Quick answer: the Net Stable Funding Ratio (NSFR) asks whether a bank funds its assets, derivatives and commitments with enough sources that are expected to remain reliable over roughly a one-year structural horizon. The basic Basel formula is NSFR = Available Stable Funding (ASF) / Required Stable Funding (RSF), with the ratio required to be at least 100% under the Basel standard. Liabilities and capital receive ASF factors according to their expected stability; assets and off-balance-sheet exposures receive RSF factors according to how much stable funding they are assumed to require. The mathematics is therefore a weighted balance-sheet problem: classify each item correctly, apply the right factor, aggregate the numerator and denominator, then test how funding choices change the ratio and the bank’s wider economics.

A bank can have enough cash for the next thirty days and still be structurally fragile if long-lived assets depend too heavily on funding that can disappear much sooner.

Page role: structural funding, not another LCR article

This page owns the one-year structural funding calculation. It does not replace HQLA buffer optimisation or liquidity stress testing. Those ask whether cash resources survive acute outflows. NSFR asks whether the bank’s funding architecture is stable enough for the assets it has chosen to hold.

1. The NSFR equation

The Basel Framework defines:

NSFR = ASF / RSF ≥ 100%.

ASF is the weighted amount of capital and liabilities expected to remain sufficiently stable. RSF is the weighted amount of stable funding considered necessary for assets and off-balance-sheet exposures. See the BIS NSFR Executive Summary and the current Basel Framework.

The key mathematical point is that neither numerator nor denominator equals an ordinary accounting total. Both are weighted sums.

2. Available Stable Funding: liabilities are not equally stable

Generic Basel ASF factors range from 100% to 0%. Long-term capital and liabilities receive the strongest recognition. Qualifying stable retail deposits receive high recognition, while short-term funding from financial counterparties can receive far less or none depending on maturity and classification.

For liability category i:

ASF = Σ Liability_i × ASF factor_i.

The familiar Basel factor set includes 100%, 95%, 90%, 50% and 0% categories. The factor is a supervisory stability assumption, not a prediction that exactly that percentage will remain in every real crisis.

3. Required Stable Funding: assets consume different amounts of structural funding

RSF factors rise as assets become less liquid, longer-dated, encumbered or otherwise harder to turn into funding capacity. Basel categories include factors such as 0%, 5%, 10%, 15%, 50%, 65%, 85% and 100% depending on the item and conditions.

For asset or exposure category j:

RSF = Σ Exposure_j × RSF factor_j.

Central-bank reserves and highly liquid unencumbered assets can require little stable funding, while long-dated loans, non-performing or illiquid assets, certain margin positions and other items can require much more. Basel’s detailed NSFR definitions and factor summary shows why classification matters.

4. Worked teaching example

Consider a deliberately simplified bank.

Funding source	Amount	Teaching ASF factor	ASF contribution
Capital + long-term debt	S$200m	100%	S$200m
Stable retail deposits	S$500m	95%	S$475m
Less-stable retail deposits	S$200m	90%	S$180m
Shorter wholesale funding	S$300m	50%	S$150m

Total teaching ASF = S$1,005m.

Asset/exposure	Amount	Teaching RSF factor	RSF contribution
Cash/reserves	S$100m	0%	S$0m
Highly liquid securities	S$200m	5%	S$10m
Shorter loans	S$300m	50%	S$150m
Residential mortgages	S$400m	65%	S$260m
Longer/less-liquid assets	S$200m	85%	S$170m

Total teaching RSF = S$590m. NSFR ≈ 1,005 / 590 = 170%.

The numbers are illustrative rather than a regulatory template. Real treatment depends on definitions, residual maturity, counterparty, collateral, encumbrance, jurisdiction and derivative/off-balance-sheet rules.

5. Why maturity transformation is the core mechanism

Banking transforms maturity: deposits and wholesale liabilities can be repayable sooner than mortgages, corporate loans and securities mature. That transformation is economically useful, but excessive reliance on short-term refinancing creates a rollover problem.

Imagine a five-year asset funded by one-month wholesale borrowing. The asset may be perfectly performing, yet the bank must repeatedly refinance the liability about sixty times before the asset matures. NSFR makes that structural dependency visible by rewarding stable funding and charging less-liquid assets more heavily.

6. LCR and NSFR can disagree without either being wrong

The Liquidity Coverage Ratio focuses on a severe 30-day stress and the stock of HQLA available against stressed net cash outflows. NSFR focuses on a roughly one-year structural funding horizon.

A bank can therefore have:

a strong LCR because it holds a large liquid-asset buffer, but a weaker NSFR because long-lived assets depend on unstable funding;
a strong NSFR because assets are funded conservatively, but a short-term liquidity weakness because immediate cash outflows are concentrated.

The US agencies explicitly describe NSFR as a one-year complement to the shorter-horizon LCR. See the Federal Reserve’s October 20, 2020 final-rule announcement.

7. Off-balance-sheet commitments still consume stable funding

A loan commitment can be undrawn today and still become a funded asset tomorrow. Basel therefore assigns RSF treatment to off-balance-sheet exposures rather than pretending that zero current balance means zero structural funding need.

