The Treasury Market's Toxic Codependency

The April 2025 tariff-announcement shock produced a genuine, quantifiable unwind of leveraged Treasury positions -- an estimated $60 billion (20%) of swap-spread positions unwound in April 2025 alone, with a further $40 billion unwinding in May -- but Treasury market liquidity deterioration was materially smaller than in March 2020 by standard transaction-cost measures. [Fact, Federal Reserve Board and Dallas Fed research]

The clearest, most directly evidenced explanation for the improved April 2025 outcome relative to March 2020 is the Federal Reserve's Standing Repo Facility, established July 2021: Dallas Fed research found no rise in implied-versus-term-repo-rate spreads in April 2025, in contrast to a 50-basis-point widening in March 2020 before policy intervention restored market functioning. [Fact, attributed to Dallas Fed research; the causal attribution to the SRF specifically is that research's own interpretation, which we find well-supported but distinguish as interpretation rather than directly observed fact]

The US repo market reached $11.9 trillion in 2024, with Treasury collateral backing approximately 89% of centrally cleared repo exposures but only 62% of bilateral repo -- a meaningful, quantified gap in central-clearing coverage that the SEC's Treasury Clearing Rule (mandatory cash Treasury clearing effective December 2026, Treasury repo clearing effective June 2027) is specifically designed to close. [Fact, Federal Reserve data as cited in industry and central-bank research]

We find genuine, primary-sourced evidence for this piece's central paradox: the same leveraged, repo-financed hedge fund intermediation that plausibly deepens Treasury market liquidity and helps absorb record government issuance in normal conditions is the mechanism identified by the Federal Reserve, Dallas Fed and ECB as the principal amplification channel in both the March 2020 and April 2025 stress episodes.

We could not independently verify several numerical claims found in secondary, non-institutional commentary on this topic (including specific claims about hedge fund gross leverage levels and cash allocation levels as of late 2025) against a primary source meeting this piece's sourcing standard, and have excluded them from this analysis rather than repeat unverified figures.

1. Data & Evidence: Sourcing Standard for This Piece

This piece uses recent Financial Times commentary on Treasury market fragility as its intellectual starting point only; no text, structure or argument from that commentary is reproduced here. Every quantitative claim in this piece is sourced directly to a named primary institution -- the Federal Reserve Board, the Federal Reserve Bank of Dallas, the Federal Reserve Bank of New York, the ECB, or the Office of Financial Research -- or is drawn from lucabindi.com's own canonical database, with the distinction stated explicitly throughout. Where secondary commentary contained numerical claims we could not trace to a primary source meeting this standard, we have excluded them rather than repeat them, consistent with this piece's explicit brief.

2. Treasury Market Structure and the Changing Investor Base

Our own database shows US federal debt-to-GDP rising from 100.5% (2015) to roughly 122.6% (2026), a structural increase this programme's global sovereign bond research has documented in detail, alongside a persistent 5-6% of GDP fiscal deficit since 2022. This growth in the stock of outstanding Treasury debt has occurred at the same time official-sector foreign demand -- historically the most price-insensitive segment of the investor base -- has grown only gradually relative to the overall market, per this programme's dollar-dominance companion research, which found the dollar's reserve share declining gradually (62.2% to 57.1% since 2019) without acceleration. The combination -- a larger debt stock, without a proportionally larger price-insensitive buyer base -- is the structural precondition for the more price-sensitive, leveraged intermediation this piece examines: someone has to absorb the marginal issuance, and hedge funds, per the Federal Reserve's own data (Section 3), have become a materially larger part of that marginal absorption than a decade ago.

