Quick answer: leverage constraints ask a deliberately different question from risk-weighted capital: how large is the bank’s exposure relative to high-quality capital, without relying entirely on modelled risk weights? The Basel leverage ratio divides Tier 1 capital by a broad exposure measure that includes on-balance-sheet assets plus specified derivatives, securities-financing and off-balance-sheet exposures. Its purpose is to act as a backstop to risk-based capital requirements. In the United States, the supplementary leverage ratio (SLR), enhanced SLR (eSLR) for systemically important firms, the ordinary Tier 1 leverage ratio and the optional Community Bank Leverage Ratio (CBLR) are related but distinct frameworks. The algorithmic lesson is that a low-risk asset can consume substantial leverage capacity even when it consumes little risk-weighted capital.

Risk-weighted capital asks how dangerous an asset is. A leverage ratio also asks how much balance sheet exists at all.

Page role: this article owns the non-risk-based capital constraint. It is complementary to How Bank Capital Models Turn Risk Into Constraints, which owns risk-weighted assets and capital ratios, and to balance-sheet optimisation, which uses several constraints at once.

1. The core ratio

The Basel leverage ratio is:

Leverage ratio = Tier 1 capital / leverage exposure measure.

The Basel minimum is 3%, subject to additional requirements for global systemically important banks under the framework. See the current Basel leverage-ratio framework.

If a bank has S$60 billion of Tier 1 capital and S$1.5 trillion of leverage exposure, its ratio is:

60 / 1,500 = 4.0%.

If exposure grows to S$1.8 trillion while Tier 1 capital is unchanged, the ratio falls to 3.33% even if the new assets are low-risk government securities.

2. Why a leverage ratio exists beside RWA

Risk-weighted capital can assign very different weights to a sovereign bond, mortgage, unsecured corporate loan and trading exposure. That is useful because their credit and market risks differ. But risk weights also depend on rules, models and classifications that can be imperfect.

A leverage ratio provides a simpler backstop by keeping a broad measure of exposure in the denominator. It therefore reduces the chance that a bank can become extremely large relative to capital merely because its assets receive low measured risk weights.

The two constraints are not rivals. They test different failure modes:

Constraint	Main question
Risk-weighted capital	How much capital is held relative to measured risk?
Leverage ratio	How much Tier 1 capital supports the overall exposure footprint?

3. The denominator is broader than accounting assets

The Basel exposure measure contains four broad components:

on-balance-sheet exposures;
derivative exposures;
securities-financing transaction (SFT) exposures such as repo;
off-balance-sheet items converted using specified credit-conversion factors.

This means “total assets” is not enough to reconstruct the leverage ratio. A bank with large derivatives, commitments or securities-financing activity can have a leverage exposure measure materially different from its simple accounting balance-sheet total.

4. On-balance-sheet exposure creates the simplest capacity trade-off

Suppose a bank can add either S$10 billion of highly rated government securities or S$10 billion of riskier loans. Their RWA can differ substantially, but before detailed adjustments their gross balance-sheet footprint is similar in leverage terms.

If leverage rather than RWA is the binding constraint, the bank can prefer the activity that produces more economic return per unit of leverage exposure—even if both use the same gross denominator capacity.

A simplified leverage-efficiency measure is:

Return on leverage exposure = risk-adjusted income / incremental leverage exposure.

This is not a regulatory ratio; it is an internal teaching metric that explains why a non-risk-based constraint can influence product economics.

5. Repo and securities financing need their own exposure treatment

A repo is collateralised, but it still creates balance-sheet and counterparty exposure. The leverage framework therefore contains specific SFT treatment rather than simply assuming high-quality collateral makes the denominator disappear.

Netting can matter when transactions meet required criteria, and counterparty exposure can depend on the relationship between cash lent/received and securities delivered/received. The exact regulatory calculation is detailed, but the key computational point is:

SFT leverage exposure ≠ repo notional alone and ≠ zero because collateral exists.

This connects to repo pricing: leverage capacity can become part of the repo desk’s all-in price.

