Banks play an important role in Treasury markets, where investors trade federal debt. Congress has focused on Treasury market resilience, which is necessary to finance federal borrowing and maintain broader financial stability. The enhanced supplementary leverage ratio (eSLR) is a capital requirement that applies to the eight globally systemically important banks (G-SIBs). In June 2025, the federal bank regulators proposed a rule to reduce the eSLR to make it "a backstop to risk-based capital requirements" that does "not discourage [the G-SIBs] from engaging in low-risk activities," such as Treasury market making. Improved market making could make Treasury markets less fragile. But lower capital requirements could make the G-SIBs more likely to fail and cause financial instability.
Treasury markets are typically highly liquid—trading is relatively robust and at low cost. However, Treasury cash or lending markets have recently had short-lasting bouts of instability in 2014, 2019, 2020, and 2025, where liquidity suddenly dried up. In each case, calm was quickly restored, sometimes through injections of liquidity by the Federal Reserve (Fed). None of these episodes had broader, lasting negative effects on the financial system.
Historically, Treasury markets were less regulated than other securities markets. In addition to the June eSLR proposal, policymakers have initiated a series of reforms to make Treasury markets more resilient. Some of these reforms have been largely implemented, such as greater transparency, debt buybacks, and new standing Fed facilities. Some have been finalized but are not yet in effect, such as more central clearing. Some have been abandoned, such as registration requirements for market participants.
Most banks hold Treasuries as investments. Large banks are required to hold a minimum amount of liquid assets, such as Treasuries. Banks holding Treasuries as long-term investments does not affect Treasury market resilience.
In addition, most primary dealers, who are the largest broker-dealers in Treasury markets, are owned by either large domestic or foreign banks. Broker-dealers (also called Treasury dealers in the context of Treasury markets) are market makers in bond markets: They hold an inventory of securities to fulfill client orders to buy, sell, borrow, and lend. For the market to be liquid, broker-dealers must be willing to increase their inventories when more clients want to sell and decrease their inventories when clients want to buy. Broker-dealers are not the only major institutional participants in the Treasury market, but they are the only ones that routinely act as market makers.
Capital adequacy is one of the primary safeguards against bank insolvency. Safety and soundness regulation requires banks to maintain various capital-to-asset ratios or face remedial actions. One of these capital requirements is the leverage ratio. Unlike risk-weighted capital ratios, in the leverage ratio all assets are counted at full value.
The G-SIBs must meet an eSLR of 5% company-wide and 6% for their bank subsidiaries to avoid restrictions on dividends, buybacks, and bonuses. The eSLR includes assets and (unlike the leverage ratio) off-balance-sheet exposures in the denominator. It is the ratio of Tier 1 capital in the numerator and unweighted exposures in the denominator. Tier 1 capital is high-quality, loss-absorbing forms of capital, such as common equity. For more information, see CRS Report R47447, Bank Capital Requirements: A Primer and Policy Issues.
The G-SIBs are the eight banks (see Table 1) that regulators believe pose the greatest risk to financial stability if they were to fail. They are subject to the most stringent safety and soundness regulations, including the eSLR and an additional G-SIB surcharge, which is added to their risk-weighted capital requirements based on each's systemic importance. For more information, see CRS Report R47876, Enhanced Prudential Regulation of Large Banks.
Banks must simultaneously comply with multiple capital requirements, some risk-weighted and some unweighted. At a given point in time, whichever one requires the most capital is the binding requirement for any given bank. Leverage ratios, such as the eSLR, require banks to hold the same amount of capital against relatively low-risk assets (such as Treasury securities) as they are required to hold against high-risk assets. But leverage ratios discourage banks from holding Treasuries only if they are the binding (or close to binding) requirement. If risk-weighted requirements are binding, then banks face little disincentive to hold Treasuries, because their risk weight is zero.
G-SIBs are also required to hold minimum total loss absorbing capacity (TLAC) at the holding company level that is composed of Tier 1 capital and long-term debt. TLAC is intended to "bail in" creditors and recapitalize a failed G-SIB. There are currently weighted and unweighted TLAC requirements—the latter is tied to the eSLR.
