-axis and share of total OTC (over-the-counter) repo volume as a percentage on the y-axis (ranging from 0 to 50). The x-axis contains 10 bins, each with a range of 10 percentage points representing a range of duration of the dealer-lender relationship as a percent of the sample period. Around 8% of total OTC (over-the-counter) repo volume is from dealer-lender relatioships which only last for 0-10% of the sample period. Around 4% of total OTC (over-the-counter) repo volumn is in each of the next four bins, representing dealer-lender relationships that last for 10-20%, 20-30%., 30-40%, and 40-50% of the sample period, respectively. There is a spike at dealer-lender relationships which exist for 50-60% of the sample period, containing around 8% of total OTC (over-the-counter) repo volumne. The three bins for 60-70%, 70-80%, and 80-90% for duration of dealer-lender relationship all have around 5-6% of the share of total OTC (over-the-counter) repo volume each. Lastly, the bin for dealer-lender relationships with a duration of 90-100% of the sample period contains nearly 50% of total OTC (over-the-counter) repo volume.
Figure 3: Trading volume at new, dissolved, and short-term trading
relationships
Note: The
graph plots the volume of repo trades at newly established, dissolved,
and short-term dealer-lender trading relationships. Short-term trading
relationships are defined as those that last no more than a fifth of the
sample period or two years or less. Source: FRBNY Tri-party repo and authors’
calculations.
A line chart plotting the volume of repo trades for newly established, dissolved, and short-term dealer-lender relationships with years on the x-axis ranging from 2014 to 2022. The y-axis shows percent of total OTC (over-the-counter) repo volume, ranging from 0 to 8 percent. Short-term relationships account for nearly 8 percent of total OTC (over-the-counter) repo volume at the beginning of the period, but their share steadily decreases until a larger drop to below 1 percent in 2016. The short-term relationships continue to account for around 1 percent of total OTC (over-the-counter) repo volume until the end of the sample, 2022, when they see a slight increase to just over 2 percent. The new dealer-lender relationships are at an even lower level than the short-term relationships for most of the sample except for four spikes in 2014, 2015, 2017, and 2018. At its max, in 2018, new dealer-lender relationships make up more than 5 percent of total OTC (over-the-counter) repo volume. Lastly, dissolved dealer-lender relationships make up the lowest percent of total OTC (over-the-counter) repo volume throughout the time period, being consistently below 1 percent.
Figure 4: Alternative wirings of a market with 2 dealers and 3 lenders
A. Symmetric but incomplete wiring (\(\mathcal{T}^{sym}\))
B. Asymmetric wiring (\(\mathcal{T}^{asym}\))
C. Fully wired (\(\mathcal{T}^{full}\))
Note: Lines between a dealer and a
lender represent an established relationship. Lines between dealers and
the \(\mathcal{C}\) market represent
access to the centrally-cleared market, which could be either borrowing
or lending. In this example, we assume that \(q_{\mathcal{B}1}>q_{\mathcal{B}2}\) and
that dealer \(d_1\) needs additional
cash and borrows from dealer \(d_2\)
through the \(\mathcal{C}\) market.
A diagram illustrating three types of alternative wirings of a market. All three markets contain a node at the top to represent a centrally cleared market between dealers and contain two dealers and three lenders. Panel A has two dealers which share one lender and each have their own unique lender as well. This represents symmetric but incomplete wiring. Panel B is the same as panel A except lender three also lends to both dealers, meaning dealer 1 has three lenders while dealer 2 only has two lenders, making this market have asymmetric wiring. Lastly, panel 3 is fully wired since all three lenders are connected with both dealers.
Figure 5: Line graph of 3\(\times\) 2
market wiring
A. Symmetric but incomplete wiring
B. Asymmetric and incomplete wiring
C. Fully wired
Note: The line graph of the bipartite
networks in Figure 4. Dashed lines connect edges with common dealers and
thick lines connect edges with common lenders.
A diagram illustrating the bipartite networks in Figure 4. In panel A, the symmetric but incompletely wired network, dealers 1 and 2 are connected through a common relationship to lender 2, while lenders one and three are only connected to dealers one and two, respectively. In panel B, the asymmetric and incompletely wired network, the same is true as in A except for dealers 1 and 2 are also connected through lender 3 and all three lenders are connected through dealer 1. In panel C, the fully wired network, the same is true as in B except for dealers 1 and 2 are also connected through lender 1 and the three lenders are all connected through dealer 2 as well.
Figure 6: Repo volumes by market segment
Note: The black line is
trading volume in the decentralized OTC segment of the tri-party repo
market. The light blue line is trading volume in the centrally cleared
GCF market. The full list of ON RRP counterparties can be found at the
https://www.newyorkfed.org/markets/rrp_counterpartiesFRB NY
website. Vertical dashed lines indicate quarter-ends. Source: FRBNY Tri-party repo.
