Reader question: A bank can have thousands of individually small loans and trades. How does a regulator decide whether too much of the bank’s capital is actually exposed to one economic failure?
The Basel large-exposures framework answers by aggregating exposures to a single counterparty or a group of connected counterparties and comparing the result with the bank’s Tier 1 capital. A concentration at or above 10% of Tier 1 capital is a large exposure for reporting purposes. The standard limit is 25% of Tier 1 capital. For a global systemically important bank’s exposure to another G-SIB, the Basel limit is tighter at 15%.
The difficult part is not the final division. It is deciding what belongs in the numerator. A bank must recognise legal control, economic interdependence, off-balance-sheet commitments, derivatives, securities financing, trading-book positions, guarantees, collateral substitution and indirect exposures through funds or securitisations.
What this page owns — and what it does not
This page owns:
counterparty relationships + measured exposures + credit-risk mitigation → aggregated exposure / Tier 1 capital → large-exposure status and limit test.
It does not replace SA-CCR derivative exposure measurement, standardised credit-risk RWA, or balance-sheet optimisation. The large-exposure framework is a concentration constraint, not another risk-weight formula.
This is public prudential mathematics. Jurisdictions may implement stricter limits or additional rules.
Why ordinary risk weights are not enough
A portfolio can have acceptable average credit risk and still be dangerously concentrated.
Suppose a bank has 100 loans of 1 million each. If all 100 borrowers are independent, one default may be manageable. If the 100 legal borrowers are actually subsidiaries of one corporate group, the bank has a 100 million concentration to one economic failure.
Basel therefore measures large exposures separately from ordinary risk-weighted capital. The objective is to limit the maximum loss from a sudden counterparty failure to a level that should not endanger the bank’s solvency.
Step 1: define the capital denominator
The Basel large-exposure ratio uses Tier 1 capital as its capital base.
For a measured exposure E:
Large-Exposure Ratio = E / Tier 1 Capital.
If Tier 1 capital is 20 billion and the aggregated exposure to one connected group is 4 billion:
4 / 20 = 20%.
This is above the 10% large-exposure reporting threshold but below the standard 25% Basel limit.
Step 2: define a single counterparty correctly
The obvious case is one legal entity: a company, bank, fund or other counterparty.
But Basel’s concentration logic asks a deeper question: Could several legal entities fail together because they are controlled together or economically dependent on one another?
If yes, those entities may need to be treated as one group of connected counterparties.
Connected counterparties: control relationship
One connection test is control. If one counterparty directly or indirectly controls another, or several entities are under common control, their exposures can need to be aggregated.
The intuition is simple: legal separation does not remove concentration if one controlling party can determine the group’s financial behaviour.
A data engine therefore needs ownership and control relationships, not merely customer IDs.
Connected counterparties: economic interdependence
The second major test is economic interdependence. Basel defines connected counterparties to include cases where financial problems at one entity would likely cause financial difficulties at another.
Possible signals include dependence on the same source of revenue, funding, guarantees, customers, suppliers or repayment cash flows.
This is harder than ownership mapping because the connection can exist without common legal control.
A simple economic-dependence example
Company A manufactures one specialised component. Company B buys almost all of A’s output and has no practical substitute supplier. A collapse of B could destroy A’s cash flow; a collapse of A could halt B’s production.
Even if A and B have unrelated shareholders, their credit risks may be economically linked.
A concentration engine that aggregates only by corporate parent can miss this dependency.
Step 3: define the exposure numerator
The large-exposure numerator covers relevant positions from both the banking book and trading book, including on- and off-balance-sheet items.
The measurement depends on exposure type:
- banking-book on-balance-sheet assets generally start from accounting exposure values;
- off-balance-sheet commitments use credit-conversion factors;
- OTC derivative counterparty exposure uses the applicable counterparty-credit-risk methodology, including SA-CCR under the Basel framework;
- securities-financing transactions use prescribed exposure calculations and haircuts;
- trading-book positions use large-exposure rules linked to jump-to-default measures.
The framework tries to reuse capital-framework exposure measures where practical, but it does not apply ordinary credit risk weights to shrink the concentration numerator.
Why risk weighting would understate concentration
Suppose a bank has a 100 exposure to a highly rated counterparty that attracts a low risk weight for capital purposes.
If the bank used risk-weighted exposure as the concentration numerator, the amount could look small. Yet if that counterparty suddenly failed, the bank’s gross economic loss could still be large.
Large-exposure limits therefore focus on exposure amount rather than portfolio-average risk weighting.
Step 4: aggregate every exposure to the same connected group
Let group G contain counterparties A, B and C.
Then:
EG = EA + EB + EC
after applying the framework’s permitted netting and credit-risk-mitigation treatment.
