Exposure at Default (EAD): Calculation, Estimation, and Mitigation

Exposure at default (EAD) is a bank’s estimate of how much money it will have at risk on a loan, credit line, or derivative contract at the moment the borrower or counterparty defaults. For a simple term loan, that figure is close to the outstanding balance. For a revolving credit line or a derivative, it is a projection, because the borrower can still draw more credit or the market value of the contract can still move before default actually happens. EAD is one of three inputs — alongside probability of default (PD) and loss given default (LGD) — that drive both regulatory capital requirements under the Basel framework and loan-loss provisions under IFRS 9 and CECL.

What EAD Represents

PD tells a bank how likely a borrower is to default. LGD tells it what share of the exposure will be permanently lost after recoveries. EAD supplies the dollar amount those percentages act on. Multiply the three together and you get expected loss; the same EAD figure then flows into the risk-weighted asset calculation that sets how much capital the bank must hold against the position.

Because EAD scales the whole calculation, errors compound. Overstate it and the bank locks up capital it could lend elsewhere. Understate it and the bank may find, in a downturn, that its reserves don’t cover what borrowers actually owe. The 2008 financial crisis illustrated the second failure mode: distressed companies drew heavily on committed credit lines on their way to default, and banks that had sized their exposure to the drawn balance alone were caught short.

How EAD Is Calculated

Term Loans and Fixed Instruments

For an amortizing term loan or a fixed-rate bond, EAD is roughly the outstanding principal at the point of default. The borrower has no contractual right to draw more, so the exposure is whatever is still owed. Nothing more complicated is needed.

Revolving Credit Facilities

Credit cards, corporate lines of credit, and overdraft facilities require a different approach because the borrower can draw on unused credit at any time. EAD on these products combines what has already been drawn with a projection of future drawdowns, built using a credit conversion factor (CCF):

EAD = Current Drawn Amount + (CCF × Undrawn Commitment)

Take a corporate line of credit with a $10 million limit. The company has drawn $4 million, leaving $6 million unused. Applying a 40% CCF — the standard Basel rate for most commitments — produces an EAD of $4 million + (0.40 × $6 million) = $6.4 million. That $6.4 million, not the $4 million currently outstanding, is the figure the bank uses for capital calculations.

The Basel CCF Schedule

Banks on the Standardized Approach apply CCFs set directly by the Basel framework rather than estimating them internally. The current rates in the Basel framework’s CRE20 chapter are:1Bank for International Settlements. Basel Framework CRE20 – Standardised Approach: Individual Exposures

  • 100% for direct credit substitutes, including financial guarantees, repurchase agreements where credit risk stays with the bank, and forward asset purchases with certain drawdown.
  • 50% for note issuance facilities, revolving underwriting facilities, and transaction-related contingent items such as performance bonds and bid bonds.
  • 40% for general commitments, including revolving credit lines, regardless of the facility’s maturity, unless a lower rate applies.
  • 20% for short-term self-liquidating trade letters of credit tied to the movement of goods.
  • 10% for commitments the bank can unconditionally cancel at any time without notice, or that automatically cancel if the borrower’s creditworthiness deteriorates. National supervisors may set this higher if local legal or practical constraints limit the bank’s ability to actually cancel.

Derivatives and Counterparty Credit Risk

Derivatives create a two-sided exposure problem. A swap or forward may be worth millions to the bank one day and nothing the next, and if the counterparty defaults while the contract has positive value, the bank loses the cost of replacing it. The Basel framework handles this through the Standardized Approach to Counterparty Credit Risk (SA-CCR), which sets EAD as the sum of replacement cost and a potential future exposure (PFE) add-on.2Bank for International Settlements. Basel Framework CRE52 – Standardised Approach to Counterparty Credit Risk

Replacement cost captures the current mark-to-market value. The PFE add-on reflects how much further the contract’s value could move in the bank’s favor before it can be replaced. SA-CCR builds the add-on trade by trade: an effective notional is derived from the adjusted notional, a supervisory delta (reflecting direction and option characteristics), and a maturity factor. Those effective notionals are then aggregated within asset classes and multiplied by prescribed supervisory factors. For interest rate derivatives, the supervisory factor is 0.5%. The resulting PFE is added to replacement cost to produce the EAD for the netting set.