A time-series chart from 2014 to 2022 showing OTC (over-the-counter) and GCF trading volumes. OTC (over-the-counter) volumes peaked at $1 trillion in 2019 and declined to $700 billion by late 2022. GCF volumes dropped from $400 billion in 2019 to $200 billion in 2022. The volume in the OTC (over-the-counter) and GCF markets start the same at around $400 billion in 2014 but immediately start diverging with the volume in the OTC (over-the-counter) market always being greater than that in the GCF market after 2015. The lower panel plots the repo volume in the ON RRP facility which oscillates withini the range of $0-$500 billion before 2018 and then stays much lower until a small spike in 2020 (still less than $500 billion). The volume has rapidly increased to now closer to $2.25 trillion from 2021 through 2022.
Figure 7: Borrowing from FRB repo operations and the GCF
Note: The temporary FRB repo operations
we considered started on September 18, 2019 and were completed in June
2020. Borrowing from the repo operations in purple is indicated as FRB
in the chart’s legend. The temporary repo operations were followed by
the establishment of the permanent Standing Repo Facility (SRF) in July
2021. The SRF has not been actively used since its establishment. Source: FRBNY Tri-party repo.
A time-series graph with data on FRB repo operations, ON RRP usage, and GCF volumes. Highlights include the September 2019 and COVID-19 market stresses, showing borrowing peaks around $200 billion for the temporary FRB repo operations. This borrowing declined to zero by June 2020. The chart also highlights the connection between the excess cash deposited at the ON RRP and the centrally cleared GCF market.
Figure 8: Foreign dealer trades at quarter-ends and ON RRP take-up
A. Foreign dealer repo trades (OTC)
B. ON RRP
deposits
Note: The
figures compute the average amounts over a 5-day window around the last
trading date of the quarter. Source:
FRBNY Tri-party repo and authors’ calculations.
Two panels are time-series graphs, which both have days to quarter end on the x-axis ranging from -5 to 5 and billion dollars on the y-axis ranging from -150 to 150. They each have three separate lines representing three time periods of 2014-2017, 2018-2020, and 2020-2022. Panel A shows the average foreign dealer repo trades in the OTC (over-the-counter) within 5 days of quarter end, with a minimum at -$150B on the day of quarter end and a maximum at $200B 1 day to quarter end. Panel B shows the ON RRP deposits with a maximum for each time period on the day of quarter end and a minimum 1 day to quarter end.
Figure 9: Distribution of government repo
rates
Note: Spreads are
expressed in percentage points. Government repo includes repo trades in
Treasury securities, agency debt, and agency MBS. The sample period for
rates on repo trades in the GCF market begins in 2016. Source: FRBNY Tri-party repo and authors’
calculations.
A time-series graph which goes from 2014 to 2022 and shows the repo rates in percentage points for the GCF rate, ON RRP rate, median OTC (over-the-counter) rates, and 5/95th percentile OTC (over-the-counter) rates. The dispersion of rates is very tight around the ON RRP rate for most days, but at quarter-ends there are notable spikes in repo rates and increases in the dispersion and levels of rates. There is one large spike on September 17, 2019.
Figure 10: Dispersion in spreads over
the ON RRP rate
Note:
Quarter-ends are the last trading day for the quarter. Other includes
all trading days outside quarter-ends and the September 2019 spikes.
Source: FRBNY Tri-party repo and authors’
calculations.
Represents overlayed histograms of the spreads over the ON RRP rate for three periods: September 17, 2019; quarter-ends; and all other times. Quarter-ends have more dispersed density that hovers between 0 and 1 percentage points relative to other periods when the histogram is tightly hugging the zero line. The September 17, 2019 spreads are display the greatest dispersion, ranging from 0 to 6 percentage points and with a spike in the mass around 3 percentage points.
Figure 11: Decomposition of supply and demand
factors
Note: The figure
shows the monthly weighted averages of the daily supply and demand
factors based on the solution of the system of equations. See appendix
for details on the construction. Source:
FRBNY Tri-party repo and authors’ calculations.
A time-series chart from 2014 to 2022 showing the monthly weighted averages of the daily supply and demand factors based on the solution of the system of equations of the Amiti-Weinstein decomposition explained in more detail in the appendix. There is significant variation in both the supply and demand series over time, with the y-axis (growth, percentage points) ranging from -20 to 30. Through 2018, demand declines at quarter-ends. Starting in 2018, the magnitude of supply shocks increases.
Figure 12: Supply and demand factors
around September 19, 2019
Note: The boxplots represent the interquartile
ranges of the percent changes in supply and demand across lender and
dealers, respectively. The solid black lines are the medians and the
interquartile ranges are represented in the colored boxes. The whiskers
represent the 5th and 95th percentiles of the distribution. Source: FRBNY Tri-party repo and authors’
calculations.
Boxplots over the course of six days (September 13, 2019 to September 20, 2019) showing the interquartile ranges of the percent changes in supply and demand across lenders and dealers, respectively. There is a large negative supply shock on September 17, 2019 which resulted in a 30 percent decline in supply at the median lender and over a 50 percent decline at the 95th percentile of the distribution. On that same day, there was no notable change in aggregate demand at the median dealer, with the percent change hovering around zero. The percent change for the supply and demand for the median does not go outside of plus or minus 20 percent change for any of the other days in the chart.