3. Hedge Funds and Treasury Exposure: The Federal Reserve's Own Taxonomy

Consistent with this piece's brief, we distinguish carefully between the concepts Table 1 and the surrounding Federal Reserve research make explicit: gross long exposure ($2.4tn) is not the same as net exposure (long minus short, materially smaller than $2.4tn given $1.6tn in gross shorts); gross economic exposure is not the same as leveraged, repo-financed exposure specifically (the basis trade, at $830bn, is the largest single leveraged component but not the entirety of the $2.4tn gross long figure); and ownership of cash Treasuries is analytically distinct from derivative (futures) exposure, both of which the Fed's $2.4tn figure aggregates across. We present the Federal Reserve's own $2.4tn/$1.6tn/$830bn figures as directly sourced facts, and would caution against conflating any of these with a single 'hedge funds own $X of Treasuries' headline figure, which the underlying data does not support as a clean characterization.

4. The Basis Trade, Explained

The Treasury cash-futures basis trade exploits a persistent, small price discrepancy between the cash (physical) Treasury security and the corresponding Treasury futures contract: a hedge fund buys the cheaper cash Treasury (financed via repo borrowing) and simultaneously sells the more expensive futures contract, locking in the price gap as a return. Because the price gap itself is typically only a few basis points, the trade is only economically worthwhile at high leverage -- repo financing allows a hedge fund to control a Treasury position many multiples of its own capital, turning a few-basis-point gross spread into a meaningfully larger return on the capital actually committed. The trade is not, in itself, a directional bet on the level of Treasury yields; a fund running a basis trade is typically indifferent to whether yields rise or fall, since it holds offsetting long-cash and short-futures positions. The trade nonetheless creates systemic vulnerability through a distinct channel: if repo funding costs rise sharply, or if the fund's counterparties demand more margin as volatility increases, the fund may be forced to unwind the position rapidly regardless of the trade's own underlying profitability -- selling the cash Treasury leg into a stressed, illiquid market precisely when few other buyers are willing or able to absorb it. This is the mechanism the Federal Reserve, Dallas Fed and academic literature (Kruttli et al.; Barth and Kahn, both cited directly in the Federal Reserve's own June 2026 note) identify as the central transmission channel in both March 2020 and, to a smaller degree, April 2025.

5. Repo Markets: The Financing Infrastructure

The US repo market reached $11.9 trillion in 2024 per Federal Reserve data, with Treasury securities the dominant form of collateral: approximately 89% of centrally cleared repo exposures and 62% of bilateral (non-centrally-cleared) repo are Treasury-collateralized. The gap between these two figures is analytically important: centrally cleared repo benefits from a central counterparty absorbing and managing counterparty risk, while bilateral repo leaves that risk directly between the two counterparties -- a structural reason the SEC's Treasury Clearing Rule specifically targets expanding central clearing to a larger share of both cash Treasury trading and Treasury repo. A further data point with direct relevance to leverage and fragility: as of the most recent quarter for which data is available (Q4 2022, per the primary sources we were able to verify), approximately 73.8% of hedge fund repo borrowing occurred at zero or negative haircuts -- meaning the collateral posted was valued at, or above, the cash borrowed against it, a financing structure that maximizes leverage but leaves essentially no cushion if the collateral's value falls before the repo can be rolled or unwound.

6. March 2020: The Principal Historical Stress Test

The March 2020 Treasury market dysfunction is the standard reference episode for Treasury-market fragility, and the Federal Reserve's own June 2026 research explicitly attributes it 'partly' to the rapid unwinding of hedge funds' basis trade positions, citing academic work by Kruttli et al. and Barth and Kahn among others. The Dallas Fed's own July 2025 research quantifies the funding-market mechanism directly: the spread between the implied repo rate (derived from the basis trade's own pricing) and the actual term repo rate widened by 50 basis points for the June 2020 five-year note futures contract before policy intervention restored market functioning -- a direct, measurable signature of forced, leverage-driven deleveraging distorting the relationship between cash and futures pricing. The Federal Reserve's response -- large-scale Treasury purchases and a substantial expansion of its balance sheet -- ultimately restored market functioning, but the episode is widely regarded, including in the Federal Reserve's own subsequent research, as evidence that a market conventionally understood as the world's deepest, most liquid safe haven can itself become a source of acute dysfunction under sufficiently concentrated, leveraged stress.