If A has a loan, B a derivative and C a bond in the trading book, the exposure engine must bring those different instruments into one concentration view.
10% defines a reportable large exposure
Under Basel, an exposure equal to or above 10% of Tier 1 capital meets the definition of a large exposure for reporting purposes.
That does not mean exposures below 10% are harmless. It means 10% is the formal large-exposure reporting threshold in the standard.
Banks also report other specified concentration information, including certain pre-mitigation and exempt exposures and their largest exposures.
25% is the standard Basel limit
The sum of exposures to one counterparty or connected group must not exceed:
25% × Tier 1 Capital.
If Tier 1 capital is 20 billion:
standard Basel limit = 5 billion.
An exposure of 5.2 billion is a breach even if the counterparty has an investment-grade rating.
G-SIB to G-SIB uses a tighter Basel limit
For a global systemically important bank’s exposure to another G-SIB, Basel sets the limit at:
15% × Tier 1 Capital.
The rationale is contagion. Failure of one systemically important bank can transmit stress through another major bank more severely than a comparable exposure to a less systemically important institution.
Step 5: recognise credit-risk mitigation carefully
Eligible guarantees, credit derivatives, financial collateral and on-balance-sheet netting can reduce the original exposure under specified conditions.
But the risk does not simply disappear.
If Counterparty A’s exposure is guaranteed by Guarantor B, the framework can reduce exposure to A while creating or increasing exposure to B.
This is substitution:
original counterparty exposure ↓; protection-provider exposure ↑.
A guarantee can solve one concentration and create another
Suppose a bank has a 4 billion exposure to Borrower A, guaranteed fully by Bank B.
Recognising the guarantee can reduce the concentration to A. But if the bank already has 3 billion of other exposure to B, the guarantee can push the aggregated exposure to B toward or through its limit.
Credit-risk mitigation therefore changes the graph of concentration rather than mechanically reducing all risk.
Collateral can also substitute exposure
Where eligible financial collateral receives recognition, the bank may reduce exposure to the original counterparty and recognise exposure to the collateral issuer in accordance with the framework.
This matters for concentrated collateral. A portfolio secured by the same government or corporate securities can transfer concentration toward the collateral issuer.
Maturity mismatches weaken mitigation
If a guarantee or hedge expires before the underlying exposure, protection is incomplete.
Basel applies conservative recognition rules for maturity mismatch. A calculation engine must therefore compare:
- maturity of the original exposure;
- residual maturity of protection;
- applicable adjustment rules.
Simply subtracting the full guarantee amount can overstate protection.
Derivatives link directly to SA-CCR
For OTC derivatives, the large-exposure framework uses counterparty-credit-risk exposure measures rather than notional amount alone.
This links to the existing SA-CCR article.
The large-exposure algorithm then asks a different question:
After measuring derivative exposure, how much of Tier 1 capital is concentrated in this counterparty or connected group?
Securities-financing transactions add haircut logic
Repos, reverse repos and securities lending create exposures to both counterparties and collateral.
The large-exposure framework uses prescribed methods that recognise market-price volatility through supervisory haircut logic.
This connects with repo-pricing and collateral mechanics, but the regulatory concentration objective is distinct from financing price.
Trading-book positions need default-sensitive exposure measures
A bond, equity or derivative in the trading book can create exposure to an issuer even if the position is not recorded as a conventional loan.
Basel therefore uses default-sensitive exposure measures such as gross jump-to-default for relevant trading-book positions in the large-exposure framework.
This prevents a bank from hiding issuer concentration simply because the exposure sits in the trading book.
Funds and securitisations create look-through problems
Suppose a bank invests 1 billion in a fund. The direct legal counterparty is the fund vehicle, but the economic exposure may be spread across dozens of underlying issuers.
Basel requires look-through treatment in specified cases for collective investment undertakings, securitisations and similar structures.
The data problem becomes:
fund position → underlying assets → underlying counterparties → connected-group aggregation.
Incomplete transparency can itself require conservative treatment.
Sovereign and central-bank exposures can be exempt
The Basel framework provides exemptions for specified sovereign, central-bank and related exposures.
An exemption means the exposure is outside or specially treated under the large-exposure limit framework. It does not mean the exposure has zero market or credit risk in every other context.
Rule engines must distinguish regulatory exemption from economic risk absence.
Evidence polarity: what supports confidence?
Evidence for a reliable large-exposure calculation includes complete counterparty IDs, ownership/control graphs, documented economic-interdependence analysis, instrument-level exposure measures that reconcile to source systems, CRM substitutions that balance between protected and protection-provider exposures, and reported ratios that reproduce from Tier 1 capital and exposure data.
