Reader question: A bank can pass a 30-day liquidity test and still depend too heavily on fragile short-term funding. How does the Net Stable Funding Ratio test whether the structure of the balance sheet is funded on a more durable one-year basis?
The Basel Net Stable Funding Ratio (NSFR) is a weighted balance-sheet algorithm. It compares how much of the bank’s capital and liabilities are considered reliably available over a one-year horizon with how much stable funding its assets and off-balance-sheet activities are considered to require.
The core equation is:
NSFR = Available Stable Funding (ASF) / Required Stable Funding (RSF).
Under the Basel standard:
NSFR ≥ 100%.
The ratio is simple. The computational difficulty lies in classifying every balance-sheet position by counterparty, residual maturity, liquidity, encumbrance, deposit stability and derivative treatment before applying the prescribed factors.
What this page owns — and what it does not
This page owns the transformation:
capital + liabilities + assets + off-balance-sheet positions → ASF and RSF weights → one-year funding-stability ratio.
It does not replace the Liquidity Coverage Ratio, which owns the 30-day stressed-liquidity buffer; matched-maturity funds-transfer-pricing algorithms, which allocate internal funding costs; or collateral optimisation, which allocates actual collateral assets under eligibility and haircut constraints.
This is regulatory-liquidity mathematics and education. It is not an assessment of any individual bank and not financial advice.
Why NSFR exists
The Basel Committee describes the NSFR as a measure intended to reduce excessive reliance on short-term wholesale funding and encourage banks to fund less-liquid assets with more stable sources of financing.
The one-year horizon complements the LCR. LCR asks whether the bank has enough usable liquidity to survive a severe 30-day outflow scenario. NSFR asks whether the balance sheet is structurally funded in a way that is less vulnerable to a prolonged funding disruption.
Step 1: calculate Available Stable Funding
ASF is not the face amount of every liability. Each category receives an ASF factor between 0% and 100% reflecting the Basel assumption about how reliably that funding remains available over the one-year horizon.
Conceptually:
ASF = Σ carrying value of funding item × ASF factor.
More stable funding sources receive larger weights.
Selected Basel ASF factors
The current Basel application guidance summarises several important categories:
- 100% ASF: qualifying regulatory capital and certain liabilities with effective residual maturity of one year or more;
- 95% ASF: qualifying stable retail and small-business deposits;
- 90% ASF: qualifying less-stable retail and small-business deposits;
- 50% ASF: several categories including shorter-term funding from non-financial corporates and certain sovereign/public-sector sources, operational deposits, and other funding with six months to less than one year residual maturity;
- 0% ASF: other liabilities not assigned positive stable-funding value under the rule.
The full rule contains detailed definitions and exceptions. A production engine must use the jurisdictionally applicable taxonomy rather than this educational summary.
A simple ASF example
Suppose a bank has:
- 100 of qualifying capital at 100%;
- 300 of stable retail deposits at 95%;
- 200 of less-stable retail deposits at 90%;
- 250 of qualifying shorter-term non-financial corporate funding at 50%;
- 150 of funding receiving 0% ASF.
Then:
ASF = 100 + 285 + 180 + 125 + 0 = 690.
The balance sheet contains 1,000 of these funding sources, but only 690 counts as available stable funding under this simplified weighting.
Deposit stability is a classification problem
The 95% and 90% weights depend on whether a retail or small-business deposit qualifies as stable or less stable under the Basel definitions inherited from the LCR framework.
The engine therefore needs more than account balance and maturity date. It can require customer type, deposit-insurance coverage, withdrawal characteristics and relationship information.
A one-line general-ledger category called “retail deposits” cannot by itself determine the ASF factor.
Residual maturity creates step changes
Funding with less than six months, six months to less than one year, and one year or more can fall into different Basel treatment buckets depending on the funding type.
That makes the NSFR a piecewise function of time to maturity. A liability crossing a regulatory maturity boundary can change ASF even if its notional amount is unchanged.
This is a useful diagnostic: a day-on-day NSFR movement can be caused by aging rather than new business.
Step 2: calculate Required Stable Funding
RSF applies a stable-funding requirement to assets and certain off-balance-sheet exposures.
Conceptually:
RSF = Σ carrying value of asset or exposure × RSF factor.
Assets that are very liquid or short-dated generally require less stable funding. Assets that are illiquid, long-dated or encumbered generally require more.
Selected Basel RSF examples
The Basel application guidance assigns low RSF factors to the most liquid assets and higher factors to less-liquid assets. Selected examples include:
- very low or zero RSF for certain cash and central-bank-reserve categories under the standard;
- 5% for certain unencumbered Level 1 HQLA;
- 15% for qualifying unencumbered Level 2A HQLA;
- higher prescribed factors for less-liquid securities, loans and other assets;
- 100% RSF for assets encumbered for one year or more.