7. April 2025: A Direct Comparison

Table 2 is, in our assessment, the empirical core of this piece: a comparably large macro shock (a 10-year yield move in the tens of basis points, a VIX spike above 30, a sharp equity drawdown) produced a materially smaller Treasury-market liquidity deterioration in April 2025 than in March 2020, with the clearest evidenced explanation being that basis-trade and swap-spread positions, while genuinely unwound to a meaningful degree ($60bn plus $40bn), did so without the disorderly, self-reinforcing repo-funding spiral that characterized March 2020. The ECB's own November 2025 Financial Stability Review corroborates this from an independent, non-US-institution perspective, describing the April 2025 swap-spread move as 'partly driven by hedge funds unwinding leveraged positions amid trade uncertainty and tighter liquidity, possibly also linked to foreign capital flight' -- a genuine, if more contained, instance of the same mechanism, not evidence the mechanism has been eliminated.

8. Why April 2025 Was Contained: The Standing Repo Facility

The Federal Reserve established its Standing Repo Facility in July 2021, explicitly in response to the lessons of March 2020, to provide primary dealers and eligible counterparties with a reliable, pre-authorized source of overnight repo funding against Treasury and agency MBS collateral. The Dallas Fed's own research finds no evidence of the kind of implied-versus-term-repo-rate distortion in April 2025 that characterized March 2020, and attributes this improved resilience to the availability of repo funding and intermediation capacity remaining stable through the April 2025 volatility -- an interpretation we find well-supported by the data in Table 2 but distinguish explicitly as the Dallas Fed's own causal attribution rather than a directly observed, controlled-experiment result, since April 2025 and March 2020 also differed in shock type, market starting conditions, and the broader policy and communication environment.

9. Monetary Policy Transmission and the Question of a Treasury Market 'Put'

This programme's own Federal Reserve communication companion research documented a further, more recent complication directly relevant to this piece: the Federal Reserve's own deliberate reduction in forward guidance under Chair Warsh has been associated with a genuinely large, real-yield-driven bear-steepening episode around the July 2026 FOMC meeting (10-year yield +14bp, 30-year +18bp, over a meeting where the policy rate itself did not move), even as that episode's VIX and credit-spread reaction remained orderly and did not, on the evidence in that companion piece, show signs of the kind of leveraged-unwind dynamic examined in this paper. We would characterize the open question this section's brief poses -- does the Fed have an implicit Treasury-market 'put,' and does expectation of Fed intervention encourage additional leverage -- as a genuine, unresolved moral-hazard concern rather than a settled fact: the Standing Repo Facility (Section 8) is explicitly designed as a liquidity backstop, and its mere existence and pre-authorization plausibly does reduce the marginal cost, and therefore plausibly increases the marginal willingness, to run leveraged Treasury strategies relative to a world without such a backstop -- though we do not have data directly measuring this specific behavioural response and flag the claim as a reasoned inference rather than a directly evidenced fact.

10. Term Premium and the Changing Investor Base

Table 3, reproduced from this programme's own Federal Reserve communication research, shows a monotonic, curve-wide steepening pattern this programme's separate analysis attributed substantially to real-yield and term-premium dynamics rather than to a simple shift in expected policy rates. We would connect this directly to the investor-base question this section's brief raises: a market increasingly intermediated by price-sensitive, leveraged, non-bank participants (Section 3) is, in principle, more likely to show exactly this kind of curve-wide, term-premium-driven sensitivity to fiscal, geopolitical, and communication-related news than a market dominated by price-insensitive official-sector buy-and-hold demand -- though we would caution that this programme's own global sovereign bond research found broadly parallel yield moves across Japan, Germany, France and the UK over a similar window, none of which share the US market's specific hedge-fund-intermediation structure to the same degree, meaning the changing US investor base is, at most, a contributing rather than sole explanation for the term-premium moves documented in Table 3.