Evidence against confidence includes multiple customer IDs for the same legal entity, missing subsidiaries, stale corporate-control data, guarantees reducing one exposure without adding the protection-provider exposure, fund holdings without look-through data, or derivative exposures measured differently from the approved SA-CCR engine.
Counterexample: 20 legal borrowers can be one concentration
A bank can comply with a per-legal-entity limit and still have excessive concentration if all 20 borrowers are controlled by the same parent.
The connected-counterparty graph is therefore as important as the exposure arithmetic.
Counterexample: an exposure can be below 25% but still dangerous
Suppose a bank has five unrelated exposures of 20% of Tier 1 capital each to one fragile industry.
No single exposure breaches the 25% counterparty limit, but sector concentration can still be severe.
The large-exposure framework addresses single-counterparty concentration, not every form of concentration risk. Sector, geography and common-risk-factor concentration require additional management and supervisory analysis.
Counterexample: collateral can increase concentration elsewhere
One hundred loans secured by securities from the same corporate issuer can look diversified by borrower but concentrated by collateral issuer after substitution treatment.
A correct engine must examine where the risk goes after mitigation.
Counterexample: a rating upgrade does not increase the 25% limit
The large-exposure limit is tied to Tier 1 capital, not the counterparty’s credit rating.
A highly rated private-sector counterparty does not receive a larger Basel concentration limit merely because its probability of default is low.
Weak links in implementation
entity-resolution failure. One counterparty appears under multiple identifiers.
ownership-graph lag. A merger or acquisition is not reflected promptly.
economic-dependence omission. Connections are tested only by legal control.
pre/post-CRM confusion. Reporting and limit calculations use inconsistent stages.
guarantee substitution leak. Original exposure is reduced but protection-provider exposure is not created.
fund look-through gap. Indirect concentrations are invisible.
Tier 1 timing mismatch. Exposure and capital denominator come from different reporting dates.
jurisdiction hard-coding. Basel limits are assumed to be the only local limits when a jurisdiction is stricter.
Diagnostics: how to test the engine
- 10% threshold test: exposures at 9.99%, 10% and 10.01% of Tier 1 should classify correctly.
- 25% limit test: test just below, at and above the standard Basel limit.
- G-SIB test: apply 15% only when both applicable systemic-status conditions are satisfied under the relevant rules.
- control-group test: aggregate parent and controlled subsidiaries.
- economic-dependence test: aggregate unrelated legal entities when the defined interdependence conditions are met.
- guarantee test: reduce A and create exposure to B.
- derivative test: reconcile counterparty exposure with the approved SA-CCR output.
- fund test: compare look-through and no-look-through treatment under the applicable rule.
- capital-date test: ensure Tier 1 capital and exposure values use the required reporting basis.
- breach test: a limit breach must trigger the required escalation and reporting workflow rather than merely a red dashboard cell.
What would falsify confidence?
Confidence should be withdrawn if known connected counterparties are not aggregated; if total exposure changes merely because one entity receives a new customer ID; if guarantees reduce exposure without substitution; if a 26% ratio passes the standard 25% Basel test; if the G-SIB limit is applied incorrectly; or if the ratio cannot be reproduced from source exposure data and Tier 1 capital.
Alternatives and limits
Internal concentration limits can be tighter than regulatory limits and can cover industries, countries, collateral types, wrong-way risk and correlated counterparties. Stress testing can estimate losses from common shocks across exposures that are legally independent. Network models can identify indirect contagion.
The Basel large-exposure framework provides a hard single-counterparty concentration boundary. It should be treated as a minimum prudential constraint, not a complete concentration-risk model.
How this connects to the surrounding knowledge estate
The exposure numerator can consume SA-CCR outputs, off-balance-sheet CCF logic from the standardised credit-risk framework, and financing/collateral information from repo algorithms. The resulting concentration constraints become inputs into balance-sheet optimisation.
Verification and update triggers
Preserve the Basel/local rule version, Tier 1 capital source, legal-entity graph, economic-dependence decisions, exposure-measure versions, CRM substitutions, fund look-through data and systemic-status flags. Revalidate after mergers, major restructurings, G-SIB-list changes, collateral-policy changes, derivative-methodology changes or any breach/near-breach incident.
Primary and high-quality references
- Basel Committee on Banking Supervision, Large exposures.
- Basel Committee, LEX10 — Definitions and application.
- Basel Committee, LEX20 — Requirements, including the 10% reporting threshold and 25% Tier 1 limit.
- Basel Committee, LEX30 — Exposure measurement.
- Basel Committee, LEX40 — Large exposure rules for G-SIBs.
Educational boundary: This article explains public concentration-limit mathematics. It does not determine the legal or regulatory status of any real exposure and does not provide personalized financial advice.