The point is not that every loan is “100% illiquid.” The factors are supervisory weights designed to require a funding structure consistent with the asset’s liquidity and maturity characteristics.
Encumbrance changes the funding requirement
An asset can be high quality yet unavailable for funding because it is pledged or otherwise encumbered.
Basel’s NSFR rules therefore adjust RSF for encumbrance duration. The current framework states that assets encumbered for one year or more receive a 100% RSF factor. Assets encumbered for between six months and less than one year receive treatment depending on the factor they would otherwise attract.
So:
same security + different encumbrance horizon → different RSF.
Why encumbrance is not merely a collateral-system field
Suppose a government bond normally attracts low RSF because it is highly liquid. If it has been pledged for 18 months and cannot be used to raise funding, its structural funding need changes materially.
A stale collateral feed can therefore overstate NSFR even when the security master and market value are correct.
Step 3: include off-balance-sheet funding needs
NSFR is not limited to recognized accounting assets. Certain off-balance-sheet commitments receive required stable funding treatment because a future draw can create funding needs.
This mirrors a broader principle seen in liquidity regulation:
contractual optionality can matter before cash is actually drawn.
The engine therefore needs committed-but-undrawn exposures, not only funded loans.
Derivatives require special treatment
Derivatives do not fit naturally into a simple “asset maturity versus liability maturity” framework. They can have positive and negative replacement values, collateral, initial margin, variation margin and legally enforceable netting.
The Basel NSFR therefore defines specific treatment for derivative assets, derivative liabilities and related margin. The 2014 final standard specifically highlighted derivatives and initial margin as areas of calibration.
A safe implementation rule is:
do not substitute accounting net derivative value for the Basel NSFR derivative calculation.
Netting permissions matter
Two derivatives can have equal and opposite accounting values but sit under different legal netting agreements. If the regulatory rule does not permit those positions to offset, collapsing them into one number understates gross funding requirements.
This connects directly to the same legal-netting problem that appears in counterparty-credit calculations such as SA-CCR.
Step 4: compute the ratio
Once ASF and RSF are constructed:
NSFR = ASF / RSF.
Suppose:
ASF = 690
RSF = 650
Then:
NSFR = 690 / 650 ≈ 106.15%.
The bank is above the 100% Basel minimum in this simplified example.
A one-unit balance change can move both numerator and denominator
Suppose the bank originates a long-dated loan funded by issuing a long-dated liability. The new liability can add ASF while the new loan adds RSF.
Whether NSFR rises or falls depends on the relative regulatory weights, not simply on whether total assets increased.
This makes NSFR useful as a balance-sheet-structure metric rather than a size metric.
The marginal NSFR effect
For a new transaction with ASF contribution ΔASF and RSF contribution ΔRSF:
New NSFR = (ASF + ΔASF) / (RSF + ΔRSF).
If the current ratio is above 100%, a trade can still reduce the ratio while remaining compliant. If the current ratio is close to 100%, small changes in funding tenor or asset encumbrance can become material.
The derivative of a ratio is not the same as the ratio of derivatives; desk-level intuition based only on “this funding is 95% stable” can therefore be misleading without considering the denominator effect.
NSFR is not a maturity-matching identity
A bank is not required to fund every individual one-year asset with one specific one-year liability. The algorithm operates on weighted aggregate categories.
This means a portfolio can satisfy NSFR without exact cash-flow matching instrument by instrument.
Conversely, a bank can appear maturity matched in a coarse accounting report yet fail NSFR because the stability, liquidity or encumbrance characteristics receive different regulatory weights.
Inputs and outputs
A production NSFR engine can require:
- legal-entity and consolidation perimeter;
- capital and liability balances;
- customer and deposit classifications;
- contractual and effective residual maturity;
- asset type and HQLA status;
- loan counterparty and performance status;
- encumbrance amount and expiry date;
- securities-financing transactions;
- derivative replacement values and netting-set identifiers;
- initial and variation margin data;
- off-balance-sheet commitments;
- jurisdictional ASF/RSF rule tables.
Outputs can include gross funding, weighted ASF, weighted RSF, NSFR, factor-level decomposition, maturity-bucket migration, encumbrance effects, derivative contributions and data-quality exceptions.
Evidence polarity: what supports confidence?
Evidence for a reliable NSFR calculation includes source balances reconciling to the regulatory perimeter, ASF/RSF mappings traceable to the current standard, deposit stability based on documented attributes, residual maturities generated from validated dates, encumbrance tied to live collateral records, derivative netting aligned with enforceable agreements and explainable period-to-period movements.
Evidence against confidence includes one factor applied to all customer deposits, maturity buckets based on original rather than residual maturity, Level 1 assets receiving low RSF despite long encumbrance, derivative netting across unrelated agreements, unexplained 100% RSF spikes, or a ratio that changes materially when only report ordering changes.
