Money markets are the foundational layer of the financial system — the infrastructure through which banks fund overnight positions, non-bank intermediaries roll short-term liabilities, corporations manage cash, and central banks transmit policy. The visible layer of interest rates, yield curves, and cross-border capital flows rests upon a plumbing of repo agreements, money market funds, securities lending, FX swaps, and commercial paper that most observers never examine. Over the past fifteen years, that plumbing has been reengineered — by post-crisis regulation, by quantitative easing and its reversal, by the collateral demands of cleared derivatives, and by the dramatic expansion of non-bank intermediation. The system that emerges is structurally different from the one that existed in 2008, and it operates under constraints that intermittently produce stress episodes — the September 2019 repo spike, the March 2020 Treasury market dislocation, the September 2022 UK gilt crisis — that reveal how thin the margin has become between ordinary market function and acute malfunction.

This analysis examines the structural collateral shortage that sits at the heart of modern money market dynamics. The question is not whether there is enough Treasury debt outstanding in absolute terms — there is an enormous quantity. The question is whether there is enough high-quality liquid collateral available at the margin, in the hands of the institutions that need it, at the times and prices that preserve orderly funding market function. The answer, across a growing body of evidence, is that the effective supply of collateral relative to demand has become structurally constrained in ways that require ongoing central bank accommodation to manage — and that the shift represents a permanent transition rather than a transient post-crisis adjustment.

Why Collateral Matters

Money markets are, at their core, collateralized lending markets. A repo transaction is a loan of cash secured by a security; a reverse repo is the mirror image. Securities lending involves lending a specific security against cash or other securities as collateral. Cleared derivatives require the posting of initial and variation margin in eligible collateral. Bank liquidity regulations require the holding of high-quality liquid assets (HQLA) against short-term funding runoff assumptions. In each of these contexts, the availability of acceptable collateral — predominantly U.S. Treasuries, agency mortgage-backed securities, and to a lesser extent high-grade sovereign debt from other major jurisdictions — determines the capacity of the system to function.

Three structural forces have pushed collateral demand upward while constraining effective collateral supply. The first is post-crisis regulation. Basel III liquidity coverage and net stable funding rules impose large HQLA holding requirements on banks. Mandatory clearing of standardized over-the-counter derivatives requires initial margin at central counterparties that is almost exclusively posted in Treasuries or cash. Uncleared margin rules extend similar requirements to bilateral derivative trading. The cumulative effect has been to lock substantial quantities of Treasuries into regulatory buffers where they are unavailable for repo market intermediation or securities lending.

The second force is the concentration of Treasury holdings in price-insensitive hands. The Federal Reserve's System Open Market Account holds several trillion dollars of Treasuries as a legacy of quantitative easing, and although quantitative tightening has reduced the stock, the residual remains historically elevated. Foreign central banks hold additional trillions. Money market funds hold Treasuries against short-term fund share liabilities. These holders do not, in general, lend their securities into repo markets at scale. The free float of Treasuries available for repo intermediation is therefore substantially smaller than the aggregate outstanding stock suggests.

The third force is the expansion of non-bank demand. Hedge funds, relative-value trading desks, asset managers, insurance companies, and corporate treasuries all require collateral — for leveraged positions, for derivative margin, for securities finance strategies, for short-selling programs. As non-bank intermediation has grown in importance, this demand has pressed against a supply that has not expanded commensurately.

$28T+
U.S. Treasury Debt Outstanding
$6T+
Daily Tri-Party Repo Volume
$6T+
Money Market Fund Assets

The Repo Complex

The repurchase agreement market is the single most important funding market in the United States and the principal transmission channel for collateralized short-term interest rates. A repo is functionally a collateralized loan: one counterparty sells a security to another with an agreement to repurchase it at a specified price on a specified future date, with the difference between sale and repurchase prices implying an interest rate. The market operates in several tiers — bilateral repo between primary dealers and their customers, tri-party repo cleared through the Bank of New York Mellon, centrally cleared repo through the Fixed Income Clearing Corporation's sponsored and GCF programs, and direct repo with the Federal Reserve through the standing repo facility and reverse repo facility.