Who Estimates EAD: Standardized vs. IRB

Which methodology a bank uses to arrive at EAD depends on the regulatory approach it has adopted.

Under the Standardized Approach, the bank applies the CCFs and risk weights set by regulators, with no internal estimation. The approach is simpler to operate, works for institutions without deep historical data, and gives supervisors uniform outputs that can be compared across banks.

Under the Internal Ratings-Based (IRB) Approach, the bank estimates risk components itself, subject to supervisory approval.3Bank for International Settlements. Basel Framework CRE32 – IRB Approach: Risk Components IRB comes in two tiers:

  • Foundation IRB (F-IRB): the bank estimates its own PD, but uses supervisory values for LGD and CCFs. The CCFs mirror the standardized schedule.
  • Advanced IRB (A-IRB): the bank uses its own estimates for all risk components, including EAD.

An A-IRB EAD model generally analyzes historical borrower behavior in the run-up to default: how fast distressed borrowers draw on available credit, how utilization shifts across economic cycles, and how facility-specific features such as covenants affect drawdown patterns. For a well-managed portfolio, the statistically derived EAD can sit below what the blunt standardized CCFs would produce, which lowers risk-weighted assets and the capital the bank must hold against them. The trade-off is cost: years of clean default and drawdown data, dedicated quantitative teams, independent model validation, documentation, and ongoing backtesting under supervisory examination. Only large banks typically find the investment worthwhile.

How Banks Reduce EAD

EAD is not purely a measurement exercise. Banks actively manage it down using recognized credit risk mitigation techniques, each with its own regulatory requirements.

Collateral

Under the Basel framework’s comprehensive approach, the bank reduces its exposure by the volatility-adjusted value of collateral it holds. Regulatory haircuts account for potential changes in the value of both the exposure and the collateral, and for any currency mismatch between the two.4Bank for International Settlements. Basel Framework CRE22 – Standardised Approach: Credit Risk Mitigation The reduction only counts if the collateral agreement is legally enforceable in every relevant jurisdiction.

Netting Agreements

For derivative portfolios, a master netting agreement allows the bank to offset positive and negative contract values against the same counterparty. Without netting, only positive values count as exposure. With a legally enforceable netting agreement, a bank holding $5 million in positive positions and $3 million in negative positions against the same counterparty records exposure of $2 million rather than $5 million. SA-CCR calculates replacement cost and PFE at the netting-set level, so netting is built into the framework rather than bolted on.

Guarantees and Credit Derivatives

Third-party guarantees and credit default swaps let a bank substitute the credit risk of the borrower with the risk of the guarantor or protection seller. Under the standardized approach, the guaranteed portion is assigned the risk weight of the protection provider instead of the borrower, but only when the guarantor carries a lower risk weight than the original counterparty. Capital is still held against the unguaranteed portion at the borrower’s original weight.

All three techniques depend on legal certainty, sound operational processes, and proper documentation. A collateral agreement that wouldn’t hold up in a foreign court, or a netting opinion that has not been refreshed, costs the bank the right to recognize the EAD reduction for capital purposes. Risk mitigation most often fails here, in the legal and operational plumbing, rather than in the arithmetic.

EAD Beyond Regulatory Capital

EAD is not only a Basel concept. It also drives loan-loss provisioning under IFRS 9 internationally and under CECL in the United States. Both frameworks require banks to estimate expected credit losses using the same basic structure: Expected Credit Loss = EAD × PD × LGD, discounted at the instrument’s effective interest rate.

Under IFRS 9, the horizon over which EAD is projected depends on the loan’s credit stage. For performing loans (Stage 1), EAD is estimated over a 12-month window. For loans showing significant credit deterioration (Stage 2) or already in default (Stage 3), the projection extends over the full remaining life of the instrument. Accurate EAD modeling therefore affects the size of a bank’s provisions on its income statement, not only its regulatory capital ratios. Banks typically run EAD models that serve both masters at once, and the inputs do not always align perfectly, which creates operational complexity between what the capital rules require and what the accounting standards demand.