Exposure at Default is one of three parameters that decide how much a bank stands to lose when a borrower stops paying. Alongside Probability of Default and Loss Given Default, it determines the size of the loss a lender must plan for. However, the calculation method changes depending on which regulatory rulebook applies. This guide explains what Exposure at Default means and how banks calculate it. It also covers four different frameworks that treat the concept differently: the Reserve Bank of India’s ECL Directions, the international accounting standard behind global provisioning rules, the Basel capital framework, and IFRS 9 specifically.
Exposure at Default measures the amount a bank expects to have outstanding at the moment a borrower defaults. It sits inside the standard expected credit loss formula:
Expected Credit Loss = Probability of Default × Loss Given Default × Exposure at Default × Discount Factor
People often mistake Exposure at Default for the current outstanding balance on a loan. In reality, the two aren’t always the same. For a term loan with a fixed repayment schedule, the outstanding balance and the expected exposure at default stay close enough to treat as one figure. A revolving facility works differently. Consider a credit card, an overdraft, or a working capital line. The amount drawn today can sit far below what a bank could face if the borrower draws down the rest of the available limit right before defaulting. Exposure at Default exists specifically to capture that gap.
Exposure at Default directly shapes how much capital and provisioning a bank must hold. If a bank underestimates it, provisions fall short exactly when a portfolio comes under stress. If a bank overestimates it, capital sits idle instead of supporting new lending. Either error eventually shows up on the P&L and on regulatory capital ratios.
The 2008 collapse of Lehman Brothers offers a widely cited real-world illustration of this risk. In the run-up to default, distressed borrowers connected to Lehman-related exposures drew down undrawn credit lines aggressively. As a result, actual exposure climbed well above what static, pre-crisis balance-sheet snapshots suggested. Banks that modeled Exposure at Default using only current drawn balances ended up understating their true risk. Credit risk teams still design for this pattern today, because exposure tends to rise, not fall, as a borrower approaches default. Our post on why loss given default matters explores this further: exposure size and recovery prospects together determine how severe a loss ultimately is.
For revolving or partially drawn facilities, banks calculate Exposure at Default as:
EAD = Drawn Amount + (Credit Conversion Factor × Undrawn Commitment)
The Credit Conversion Factor, or CCF, represents the share of the currently undrawn commitment that a bank expects the borrower to draw before default. In other words, it converts an off-balance-sheet commitment into an on-balance-sheet-equivalent exposure.
Worked example: a borrower holds a working capital facility with ₹50 lakh drawn and ₹30 lakh undrawn. If the bank’s CCF for this facility type is 40%, the expected additional drawdown comes to ₹12 lakh (40% of ₹30 lakh). Therefore, total Exposure at Default reaches ₹62 lakh — ₹12 lakh higher than the current outstanding balance alone would suggest. That ₹12 lakh gap represents precisely the risk a static balance-sheet view would miss.
Term loans have a fixed disbursement and a contractual amortization schedule. As a result, banks know their Exposure at Default largely in advance, since it simply tracks the remaining principal outstanding at any point, with no undrawn commitment left to model.
Revolving facilities work differently. Credit cards, cash credit accounts, overdrafts, and working capital limits all give the borrower discretion over how much of an approved limit to draw at any time. Borrowers tend to exercise that discretion most aggressively just before defaulting. Consequently, banks cannot rely on the current drawn balance alone. Instead, they must model Exposure at Default for these facilities using a Credit Conversion Factor, segmented by facility type and borrower risk profile.
Banks typically estimate CCF from historical data. They study defaulted accounts and measure how much of the undrawn limit borrowers actually used in the period leading up to default. From that observed behavior, they build a CCF estimate — often segmented by product type, risk grade, or vintage.
However, the source of that CCF, and how much discretion a bank has in setting it, differs sharply depending on which regulatory lens applies. A capital-adequacy regulator, an accounting standard-setter, and a provisioning regulator each care about CCF for different reasons. Each also imposes a different level of prescription on how banks calculate it. The next four sections walk through those perspectives one at a time.