Tri-party repo is the clearing platform through which the bulk of general collateral repo settles. Tri-party volumes of $6 trillion or more per day pass through the system, collateralized primarily by Treasuries and agency mortgage-backed securities. The tri-party infrastructure — with its central custody, automated allocation of collateral from general pools, and intraday credit extended by the clearing bank — is what makes large-scale repo intermediation possible. Stresses in the tri-party plumbing tend to manifest as stresses in overnight rates across the broader system.

Centrally cleared repo, through FICC, has grown dramatically as dealers and hedge funds have migrated to the efficiency and netting benefits of central counterparty clearing. The sponsored service extends cleared access to non-dealer counterparties including money market funds and other institutional cash lenders. The expansion of FICC clearing has altered the distribution of intermediation — dealers can net exposures against other dealers through the central counterparty, reducing balance-sheet consumption and enabling larger intermediation volumes than bilateral arrangements would support. But this migration has also concentrated systemic exposure in the central counterparty, whose own stress tolerance becomes a macro-critical variable.

The September 2019 Episode

The clearest recent demonstration of how collateral scarcity can manifest as funding stress occurred in September 2019. Overnight repo rates, which typically trade within a narrow band around the Federal Reserve's policy target, spiked to as high as ten percent intraday on September 17 — a move of several hundred basis points in a market where typical daily volatility is measured in single basis points. The proximate causes were a coincidence of corporate tax payment date (draining cash from the banking system) and a Treasury coupon settlement (draining cash against newly issued securities), but the underlying cause was structural: the banking system did not hold sufficient excess reserves to absorb the transient cash demands, and the plumbing of repo intermediation could not re-pool collateral and cash efficiently enough to prevent a rate spike.

The Federal Reserve's response was to inject liquidity through open market operations, to expand the System Open Market Account's Treasury holdings (halting and reversing the first episode of quantitative tightening), and eventually to establish the standing repo facility as a permanent backstop. The episode demonstrated that the level of bank reserves that had seemed comfortably abundant in earlier periods was, in fact, close to a structural floor below which money market function deteriorates rapidly. That floor has shifted upward since — reflecting regulatory changes, the growth of non-bank demand, and the accumulation of Treasury debt — and the Fed's operating framework has been adjusted to maintain reserves above it as a matter of policy.

The Standing Repo Facility

The Standing Repo Facility (SRF), established in 2021, is the Federal Reserve's permanent mechanism for supplying overnight liquidity against Treasury and agency collateral. Eligible counterparties — primary dealers and, under expanded access, depository institutions — can borrow from the Fed at a pre-announced rate against eligible securities. The facility is sized generously (at a published aggregate cap that has been raised over time) and operates at a rate set at or modestly above the top of the federal funds target range. The SRF serves as a ceiling on overnight rates: if market repo rates spike above the SRF rate, counterparties can substitute the facility for market funding and arbitrage the rate down.

The effectiveness of the SRF depends on the willingness of eligible counterparties to use it. Historically, banks have been reluctant to access Fed facilities out of stigma concerns and balance-sheet management considerations. The SRF is explicitly designed to be non-stigmatized — usage is aggregated and reported only with a lag, rate-setting is routine rather than discretionary, and frequent use is encouraged as normal operational activity. Whether the facility will be utilized effectively in a stress scenario remains a matter of ongoing observation; the design is sound, but the behavioral response under stress has not been fully tested.

The Overnight Reverse Repo Facility

The Overnight Reverse Repo Facility (ON RRP) is the mirror of the SRF — the mechanism through which the Federal Reserve absorbs excess cash from eligible counterparties in exchange for Treasury collateral. Eligible counterparties include money market funds, government-sponsored enterprises, and certain depository institutions. The facility operates at a rate set at the bottom of the federal funds target range. The ON RRP serves as a floor on overnight rates: if market rates fall below the ON RRP rate, eligible counterparties can substitute the facility for market lending, supporting rates from below.

Usage of the ON RRP has varied dramatically with Federal Reserve balance-sheet policy. During the post-pandemic period of peak quantitative easing, ON RRP balances rose to over $2 trillion as money market funds found themselves with surplus cash and inadequate alternatives. As quantitative tightening has drained system cash and as Treasury bill issuance has provided alternative investment vehicles for MMF cash, ON RRP balances have declined substantially. The trajectory of ON RRP usage is a real-time indicator of the supply-demand balance in the short-end of the curve — rising usage signaling excess cash relative to private investment opportunities, declining usage signaling cash being reabsorbed by the private market.