11. Fiscal Policy, Debt Issuance and the Central Paradox

This piece's brief poses what we regard as its most analytically important question directly: could the growing involvement of hedge funds actually allow the US government to issue more debt at lower yields in normal times, while simultaneously increasing fragility during stress? We find the evidence assembled in this piece genuinely consistent with both halves of this proposition holding simultaneously, rather than being in tension. On the first half: a $2.4 trillion pool of long Treasury demand from hedge funds (Table 1), even where partly financed through basis-trade arbitrage rather than genuine buy-and-hold appetite, represents real, additional absorption capacity for the record issuance this programme's global sovereign bond research has documented (US federal debt-to-GDP at 122.6% in 2026, a structural 5-6% of GDP deficit since 2022) -- absorption capacity that did not exist at anything like this scale a decade ago. On the second half: Table 2's direct comparison of March 2020 and April 2025 shows precisely the fragility this same leveraged demand can introduce under stress, even in the more contained April 2025 case. We regard this as a genuine paradox rather than a contradiction to be resolved: both effects are real, and both are, on the evidence in this piece, growing in tandem as the basis trade and broader hedge fund Treasury exposure continue to expand past their pre-pandemic peaks (Table 1).

12. Foreign Investors and the Limits of the De-Dollarisation Narrative

This programme's dedicated dollar-dominance companion research examined the foreign-official-demand side of this question in detail and found the dollar's global reserve share declining gradually (62.2% to 57.1% since 2019) without acceleration, alongside continued active US stewardship of the dollar-centred system (the July 2026 coordinated US-Japan yen intervention, examined in that programme's separate case study). We treat that analysis as directly incorporated here, and would note its direct relevance to this piece's central question: a gradually, not rapidly, declining foreign-official investor base is consistent with hedge funds and other leveraged non-bank intermediaries becoming a structurally larger share of marginal Treasury demand (Section 11) even without any dramatic 'foreign investors are fleeing Treasuries' narrative being required to explain it.

13. Banks, Dealers and Non-Bank Intermediation

Primary dealer balance-sheet capacity has, per widely documented post-2008 regulatory changes (Basel III capital and leverage-ratio requirements applied to bank-affiliated dealers), grown more slowly than the Treasury market itself over the past decade -- a structural constraint this piece's brief correctly identifies as a plausible contributor to the rise of non-bank, hedge-fund intermediation examined throughout this piece, since dealers facing binding balance-sheet constraints have reduced capacity to warehouse Treasury inventory and instead increasingly rely on other market participants, including leveraged relative-value funds, to perform functions dealers historically performed more directly themselves. We do not have dealer-balance-sheet time series in our own database to quantify this relationship directly and flag it as a well-established, directly-cited structural feature of the post-2008 regulatory environment rather than an independently measured finding of this piece.

14. Systemic Risk: The Transmission Chain

The transmission chain this piece's brief specifies -- a Treasury shock triggering margin increases, repo funding pressure, forced selling, deteriorating Treasury liquidity, a wider basis, further margin calls, and potential contagion into broader asset markets -- is, on the evidence in Section 6, precisely the mechanism that operated in March 2020, and, on the evidence in Sections 7-8, operated only partially and in a contained fashion in April 2025 specifically because the repo-funding link in that chain (Section 8) did not show the same distortion. We would characterize the chain as a genuine, evidenced risk architecture rather than a purely hypothetical one, with the Standing Repo Facility functioning, on the evidence available, as a specific circuit-breaker at the repo-funding-pressure link -- meaning the chain's remaining links (margin calls, forced selling, basis widening) remain live risks even where the specific link this reform targeted has, so far, held.

