Skip to content

Counterparty credit risk

Follow the positive derivative claim that can be lost at default, and distinguish it from notional, future exposure, regulatory EAD and the effects of netting and collateral.

Counterparty credit risk is the risk of losing a claim because the other party to a transaction defaults before its cash flows have finished settling. For a derivative, the claim changes with market prices: a trade that starts at fair value can later become a substantial receivable. The desk needs to know both whether its trading partner can pay and how much the partner could owe when it fails.
The familiar derivative case is bilateral. A signed contract value can favour either party, so either may become the exposed creditor. This differs from the simple lender–borrower picture of credit risk. OSFI's Canadian bank-capital definitions describe that distinction; the wider CCR framework also covers securities-financing and other specified transactions.

Start with the claim, not the contract size

Consider one fictional unsecured derivative with notional amount USD 100 million. The counterparty defaults when the stipulated close-out amount is positive USD 10 million to us. Assume we survive, there are no other trades to offset it, no collateral or guarantees, and no close-out costs. Recovery is stipulated to be 40% of this claim, measured as a cash-equivalent value on the same valuation date. There is no additional discounting or settlement delay in this example.
Claim ledger at defaultUSD millions
Positive close-out claim10
Recovery: 40% of the claim4
Unrecovered loss6
The loss given default fraction is 60%, so the loss is 10×(1−0.40)=610\times(1-0.40)=6 million. Applying that fraction to the USD 100 million notional would answer the wrong question. Notional scales contractual payments; it is not automatically the replacement claim.
A fictional derivative has USD 100 million notional and a USD 10 million positive close-out claim. At stipulated 40% recovery, USD 4 million is recovered and USD 6 million is lost. Signed contract values of minus 10, zero and plus 10 million have positive current exposures of zero, zero and 10 million.
A fictional derivative has USD 100 million notional and a USD 10 million positive close-out claim. At stipulated 40% recovery, USD 4 million is recovered and USD 6 million is lost. Signed contract values of minus 10, zero and plus 10 million have positive current exposures of zero, zero and 10 million.
Original synthetic claim ledger. All amounts are USD millions at one valuation date. Single unsecured derivative, no netting or collateral, stipulated close-out and recovery; not market data.
If the contract were worth negative USD 10 million to us, we would have a payable, with zero positive current exposure from this contract. Counterparty default does not turn that payable into a positive claim. Under this example's clean close-out convention, we still pay our obligation in full. Actual termination, set-off and insolvency outcomes require the applicable contract and law; the assumed settlement rule is not universal.
The USD 6 million is a loss conditional on this default scenario. It is neither the trade's expected credit loss today nor its credit valuation adjustment, which requires a pricing model for discounted losses across possible default times and exposures.

Four quantities that answer different questions

For this single unsecured contract, let VtV_t be its signed replacement value to us, before allowing for either contracting party's default. Its positive current exposure is

Et=max⁡(Vt,0).E_t=\max(V_t,0).
QuantityWhat it tells the desk
Current exposureThe positive replacement claim now, before recovery. It can be zero while future risk remains.
Potential future exposureIn a scenario report, a chosen high percentile of exposure at a future date; state the probability model. A regulatory PFE add-on has its own prescribed construction.
Exposure at defaultThe amount exposed when the counterparty defaults, or an estimate of that amount. A regulatory EAD estimate follows the applicable method and is not simply today's positive value.
NotionalA contractual reference amount, not a measure of loss by itself.
For example, OSFI's 2026 Canadian SA-CCR framework constructs regulatory EAD from replacement cost and a PFE component with a prescribed multiplier. That component should not be silently substituted for a simulated exposure percentile. See Chapter 7, §7.1.7. The Exposure profiles lesson develops the scenario-based calculation and the distinction between mean exposure and a quantile; its public page is a preview and the full lesson requires membership.

The agreement and the future path matter

Close-out netting can combine covered receivables and payables into a single claim. Sharing a counterparty name does not establish an enforceable netting set. Collateral can cover part of the claim, but the usable balance, its value and the time needed to close out matter. A margin call that has not settled is not cash already held.

Exposure and credit quality can also deteriorate together. Wrong-way risk occurs when the exposure is adversely related to the counterparty's creditworthiness; an unconditional average exposure can then conceal the amount at risk in default states.

These are reasons to monitor more than a single current-value number. The Basel Committee's CCR guidance calls for complementary exposure metrics alongside counterparty due diligence, mitigation and limits. A useful report preserves the legal agreement, valuation currency, exposure horizon and collateral chronology behind each number.

Sources

  • OSFI, Capital Adequacy Requirements 2026, Chapter 7, §§7.1.1 and 7.1.7: definitions, netting sets, current exposure and the Canadian SA-CCR construction. This edition states validity through the reporting period ending 31 January 2027.
  • Basel Committee, Counterparty credit risk, CRI40, based on the December 2024 management guidelines: interacting exposure and credit quality, complementary metrics, mitigation and governance.
  • McNeil, Frey and Embrechts, Quantitative Risk Management: Concepts, Techniques and Tools, revised edition, Princeton University Press, 2015, §17.2, pp. 603–606, especially §17.2.1 and equations 17.3–17.8. The ledger is an original stipulated example.
Last updated
  • counterparty credit risk
  • derivatives
  • credit risk
  • exposure
  • close out
← Back to the glossaryBack to top ↑