The Reserve Bank of India issued its final Expected Credit Loss Directions on April 27, 2026. The framework takes effect April 1, 2027. Scheduled commercial banks must comply, though the Directions exclude Regional Rural Banks, Small Finance Banks, payments banks, and local area banks. Under the framework, banks must estimate Probability of Default, Loss Given Default, and Exposure at Default as the three core parameters of expected credit loss. This replaces India’s earlier incurred-loss provisioning approach.
For revolving facilities, banks calculate Exposure at Default as the drawn balance plus a CCF-adjusted portion of the undrawn commitment, consistent with the general formula above. Additionally, RBI has built in product-wise prudential floors as a backstop. As a result, a bank’s internally modeled EAD cannot fall below a regulator-set minimum, regardless of how favorable its own historical data looks. Banks also get a four-year transition window, running through FY2031, to phase in the capital and provisioning impact.
It’s worth being precise here: this is a provisioning framework, not a capital framework. India applies the Standardised Approach for Basel capital adequacy rather than the Internal Ratings-Based approaches. Therefore, a bank builds its own EAD models for RBI’s ECL Directions, while it uses regulator-prescribed factors for capital purposes under Basel. In short, the two frameworks answer different questions and shouldn’t be conflated.
RBI’s Directions didn’t emerge in isolation. Instead, they build on a global accounting standard set by the International Accounting Standards Board (IASB). Before 2018, banks worldwide followed IAS 39, which used an incurred-loss model: a loss event had to occur before a bank could recognize impairment. Under that regime, Exposure at Default was effectively a backward-looking snapshot, and banks only revised it once trouble was already visible.
The IASB issued the complete version of IFRS 9 in July 2014. It became effective for annual reporting periods beginning on or after January 1, 2018, and fully replaced IAS 39’s impairment model. IFRS 9 introduced a single, forward-looking expected credit loss model that applies to all in-scope financial instruments. This shift turned Exposure at Default into a modeled, forward-looking figure rather than a static balance check. Consequently, this accounting foundation underpins RBI’s own Directions, along with similar national frameworks elsewhere.
Basel III’s capital adequacy framework uses Exposure at Default for an entirely different purpose: calculating Risk-Weighted Assets, the base on which regulators set minimum capital requirements. Here, banks convert off-balance-sheet items — undrawn commitments, guarantees, letters of credit — into credit-equivalent exposure using a Credit Conversion Factor. This is the same underlying concept covered above, but applied to a capital rather than a provisioning question.
Under the Foundation Internal Ratings-Based approach, regulators set the CCF values directly. Under the Advanced IRB approach, however, banks estimate their own CCFs from internal data, subject to supervisory validation. Indian banks sit outside both approaches: because RBI mandates the Standardised Approach for capital adequacy, Basel’s IRB-based CCF modeling doesn’t apply to their capital calculations. Even so, those same banks still build bank-estimated EAD models for RBI’s separate ECL provisioning framework.
IFRS 9 organizes credit exposures into three stages, and the estimation horizon for Exposure at Default changes by stage. Stage 1 (performing) uses a 12-month expected credit loss estimate. Meanwhile, Stage 2 (significant increase in credit risk) and Stage 3 (credit-impaired) both require a lifetime expected credit loss estimate. In other words, the EAD estimation window itself shifts as a borrower’s credit quality deteriorates.
IFRS 9 also requires banks to incorporate forward-looking, probability-weighted macroeconomic scenarios into EAD and CCF estimates, rather than relying purely on historical averages. Critically, IFRS 9 remains principle-based. Unlike RBI’s Directions, it does not prescribe fixed CCF values or product-wise floors. Instead, it leaves more estimation judgment to the reporting entity, subject to auditor and regulator scrutiny.