6. Derivatives can create exposure without full notional entering the denominator

A S$1 billion interest-rate swap notional is not economically equivalent to lending S$1 billion of cash. The leverage framework therefore uses a derivatives exposure calculation that includes current replacement cost plus an add-on for potential future exposure, with specified recognition of eligible cash variation margin and other rule-defined adjustments.

The model distinguishes:

notional used to calculate contractual cash flows;
current mark-to-market exposure;
potential future exposure;
collateral/margin that qualifies for specific recognition.

This links to How Banks Calculate Derivatives XVA, but the jobs differ: XVA adjusts derivative value for credit, funding and margin economics; leverage rules define a regulatory exposure measure.

7. Off-balance-sheet commitments are not “zero exposure”

A bank can promise a revolving credit line without funding the full amount today. Leverage rules therefore convert specified off-balance-sheet commitments into exposure using rule-defined credit conversion factors rather than ignore the unused line.

Conceptually:

Leverage exposure from commitment = off-balance-sheet amount × applicable conversion factor.

The exact factor depends on the applicable framework and exposure type. This connects to revolving-credit EAD: both recognise that an unused line can become real exposure, but they use it for different purposes.

8. SLR and the ordinary Tier 1 leverage ratio are not interchangeable

US banking rules contain more than one leverage measure. The ordinary Tier 1 leverage ratio uses Tier 1 capital relative to an average-asset measure under the applicable capital rule. The SLR uses a broader total leverage exposure measure that also captures specified derivatives, SFTs and off-balance-sheet items.

A reader should therefore not compare a 7% Tier 1 leverage ratio with a 5% SLR as if one were a simple recalculation error. They use different denominators and apply to different regulatory contexts.

9. The enhanced SLR changed in 2026

US regulators finalised changes to the enhanced supplementary leverage ratio in November 2025, effective 1 April 2026. The stated objective was to make the enhanced leverage requirement operate more consistently as a backstop to risk-based capital rather than frequently become the binding constraint for US global systemically important banks and their covered depository institution subsidiaries.

For covered bank subsidiaries, the final rule replaced the former fixed 3-percentage-point enhancement above the 3% SLR with an enhancement based on the controlling GSIB’s Method 1 surcharge, subject to a 1-percentage-point cap. See OCC Bulletin 2025-41 and the joint-agency final-rule summary.

The educational significance is not political: a changed constraint changes the shadow price of balance-sheet capacity. Repo, Treasury intermediation and other low-risk, low-margin activities can become more or less attractive when the leverage requirement changes.

10. The Community Bank Leverage Ratio is a separate simplification

The CBLR is an optional simplified capital framework for qualifying US community banking organisations, not a smaller version of the GSIB SLR.

On 1 July 2026, final US rules lowered the CBLR threshold from greater than 9% to greater than 8% and extended the grace-period structure for qualifying institutions. The FDIC says the framework applies to qualifying institutions with less than US$10 billion in total consolidated assets that meet the other eligibility criteria. See the FDIC final-rule summary and the OCC 2026 compliance guide.

This is a useful example of authority boundaries: the same phrase “leverage ratio” can refer to different calculations, populations and legal consequences.

11. Binding constraint: the smallest remaining slack matters

A bank can satisfy many ratios at once. Define slack for constraint k as:

Slack_k = actual ratio − required/internal minimum.

If CET1 risk-weighted capital has 4 percentage points of slack but SLR has only 0.4 points, leverage can be the economically binding constraint. The next low-risk asset may therefore be limited by leverage even though it barely moves RWA.

This is why balance-sheet optimisers calculate marginal constraint usage rather than ask only whether each ratio passes today.

12. Shadow price of leverage capacity

Suppose one extra S$1 billion activity uses S$1 billion of leverage exposure and produces S$6 million of annual risk-adjusted income. If leverage capacity is abundant, that can be attractive. If the bank is near its leverage floor, the same S$1 billion may displace another activity worth S$10 million.

The implicit opportunity cost—the shadow price—is therefore at least the value of the displaced scarce capacity.