The reverse repo facility is not a policy instrument — it is the marginal private-sector investment opportunity of last resort. Its utilization is a thermometer for the balance of cash and collateral in the system, not a lever with which to adjust that balance.

— HL Hunt Research Division

SOFR and the Reference Rate Transition

The Secured Overnight Financing Rate (SOFR) is a volume-weighted median of overnight Treasury repo transaction rates, published daily by the Federal Reserve Bank of New York. SOFR was designated in 2017 as the preferred replacement for U.S. dollar LIBOR, which was discontinued for most tenors at the end of June 2023 and permanently ceased publication in September 2024. The transition from LIBOR to SOFR is among the largest infrastructural changes in the history of modern financial markets — hundreds of trillions of dollars of cash and derivative products have been repapered, repriced, or fallbacked to SOFR-based terms.

SOFR differs from LIBOR in critical respects. LIBOR was an unsecured bank funding rate, reflecting bank credit risk and the term structure of interbank lending. SOFR is a secured Treasury repo rate, reflecting the cost of collateralized borrowing rather than bank credit risk. The transition therefore requires more than a rate substitution — it requires a restructuring of the reference rate framework itself. Term SOFR variants have been developed to approximate the forward-looking term structure that LIBOR provided; SOFR-based averaged rates over various lookback periods have been introduced for loan products that require compounded rate conventions; and credit spread adjustments have been calibrated to preserve economic equivalence across legacy contracts that transitioned from LIBOR to SOFR-based fallbacks.

Reference Rate Characteristic USD LIBOR (Historical) SOFR (Current)
Underlying Market Unsecured interbank lending (panel-based submissions) Overnight Treasury repo (transaction-based)
Credit Component Bank credit risk embedded in rate Collateralized; no embedded bank credit
Daily Volume Underlying Declining post-crisis; often thin $1T+ per day in eligible repo transactions
Term Structure Native forward-looking 1M, 3M, 6M, 12M Overnight; Term SOFR and averaged conventions
Stress-Period Behavior Widens with bank stress (captures credit) Can tighten with bank stress (flight to collateral)

The most consequential behavioral difference is the stress-period response. LIBOR historically widened during bank-funding stress as banks demanded compensation for counterparty risk. SOFR, as a secured rate, can behave in the opposite direction — falling during flight-to-collateral episodes as demand for Treasury collateral drives repo rates down. This inversion matters enormously for the economics of loan products and hedging instruments that were originally structured around LIBOR's credit-sensitive behavior. Credit spread adjustments attempt to compensate for this difference in fallback conversions, but the ongoing behavior of new SOFR-referenced contracts during future stress periods remains an open question.

Term SOFR and the Forward-Rate Problem

One of the most persistent operational challenges of the SOFR transition has been the construction of forward-looking term rates. LIBOR was published as a term rate — a one-month rate, a three-month rate — with a forward-looking character that aligned naturally with the payment and reset conventions of most loan products. SOFR is an overnight rate; constructing a forward-looking one-month or three-month equivalent requires either backward-looking averaging (which creates economic tension for borrowers who need to know future rates to manage cash flows) or derivative-implied forward rates (which depend on liquid SOFR derivative markets).

Term SOFR, administered by CME Group, is the industry's solution — a forward-looking term rate constructed from SOFR futures data. Its use in derivatives is restricted to dealer-customer transactions with end-users; interdealer and centrally cleared derivative markets reference averaged SOFR rather than Term SOFR. The bifurcation is intended to prevent Term SOFR from becoming a dominant benchmark in ways that would reintroduce the vulnerabilities that led to LIBOR's discontinuation. Whether the bifurcation will hold over time, and whether liquidity in averaged-SOFR derivative markets will remain robust enough to support the scale of the SOFR cash market, are open structural questions.

Money Market Funds After Reform

Money market funds are the largest cash pools in the U.S. financial system, with aggregate assets exceeding six trillion dollars. MMFs are governed by SEC Rule 2a-7, which has been amended multiple times since the 2008 crisis to reduce their vulnerability to runs, improve liquidity risk management, and limit their contribution to systemic stress. The most recent substantial reforms, completed in 2023, eliminated redemption gates for government and retail prime funds, removed the direct linkage between weekly liquid assets and liquidity fees, imposed mandatory liquidity fees on institutional prime funds during periods of high net redemptions, and increased daily and weekly liquid asset requirements.