15. Global Spillovers

This programme's global sovereign bond regime research found broadly parallel, simultaneous sovereign yield increases across Japan, Germany, France, the UK and the US since 2020 (265-429 basis points depending on market), evidence of a shared global term-premium dynamic that a US-specific Treasury-market dysfunction episode would plausibly transmit into and interact with, given how closely global sovereign yield curves have moved together over this programme's entire research programme this year. We would flag this as a genuine channel through which a future, larger-scale version of the systemic risk chain in Section 14 could propagate beyond the US Treasury market specifically, into the broader global fixed-income complex this programme has documented extensively in its companion research.

16. The 'Safe Haven' Paradox

The conceptual core of this piece is the proposition that US Treasuries can simultaneously be the world's most important safe asset and a source of systemic financial instability -- not as a contradiction, but as two faces of the same underlying fact: Treasuries' status as the deepest, most liquid, most widely used collateral asset in global finance is precisely what makes them the instrument of choice for the leveraged relative-value and financing strategies examined throughout this piece, and that same depth and ubiquity is what makes a leveraged unwind in Treasuries, when it occurs, capable of transmitting stress across the entire financial system rather than remaining contained to a narrower asset class. This is, in our assessment, this piece's most important original conceptual contribution: the safe-haven property and the fragility property are not offsetting characteristics to be netted against each other, but two consequences of the same structural fact about how central Treasuries are to the plumbing of global finance.

17. Policy Options

We would not advocate for direct leverage or position limits on Treasury basis trades without further evidence, consistent with this piece's brief to avoid endorsing a policy merely because it appears in official discussion: Table 4's own benefit/risk framing shows a genuine trade-off between reducing leverage-driven fragility and preserving the liquidity-provision and government-financing benefits the same leverage provides in normal conditions (Section 11) -- a trade-off this piece's evidence cannot resolve on its own.

18. Investment Implications

In normal market conditions, the evidence in this piece is consistent with continued, deep Treasury market liquidity supporting relatively tight bid-ask spreads and efficient price discovery across the curve, a genuinely supportive backdrop for institutional Treasury allocation. Under Treasury-market stress conditions of the kind examined in Sections 6-7, our own data and this programme's companion research point to a specific, differentiated set of consequences: a sharp, if now more contained (per Table 2), liquidity deterioration in the specific Treasury tenors most implicated in basis-trade unwinds; a flight toward, rather than away from, Treasuries at the very front end and toward cash-equivalent instruments, even as the specific securities most embedded in leveraged strategies see the largest liquidity deterioration; a corresponding widening in swap spreads (Table 2); and, per this programme's global sovereign bond research, likely spillover into other major sovereign bond markets given the shared global term-premium dynamics documented there. We would flag gold and, to a lesser extent, the dollar itself as the assets most likely to benefit from a genuine Treasury-market stress event specifically, distinct from a more conventional risk-off equity selloff, given gold's role as a hedge against exactly the kind of core-market dysfunction this piece examines rather than only against inflation or currency risk.

19. Conclusion

Has the Treasury market become too dependent on leveraged non-bank intermediation? The evidence in this piece does not support a simple yes. It supports a more precise, and in our assessment more useful, characterization: hedge fund intermediation, now at $4.0 trillion in gross exposure and still growing past its pre-pandemic peaks (Table 1), simultaneously deepens Treasury market liquidity and helps absorb record government issuance in normal conditions (Section 11), while remaining the specific mechanism identified by the Federal Reserve, the Dallas Fed and the ECB as the principal amplification channel in both the market's worst recent dysfunction episode (March 2020) and its most recent, more successfully contained stress test (April 2025, Table 2). The Standing Repo Facility appears, on the evidence in Section 8, to have genuinely improved the system's resilience at one specific, critical link in the transmission chain (Section 14) -- but the underlying leverage this piece has documented throughout has grown, not shrunk, even as that resilience has improved. The question this piece's brief poses is, in our final assessment, correctly framed not as whether hedge funds are good or bad for the Treasury market, but as whether the current architecture -- genuinely more resilient than in 2020, genuinely more leveraged than ever -- is sufficient for a shock larger than either of the two this piece has examined in detail.