A pricing desk can incorporate this through an internal charge:

Economic return = direct revenue − funding − expected loss − operational cost − leverage-capacity charge − other scarce-resource charges.

13. Why leverage can affect Treasury and repo intermediation

Treasury securities and secured repo can carry low credit risk while still expanding total leverage exposure. When leverage is binding, low-margin intermediation can compete for the same denominator capacity as higher-return businesses.

US regulators explicitly cited this issue when revising the eSLR, stating that the change was intended to reduce disincentives for low-risk activities important to market functioning, including Treasury intermediation and repo financing. See the FDIC 2026 regulatory update.

The mathematical point is broader than any one policy choice: a constraint based on gross exposure can affect business mix differently from a constraint based on measured risk.

14. Inputs and outputs

Inputs	Outputs
Tier 1 capital	Leverage numerator
On-balance-sheet exposure	Base denominator
Derivatives and margin data	Derivative exposure amount
Repo/SFT positions and netting sets	SFT exposure amount
Off-balance-sheet commitments	Converted exposure amount
Applicable rule set and institution category	Required ratio and buffer
Business revenue/cost data	Marginal leverage efficiency / shadow price

15. Failure modes

Assets-only denominator. Derivatives, SFTs and commitments are omitted.
RWA substitution. Risk-weighted exposure is incorrectly used in a leverage-ratio denominator.
Ratio-name confusion. Tier 1 leverage, SLR, eSLR and CBLR are treated as one formula.
Low-risk invisibility. Treasury or repo activities are assumed to consume no leverage capacity because credit risk is low.
Notional error. Derivative notional is used directly without the applicable regulatory exposure method.
Netting overreach. Offsetting positions are netted without satisfying rule criteria.
Static-constraint pricing. Business charges ignore whether leverage is actually scarce.
Regulatory-date drift. A model uses pre-April or pre-July 2026 US leverage settings after the rules changed.

16. Diagnostics, alternatives and falsifiers

Which denominator component contributes most to total leverage exposure?
How much leverage slack remains after a proposed transaction?
Is leverage or risk-weighted capital currently the tighter constraint?
What happens to SLR if the bank adds S$10 billion of low-risk government securities?
How much SFT exposure disappears through permitted netting?
Would the same business decision change if leverage capacity were abundant?
Is the institution actually eligible for the ratio framework being applied?
Does the regulatory engine contain the current effective-date parameters?

Falsifier: suppose someone claims, “A zero-risk-weight sovereign asset cannot make a bank’s capital constraint worse.” A counterexample is a bank for which the leverage ratio is binding: adding the asset can expand total leverage exposure and reduce the leverage ratio even if risk-weighted capital barely changes.

17. Verification and update triggers

Reconcile Tier 1 capital to regulatory capital reporting.
Reconcile exposure components to balance-sheet and off-balance-sheet systems.
Independently reproduce derivatives and SFT exposure calculations.
Check netting eligibility rather than assuming economic offsets qualify.
Re-run business shadow prices when capital or exposure changes materially.
Update rules on their legal effective dates, not announcement dates alone.
Compare leverage projections with stress-test outputs where applicable.
Keep risk-weighted and leverage constraints separate in dashboards and explanations.

Research anchors

The deeper lesson

Leverage constraints are deliberately coarse. That is their strength and their weakness. They protect against risk models becoming too optimistic, but they can make low-risk balance-sheet activities consume scarce capital capacity. The bank therefore lives inside two geometries at once: a risk-weighted geometry and a gross-exposure geometry. A strong capital algorithm knows which one is binding, calculates each denominator honestly, and never mistakes “low measured risk” for “no balance-sheet cost.”

Educational note: This article explains public bank-capital mathematics and current regulatory concepts. It is not regulatory advice, capital-planning advice or an assessment of any specific bank.

Small Group Tutorials

How Banks Calculate Leverage Constraints: Tier 1 Capital, Total Leverage Exposure, SLR, eSLR, CBLR and Balance-Sheet Trade-offs