The reforms shifted the composition of the MMF industry decisively toward government funds. Institutional prime funds — which historically intermediated non-bank-deposit credit by holding commercial paper and certificates of deposit — have shrunk substantially as investors have migrated to government funds that hold only Treasuries, agency debt, and government-secured repo. The effect on short-term credit markets has been material: commercial paper issuance, which had relied heavily on institutional prime fund demand, has adjusted in composition and in pricing, and bank wholesale funding has reorganized around a smaller pool of MMF demand for bank-issued paper.

The 2023 reforms also introduced a new regulatory tool: the mandatory liquidity fee for institutional prime and tax-exempt funds during stress periods. If net redemptions exceed five percent of net assets on a given day, funds must apply a liquidity fee based on the estimated cost of liquidating the pro rata share of portfolio assets. The tool is intended to internalize to redeeming shareholders the liquidity cost of their redemptions, rather than allowing non-redeeming shareholders to bear the dilution. Whether the mechanism will operate as intended during an acute stress event is untested; it is possible that the anticipation of imminent fee imposition could accelerate rather than dampen run dynamics.

Structural Implication

The government-fund concentration

The concentration of MMF assets in government funds that hold predominantly Treasuries, agency debt, and Treasury-secured repo means that the MMF industry is now a large structural consumer of Treasury collateral. On any given day, MMFs hold over three trillion dollars of Treasury bills and notes and intermediate multi-trillion-dollar volumes of Treasury-secured repo. This concentration reinforces the collateral demand side of the structural shortage and binds MMF behavior tightly to Federal Reserve and Treasury issuance decisions.

The Treasury General Account and Fiscal-Monetary Interaction

The Treasury General Account (TGA) is the operating checking account of the U.S. Treasury at the Federal Reserve. When the Treasury issues debt or collects tax revenue, the proceeds flow into the TGA; when the Treasury makes payments — for Social Security, Medicare, defense contracts, interest payments, and the thousands of other lines of federal spending — funds flow out. Changes in the TGA balance correspond, one-for-one, to changes in the aggregate cash position of the banking system outside the Treasury.

This mechanical accounting has large money-market implications. When the TGA balance rises — because of heavy Treasury bill issuance, tax receipts, or deliberate Treasury cash-balance management — bank reserves and MMF cash fall by an equivalent amount. The effect is equivalent to an unannounced tightening of monetary conditions. The reverse occurs when the TGA balance falls: reserves and MMF cash rise, and monetary conditions loosen at the margin. The Federal Reserve must calibrate its own balance-sheet operations to offset or accommodate these fiscal flows if it wishes to maintain stable reserve conditions — a task complicated by the fact that TGA balance decisions are made by the Treasury based on cash management considerations, not monetary policy objectives.

Debt ceiling episodes introduce acute TGA dynamics. When the debt ceiling is reached and extraordinary measures are invoked, the Treasury typically runs down the TGA to near-zero balances as it exhausts available cash. This releases a large pool of reserves into the banking system and pushes overnight rates toward the lower bound of the Fed's target range. When the debt ceiling is raised and the Treasury rebuilds the TGA through aggressive bill issuance, the reverse occurs — reserves drain, overnight rates press toward the upper bound, and money markets absorb the issuance with varying degrees of smoothness depending on the pace of rebuild. These fiscal-monetary interactions are not transitory background noise; they are first-order drivers of short-term rate dynamics during debt-ceiling episodes and their aftermath.

Cross-Currency Basis and the Dollar Funding System

U.S. dollars are the global reserve currency, and the global demand for dollar funding exceeds the supply available from U.S.-resident banks and money markets. The gap is bridged primarily through FX swap markets, where non-U.S. banks and corporates borrow dollars against their domestic currency, paying a premium over the rate implied by uncovered interest parity. This premium — the cross-currency basis — is a direct measure of dollar funding stress outside the United States. Negative basis in the euro/dollar and yen/dollar markets (the convention is that negative basis means dollars are more expensive than covered interest parity implies) has been a persistent feature of the post-crisis period and widens dramatically during stress episodes.