This connects to revolving-credit utilisation and EAD: the same undrawn line can create both future credit exposure and future funding demand.

8. Derivatives make the classification problem less intuitive

Derivative receivables, liabilities and variation margin receive specialised NSFR treatment because the economic exposure is not represented well by gross notional. Netting, collateral and margin flows affect the structural funding calculation.

A useful model audit therefore asks whether derivative balances are being mapped through the dedicated NSFR rules rather than shoved into an ordinary asset bucket. US Regulation WW contains separate calculations for NSFR derivative amounts in Subpart K.

9. NSFR optimisation is a constrained balance-sheet problem

Suppose the bank can choose quantities x of deposits, term debt, short wholesale funding, mortgages, corporate loans and securities. A stylised optimisation can be written:

Maximise expected economic profit(x)

subject to:

ASF(x) − RSF(x) ≥ 0;
LCR and internal liquidity limits;
capital/RWA constraints;
interest-rate-risk limits;
funding concentration limits;
customer/product demand constraints;
legal-entity and currency constraints.

The cheapest action to improve NSFR is not always “issue long-term debt.” The bank might instead reduce an RSF-intensive asset, shift asset maturity, attract more stable deposits, reduce encumbrance or change a commitment structure. This connects to balance-sheet optimisation and shadow prices.

10. Evidence polarity: what should raise or lower confidence?

Evidence supporting a stable funding profile includes durable term funding, diversified retail deposits, low reliance on overnight refinancing, consistent regulatory classification and a ratio that remains resilient after plausible funding migration.

Evidence against the model includes supposedly stable deposits behaving like rate-sensitive wholesale money, encumbrance not reflected in asset treatment, repeated maturity misclassification, material derivatives mapped incorrectly, or an NSFR that collapses when realistic behavioural assumptions replace contractual labels.

11. Failure modes and counterexamples

Ratio worship. A 120% NSFR is treated as proof that all liquidity horizons are safe.
Contractual-maturity blindness. Behavioural funding leaves earlier than legal maturity suggests.
Deposit-label complacency. “Retail” is assumed stable without checking concentration and rate sensitivity.
Encumbrance omission. Assets unavailable to the bank receive treatment as if freely usable.
Off-balance-sheet invisibility. Commitments create future funding demand that never reaches RSF.
Jurisdiction copying. Basel factors or US treatment are copied mechanically into a different legal framework.
Constraint substitution. Improving NSFR by buying or selling assets inadvertently worsens capital, NII, EVE or LCR.

A useful counterexample is a bank with NSFR comfortably above 100% but severe same-day payment obligations and weak intraday liquidity. Structural stability does not produce money at every minute of the day. See intraday-liquidity forecasting.

12. Diagnostics and falsifier tests

Which three liability categories contribute most ASF?
Which three asset categories consume most RSF?
How far does NSFR fall if a portion of “stable” deposits migrates to a less-stable category?
What happens if short wholesale funding cannot roll at maturity?
How much NSFR improvement comes from numerator changes versus denominator changes?
Does a derivative reclassification materially move RSF?
Does the ratio remain above the target after realistic encumbrance and behavioural adjustments?

Falsifier: “Our funding is structurally stable because NSFR is high” is falsified if the high ratio depends materially on a category that empirical behaviour, legal restrictions or corrected classification shows is less stable than assumed.

13. Alternatives and complementary tools

NSFR is a regulatory structural metric, not a complete internal liquidity model. Complement it with contractual maturity ladders, behavioural deposit models, funds-transfer pricing, survival-horizon stress tests, HQLA analysis, concentration measures and currency/entity-specific liquidity views.

Funds Transfer Pricing is particularly useful because it can turn the economic cost of structural funding into a price signal for business lines rather than leaving NSFR as a compliance-only ratio.

14. Verification and update triggers

reconcile every NSFR balance to finance/regulatory source systems;
independently validate ASF/RSF mapping rules;
retest after new funding products or securitisations;
update after regulatory factor changes;
review deposit stability after sharp rate moves;
recalculate after large acquisitions, asset sales or encumbrance changes;
compare internal structural-liquidity stress with regulatory NSFR conclusions;
retain a simple maturity-gap challenger capable of disagreeing with the production calculation.

Research anchors

The deeper lesson

NSFR is the mathematics of matching funding persistence to asset persistence. It deliberately refuses to treat every dollar of liability as equally reliable and every dollar of asset as equally easy to fund. The ratio is useful because it compresses that structure into one constraint; it is limited because the compression can hide currency, entity, timing and behavioural detail. The strongest use of NSFR is therefore not “we passed 100%.” It is “we know which pieces create the ratio, why the factors make sense, what breaks the result, and which structural funding decision improves the bank without damaging another constraint.”

Educational note: This article explains public banking mathematics and regulatory concepts. It is not treasury advice, investment advice or institution-specific regulatory guidance.

Small Group Tutorials

How Banks Calculate and Optimise the Net Stable Funding Ratio: ASF, RSF, Maturity Transformation, Structural Funding and Constraint Trade-offs