Counterexample: more deposits do not always create the same ASF benefit
Two banks each receive 100 of new deposits. One receives qualifying stable retail funding; the other receives short-term funding with little or no ASF recognition.
The accounting liability increase is identical. The stable-funding contribution is not.
Counterexample: a liquid asset can require high stable funding when encumbered
A government security may normally receive low RSF. If pledged for more than one year, Basel encumbrance treatment can push its RSF to 100%.
Liquidity quality and current availability are different dimensions.
Counterexample: a one-year contractual maturity can be the wrong maturity
If a liability has an embedded option allowing the provider to withdraw earlier without a meaningful penalty, the effective regulatory maturity treatment may differ from its nominal legal maturity.
Date fields alone are therefore insufficient for some products.
Counterexample: LCR and NSFR can point in different directions
A bank can hold abundant HQLA and therefore have a strong 30-day LCR while financing long-dated assets with unstable short-term wholesale funding, weakening NSFR.
Another bank can have a structurally stable one-year funding profile but still face an acute 30-day cash-flow stress if immediate outflows are large.
The two ratios answer different questions.
Weak links in implementation
Original-maturity bug. Original term is used instead of residual maturity.
Boundary bug. Six-month and one-year cutoffs are implemented inconsistently.
Deposit-category drift. Product labels change without updating stability rules.
Encumbrance mismatch. Collateral is released or pledged after the source snapshot but before regulatory extraction.
Derivative netting overreach. Economic hedges are netted despite lacking regulatory/legal permission.
Off-balance-sheet omission. Undrawn commitments are absent from RSF.
Currency/unit error. Source systems report thousands while the calculator assumes units.
Stale factor tables. A jurisdictional implementation changes but the engine does not.
Diagnostics: how to test the algorithm
- factor replay: reproduce Basel application-guidance examples for selected ASF and RSF categories.
- maturity-boundary test: place identical instruments just below and just above six-month and one-year thresholds.
- deposit-stability test: reclassify a retail deposit from stable to less stable and verify ASF falls from the applicable 95% to 90% treatment.
- encumbrance test: extend an asset’s pledge beyond one year and verify the 100% RSF rule applies where required.
- netting test: split two offsetting derivatives into separate legal netting sets and confirm prohibited offsets disappear.
- off-balance-sheet test: add a commitment and verify RSF increases according to the applicable rule.
- aging test: advance the reporting date without changing balances and explain changes caused only by residual-maturity migration.
- reconciliation test: weighted ASF/RSF populations reconcile to unweighted source balances plus documented exclusions.
- ratio test: verify NSFR equals ASF divided by RSF at full internal precision before display rounding.
- parallel implementation test: reproduce a sample portfolio in an independent spreadsheet or code path.
What would falsify confidence?
Confidence should be withdrawn if capital or liabilities receive incorrect ASF factors; if residual maturity boundaries are unstable; if encumbrance does not affect RSF; if derivative netting exceeds the legally recognized netting set; if off-balance-sheet exposures disappear from the denominator; or if the ratio cannot be reconstructed from line-level contributions.
Alternatives and limits
NSFR is a standardized structural-funding metric. It does not replace internal liquidity stress testing, funding concentration analysis, market-access assumptions, behavioral deposit models, funds-transfer pricing or detailed contractual cash-flow ladders.
A bank can satisfy NSFR while still being exposed to a specific funding-provider concentration, a currency mismatch or a scenario more severe than the standardized assumptions. The metric should therefore be read as one constrained projection of funding resilience, not as a complete model of solvency or liquidity.
How this connects to the surrounding knowledge estate
The 30-day complement is the LCR algorithm. Internal pricing of stable funding connects to matched-maturity FTP. Encumbrance depends on collateral allocation. Derivative netting and exposure structure connect to the SA-CCR layer without making the two calculations interchangeable.
Verification and update triggers
Preserve the regulatory version, ASF/RSF lookup tables, maturity logic, deposit-stability mapping, HQLA classifications, encumbrance feed, derivative-netting agreement identifiers and off-balance-sheet inventory. Revalidate after Basel or national rule changes, product redesigns, collateral-platform migrations, legal-netting opinions, core-banking upgrades or unexplained step changes in weighted funding.
Primary and high-quality references
- Basel Committee on Banking Supervision, Basel III: the net stable funding ratio, 31 October 2014; now consolidated into the Basel Framework.
- Basel Framework, NSF20 — Calculation and reporting, defining NSFR and the 100% minimum.
- Basel Framework, NSF30 — Available and required stable funding, including current ASF and RSF treatment.
- Basel Framework, NSF99 — Application guidance, summarising ASF and RSF factors.
- Office of the Superintendent of Financial Institutions, Liquidity Adequacy Requirements (2026), Chapter 3 — Net Stable Funding Ratio, a current implementation cross-check.
Educational boundary: This article explains standardized structural-funding mathematics. It does not rate a bank’s safety or provide personalized financial advice.