The Federal Reserve's standing U.S. dollar liquidity swap lines with the European Central Bank, Bank of Japan, Bank of England, Swiss National Bank, and Bank of Canada are the official mechanism for alleviating offshore dollar stress. These swap lines were reactivated during the March 2020 pandemic stress episode and again during the March 2023 banking stress episode, providing hundreds of billions of dollars of dollar liquidity to foreign central banks against their domestic currency collateral. The swap lines are structured as backstops rather than primary funding channels, but their availability is a critical element of the global dollar funding system's stability.

The growth of offshore dollar intermediation through non-bank channels — eurodollar-denominated commercial paper, offshore MMF structures, and FX swap intermediation by broker-dealers and custodian banks — has expanded the effective supply of dollars available to non-U.S. borrowers without expanding U.S.-resident bank balance sheets. This off-balance-sheet dollar intermediation is both a source of systemic resilience (in that it provides dollar access without U.S. bank concentration) and a source of vulnerability (in that it can retrench rapidly during stress in ways that are poorly observed by regulators). The cross-currency basis is the market's real-time signal of where the balance lies.

Implications for Market Participants

The structural reshaping of money markets has direct operational implications for banks, non-bank intermediaries, corporates, and asset managers. A framework for incorporating these implications into operational practice includes the following elements:

  • Liquidity stress testing must incorporate collateral dynamics. Liquidity stress frameworks that assume stable collateral availability at standard haircuts understate risk. Scenario analysis should include collateral scarcity episodes, haircut widening under stress, and failures of specific repo counterparties to extend or rollover financing.
  • The SRF and discount window must be operationally ready. Banks and non-bank entities eligible for Federal Reserve facilities should ensure that collateral is pre-positioned, legal agreements are executed, and operational testing is current. Stigma-driven reluctance to use facilities during stress has historically cost institutions more than the stigma itself was worth.
  • SOFR basis exposure must be measured and hedged. Institutions with mixed LIBOR-era and SOFR-era exposures should measure the residual basis risk that credit spread adjustments do not fully resolve. Bank-credit-sensitive exposures that have been converted to SOFR-based terms are particularly susceptible to basis divergence during future stress.
  • Treasury issuance forecasts should inform cash management. Large corporate treasurers and institutional investors should track Treasury bill issuance trajectories, TGA balance dynamics, and expected reserve-balance evolution as inputs to cash management. The stability of short-term rates is a function of these variables, and forecasting error on cash returns can be material.
  • Cross-currency basis should be monitored as a stress indicator. For institutions with multi-currency funding needs, widening basis is an early warning of dollar funding stress that often leads broader market deterioration by days or weeks.

Institutional Infrastructure From HL Hunt

HL Hunt provides institutional-grade financial technology across payments, credit, and banking — built on infrastructure that reflects the operational realities of modern money markets.

Explore Solutions

Conclusion

Money markets have been reengineered. The system that exists today — with its vastly expanded Federal Reserve balance sheet, its SOFR reference rate framework, its reformed money market fund industry, its centrally cleared repo complex, and its standing facilities that implicitly backstop the private market — is structurally different from the one that existed before the 2008 crisis and materially different from the one that existed five years ago. The changes reflect deliberate policy responses to identified vulnerabilities, but they also reflect emergent adaptations to collateral scarcity, regulatory constraints, and the migration of intermediation to non-bank channels.

The structural collateral shortage at the heart of this reengineered system is not a problem to be solved through a single policy intervention. It is a condition to be managed — through Federal Reserve balance-sheet policy, through standing facilities, through Treasury issuance strategy, through continued evolution of MMF regulation, and through the ongoing adaptation of private-sector intermediation. The system works when these elements are calibrated coherently; it malfunctions episodically when they are not. The episodes of malfunction — September 2019, March 2020, March 2023 — are the observable evidence of how thin the operating margin has become.

For institutional participants, the practical implication is that the operational infrastructure of funding must be treated as a first-order strategic concern rather than a back-office utility. The firms that build robust liquidity frameworks, maintain ready access to official facilities, understand the evolving dynamics of SOFR and Treasury issuance, and monitor cross-currency basis as a stress indicator will navigate future episodes with materially less disruption than those that treat money markets as a given. The reinvention of money markets is not a completed event; it is an ongoing process that will continue to shape financial conditions and institutional operational requirements for the foreseeable future.