Key Highlights

• Quarter-end repo rises, SRF tapped at quarter-end
• SLR reform presses ahead, but we view the reaction in markets as mostly played out
• Cross-border outflows from USTs have stabilized since March, but will inflows resume?

View chart pack

Repo rates climb at quarter-end, SRF tapped; normalization before the T-bill rebuild

EXHIBIT #1: MONTH-END INCREASES IN GC REPO

_{Source: BNY Markets, Bloomberg}

Month- and quarter-end pressure reappeared yesterday in money markets, with the general collateral repo rate printing well over 4.50% on Monday. The Federal Reserve’s standing repo facility (SRF), since last week open for business at 8:15 AM, received a little over $5bn in usage on Monday morning, while the long-standing afternoon offering pulled in $6bn. While that total of $11bn is small in the context of overall repo operations, it still represents one of the largest take-ups seen since the SRF was introduced.

Period-end increases in funding rates have been occurring with increasing frequency in recent months, and with larger back-ups in market rates. Exhibit #1 plots the simple spread between the tri-party general collateral rate and the offering rate on the Fed’s reverse repo facility on the last trading day of each month. This spread, generally slightly negative for most of 2023, has been rising steadily, posting particularly wide readings during 2025.

Month- and quarter-end funding hiccups in money market rate suggest that liquidity is slightly challenged during those periods, as does the relatively large usage of the SRF on Monday morning. This is not to say there is a funding problem, merely that excess liquidity in the system is receding. Several weeks ago, Roberto Perli, manager of the Fed’s SOMA portfolio pointed out the month-end developments and characterized them “as a sign that the market is returning toward more normal conditions after a period in which very abundant liquidity suppressed nearly all volatility. For now, these signals from the repo market suggest no cause for concern.”  We don’t disagree with that for now. There is no cause for concern, but the situation does warrant keeping an eye on.

This is especially true as we (hopefully) approach a resolution of the current debt ceiling suspension. Since November, T-bill supply has declined by some $350bn, while the Treasury’s general account (TGA) has fallen to just over $300bn as the government has spent out of its savings. In May 2023, the last time we ran up against an 11th hour debt ceiling impasse, the TGA fell to around $22bn. After resolution of the debt ceiling, the TGA swelled by nearly $850bn by the end of October. We expect a similar rapid ramp-up in bill supply after this year’s episode is over, with perhaps as much as $1tn in issuance.

This will lower reserves, currently sitting comfortably at well over $3tn, and could impart additional liquidity stresses into money markets. Repo rates will likely be higher and more sensitive in this case and may not necessarily mean markets are functioning under strain. The opening of the SRF for morning operations, when market participants’ liquidity needs are most apparent, was intended to address this eventuality.

SLR reform: Close at hand, but short-term effects could underwhelm

EXHIBIT #2: SWAP SPREADS HAVE ALREADY MOVED IN RESPONSE TO SLR REFORM

_{Source: BNY Markets, Bloomberg}

Momentum is building toward an eventual reduction in the enhanced Supplementary Leverage Ratio (SLR) paid by GSIBs on their risky asset holdings. Last week both the Fed and the FDIC recommended reducing this capital charge, as has been expected in recent months. We wrote on this topic several weeks ago, when discussions on SLR reform were about to commence. Merely broaching the discussions and hinting at its reform led to a widening of the 10y SOFR swap spread to nearly –16bp between March 10 and April 8, with a large 8bp move wider on April 8 alone. The announcement’s effect at that time probably represents the bulk of the market response to proposed SLR reform (see Exhibit #2).

In our piece from early June, we argued that we thought it would be unlikely for SLR reform to make a material difference in 10y rates or dealers’ willingness to hold more coupons on their balance sheet during normal, stress-free times. Yes, broker dealers’ holdings of Treasurys are historically high, at about 15% of all commercial bank assets, suggesting banks are close to capacity. In theory, reducing the SLR for Treasurys should allow for more take-up by the dealers, but in practice, we remain skeptical.

It could free up capital for banks to add other risky assets (like loans) to their balance sheets, or it might – especially given the T-bill supply dynamics we referenced above – induce banks to hold more in bills, which carry less interest rate risk than longer-dated notes and bonds. Loan demand is falling and loan standards (supply) are rising, so SLR relief is unlikely to unlock additional lending. Banks may already be internally constrained or have other capital requirements and internal policies which make adding coupon risk to their balance sheet unattractive.

To be sure, SLR reform could come in handy in times of market stress, allowing dealers to take on more UST risk during illiquid periods or other disruptions in the Treasury market, giving them more room to intermediate the market. Thus, it seems to us that SLR reform – no matter how it is presented to the public – is more of a macroprudential initiative than it is a way to lower the yields on long-dated USTs.

Cross-border Treasury outflows stabilize – what’s next?

EXHIBIT #3: FLATTENING OUT OF CROSS-BORDER UST FLOWS

_{Source: BNY Markets, iFlow}

Since the intense period of cross-border UST selling during April’s market unpleasantness, foreign appetite for U.S. sovereign debt has stabilized. Exhibit #3 shows cumulative UST purchases since the beginning of June 2024. By reporting the evolution of flows in this manner, we can, from the contours of the blue area of the graph, see trends, inflection points and turning points in cross-border demand for USTs.

We reiterate that since the end of June last year, foreign selling of Treasurys has been a year-long phenomenon, admittedly with some reversals in this trend. April 2025 represents one, but by no means the only period in which we saw concentrated selling by foreigners. Notably, such selling also occurred in August 2024 (during the market swoon early in that month), the period around the presidential election in November 2024, and in early 2025, after the inauguration and the beginning of tariff talk. Since April’s large downturn in demand, there has been some net buying, but not enough to take the change in positions since last summer back to positive.

Generally speaking, however, demand has been tepid at best. At worst, it has been waning. The rest of the year will tell, especially if the “Big Beautiful Bill” passes, how attractive U.S. sovereign debt is for investors overseas, or whether or not perceived risk remains high enough to discourage a return to the market in any meaningful way. Our view is that with yields not quite breaching 4.5% at the 10y horizon, pricing might not be sufficiently attractive for new bond buyers, especially if risks and volatility remain elevated.

Chart pack