The proposal would replace the fixed 5%/6% eSLR with a sliding scale based on the G-SIB's Method 1 surcharge, which currently yields surcharges that range from 1% to 2.5%. (The proposal would reduce required capital more than a similar 2018 proposal that was not finalized, which was based on the actual G-SIB surcharge.) Under the proposal, the eSLR would be reduced to 3% plus 50% of each G-SIB's Method 1 surcharge, yielding an eSLR that ranged from 3.5% to 4.25% for their 2024 surcharge (see Table 1.) In addition, bank subsidiaries would no longer face a higher eSLR than the holding company does. A corresponding reduction based on 50% of the Method 1 surcharge would be made to the leverage-based TLAC requirements. The proposal parallels international Basel standards, removing what critics call U.S. "gold plating."
|
G-SIB |
2024 Method 1 Surcharge |
Proposed eSLR (50%*Method 1 Surcharge+3) |
|
JP Morgan Chase* |
2.5 |
4.25 |
|
Citigroup* |
2 |
4 |
|
Bank of America* |
1.5 |
3.75 |
|
Goldman Sachs* |
1.5 |
3.75 |
|
Bank of NY Mellon |
1 |
3.5 |
|
Morgan Stanley* |
1 |
3.5 |
|
State Street |
1 |
3.5 |
|
Wells Fargo* |
1 |
3.5 |
Source: CRS calculations based on Financial Stability Board data.
Note: * = has primary dealer.
According to the proposal, the eSLR was the binding constraint for five of the eight holding companies and eight of the nine major bank subsidiaries of the G-SIBs (one G-SIB has two) in 2024. The proposal would result in a large reduction in how much capital G-SIBs would be required to hold under the eSLR. However, the eSLR would (as intended) no longer be the binding capital requirement for most G-SIBs, so the overall amount of capital G-SIBs would now be required to hold would be determined by one of their other capital requirements. Once this effect is netted out, the G-SIBs could reduce their capital under the proposal by a cumulative $13 billion, ranging by bank from 0% to 7.4%. The net decline in required capital at the bank subsidiaries is significantly larger—$213 billion, ranging from 0% to 37%. However, the G-SIBs cannot reduce their overall capital levels by $213 billion—they can only reduce them by $13 billion and shift the other $200 billion from the bank subsidiary to the parent company, where it can be used to finance activities at other subsidiaries or to pay down external debt. Shifting capital does not make the bank more likely to fail so long as existing "source of strength" requirements are effective—which critics have disputed.
The proposal would also reduce required TLAC by $90 billion. This also interacts with the capital requirements, however, so most of this reduction would come from long-term debt, not capital. Less long-term debt does not increase the likelihood of failure but leaves fewer resources to fund a bail in, making a government bailout more likely.
Bank and primary-dealer holdings of Treasuries have grown more rapidly than have Treasuries outstanding since 2014—prima facie evidence at odds with the argument that the eSLR is undermining Treasury market stability, as the eSLR has been in place since 2018. Banks must classify their securities as for trading, available for sale, or held to maturity. Only the trading securities directly affect market making and resiliency. In 2024, G-SIBs held $1.7 trillion of Treasuries, of which 33% were categorized as for trading, which were mostly with their broker-dealer subsidiaries.
Reducing the eSLR affects Treasury market making only if the G-SIB owns a primary dealer and the eSLR is binding. The six largest Treasury dealers are owned by G-SIBs. Under the proposal, the eSLR would no longer be binding for all G-SIBs with dealer subsidiaries. As an upper limit estimate of the proposal's effect, G-SIBs would have the capacity to add $2.1 trillion Treasuries (or other zero risk weighted assets) with their current Tier 1 capital.
Because one of the two main arguments made for lowering the eSLR is to remove a disincentive for G-SIBs to hold Treasuries, the proposal could have exempted Treasuries from the eSLR instead, which would remove the disincentive whether or not the eSLR is binding. The regulators estimated that exempting Treasuries and bank reserves (as was done temporarily during COVID-19 for the SLR) would lead to a smaller reduction in capital and a greater increase in capacity to hold Treasuries than the proposal. However, Treasuries are not risk free, and this alternative would undermine the eSLR's role as a neutral backstop. The proposal also combines a lower eSLR and an exemption for trading Treasury securities of dealer subsidiaries, which would reduce required capital further.
Unweighted capital requirements such as the eSLR were not intended to be binding, because then risk weights do not have the intended effect of discouraging excessive risk taking. To make them not binding, policymakers could lower unweighted or raise weighted requirements or choose a combination of the two that left aggregate required capital roughly unchanged. Bank regulators proposed to reduce the eSLR so that it would not be binding for most G-SIBs, which would lower required capital by a range of 0%-7.4%, thus making the probability of failure slightly higher for some G-SIBs. This higher probability can be weighed against the gain in Treasury market resiliency. Reducing the eSLR is arguably a blunt tool to indirectly improve resiliency—Treasury market making is a small part of G-SIBs' overall business model—only six primary dealers are owned by G-SIBs—and encouraging market making is only one tool to strengthen resiliency.