Table: Exposure at Default Across Four Regulatory Lenses
| Dimension | RBI (ECL Directions) | IASB / IFRS 9 (general) | Basel III (capital) | IFRS 9 (ECL mechanics) |
| Purpose | Provisioning (P&L/balance sheet) | Global accounting standard baseline | Capital adequacy (RWA) | Provisioning (P&L/balance sheet) |
| Governing body | Reserve Bank of India | IASB | Basel Committee (BIS) | IASB |
| Prescriptiveness | Prescriptive — product-wise prudential floors | Principle-based framework | Prescriptive under Standardised/Foundation IRB; modeled under Advanced IRB | Principle-based, no fixed floors |
| CCF treatment | Bank-modeled, subject to RBI floors | Not CCF-specific — sets the ECL principle IFRS 9 operationalizes | Supervisory CCF (Foundation IRB) or bank-estimated (Advanced IRB); India uses Standardised Approach | Bank-modeled, informed by forward-looking scenarios, no supervisory floor |
| Time horizon | Staged: 12-month / lifetime | N/A (framework-level) | Point-in-time exposure at default | Staged: 12-month (Stage 1) / lifetime (Stage 2–3) |
| India applicability | Mandatory for scheduled commercial banks from April 1, 2027 | Basis for RBI’s framework, not directly applied | Standardised Approach only — no IRB | Underlying global standard RBI’s Directions align to |
import pandas as pd
# Illustrative dataset: defaulted revolving accounts
df = pd.DataFrame({
“drawn_at_default”: [620000, 480000, 910000],
“undrawn_at_observation”: [300000, 250000, 400000],
“limit”: [900000, 700000, 1300000],
})
# CCF = proportion of undrawn amount that was drawn before default
df[“ccf_observed”] = (
(df[“drawn_at_default”] – (df[“limit”] – df[“undrawn_at_observation”]))
/ df[“undrawn_at_observation”]
).clip(lower=0, upper=1)
avg_ccf = df[“ccf_observed”].mean()
# Apply the estimated CCF to a live account
drawn, undrawn = 500000, 300000
ead = drawn + avg_ccf * undrawn
print(f”Estimated CCF: {avg_ccf:.2%}, EAD: ₹{ead:,.0f}”)
This example is illustrative, not production-grade. A real CCF model would segment by facility type and risk grade, and it would typically use a Beta regression — the same modeling approach we use elsewhere in this content series for LGD recovery rates.
What is Exposure at Default (EAD) in simple terms? It’s the amount outstanding when a borrower defaults, including any undrawn credit the borrower is likely to draw down beforehand.
What is the formula for calculating EAD? EAD = Drawn Amount + (Credit Conversion Factor × Undrawn Commitment).
Is EAD the same as the outstanding loan balance? For term loans, they’re close. For revolving facilities, EAD is typically higher because it accounts for further drawdown before default.
How does EAD differ from PD and LGD? PD estimates the likelihood of default. LGD estimates the share of exposure that won’t be recovered. EAD estimates the size of the exposure itself — together, the three combine into expected credit loss.
Does RBI treat EAD the same way as Basel and IFRS 9? No. RBI’s ECL Directions govern provisioning with prescribed floors, Basel governs capital adequacy, and IFRS 9 is the principle-based global accounting standard underlying RBI’s framework. See the comparison table above for the full breakdown.
Exposure at Default measures how much is genuinely at risk when a borrower defaults. Getting it right means going beyond the current outstanding balance and modeling undrawn commitments through a Credit Conversion Factor. However, the governing rules change depending on the lens. RBI’s ECL Directions (effective April 1, 2027) prescribe provisioning floors. The IASB’s IFRS 9 sets the forward-looking global accounting standard beneath them. Basel III, meanwhile, applies EAD to a separate, capital-adequacy calculation entirely. Understanding all three — and how they interact — remains core to credit risk work in Indian banking today.
Understanding the theory behind PD, LGD, and EAD is the first step. Building bankable, interview-ready models — in Python or SAS, on real credit datasets, aligned to Basel and IFRS 9 — is what actually moves a career forward.
Explore Dexlab Analytics’ Credit Risk Modeling certification program to build PD, LGD, and EAD models from scratch, work through IFRS 9 ECL frameworks, and learn model validation techniques used by practicing risk teams.
.
Credit Risk, credit risk analysis, Credit Risk Analytics And Modeling, credit risk analytics training, Credit risk certification India, Credit risk modeling course India, Credit Risk Modelling, Credit Risk Modelling Using SAS, Credit scorecard modeling, Expected credit loss ECL, IFRS 9 modeling course India, Logistic regression scorecard, PD LGD EAD modeling, Python credit risk modeling, Risk Analytics Market, Risk Management Courses, risk management courses online, Risk Management in Banking
Comments are closed here.