How are credit conversion factors (CCFs) estimated for off-balance-sheet exposures, and why do they matter for EAD?

Credit conversion factors (CCFs) translate off-balance-sheet commitments into on-balance-sheet equivalent exposures for the purpose of calculating Exposure at Default (EAD). They answer the question: \"How much of the undrawn commitment will the borrower draw down before defaulting?\"\n\nEAD Formula:\n\nEAD = Drawn Amount + CCF x Undrawn Amount\n\nExample:\n\nMeadowbank Corp has a $10M revolving credit facility from Ridgeline National Bank.\n- Currently drawn: $6M\n- Undrawn commitment: $4M\n- CCF (estimated): 65%\n\nEAD = $6M + 0.65 x $4M = $6M + $2.6M = $8.6M\n\nThis means the bank expects Meadowbank to draw an additional $2.6M before defaulting, bringing total exposure to $8.6M rather than just the current $6M.\n\nWhy Borrowers Draw Down Before Default:\n\nDistressed borrowers tend to draw on available credit lines as a last resort for liquidity. Empirical evidence consistently shows that defaulting borrowers have higher utilization at default than at any earlier observation point. This \"drawdown before default\" phenomenon is why CCFs are critical.\n\nStandardized Approach CCFs (Basel):\n\n| Commitment Type | Maturity | CCF |\n|---|---|---|\n| Unconditionally cancellable | Any | 0% (10% under Basel III) |\n| Other commitments | <= 1 year | 20% |\n| Other commitments | > 1 year | 50% |\n| Direct credit substitutes | Any | 100% |\n| Trade-related contingencies | Any | 20% |\n| NIF / RUF | Any | 50% |\n\nAdvanced IRB Estimation:\n\nUnder the Advanced IRB approach, banks estimate their own CCFs using internal data. The process at Ridgeline National Bank:\n\n1. Collect historical data: for every defaulted borrower over the past 10 years, record the drawn amount 12 months before default and at the point of default\n2. Compute realized CCFs: CCF_realized = (EAD - Drawn_12m) / Undrawn_12m\n3. Segment by risk factors: facility type, borrower rating, industry, drawn percentage\n4. Estimate conditional mean: for each segment, compute the average realized CCF\n\nSample Data (Ridgeline's revolving credit portfolio):\n\n| Borrower | Drawn (12m prior) | Undrawn (12m prior) | EAD at default | Realized CCF |\n|---|---|---|---|---|\n| Firm A | $5.0M | $5.0M | $8.5M | 70% |\n| Firm B | $3.2M | $6.8M | $7.4M | 62% |\n| Firm C | $7.5M | $2.5M | $9.8M | 92% |\n| Firm D | $4.0M | $6.0M | $6.8M | 47% |\n| Average | | | | 68% |\n\nNotably, Firm C had very little headroom remaining, and distressed borrowers with high existing utilization tend to draw nearly everything, pushing CCFs toward 100%.\n\nKey Insight: CCFs are positively correlated with borrower credit quality deterioration. This creates a double-hit: the borrower becomes riskier (higher PD) AND draws more (higher EAD) simultaneously.\n\nExplore credit risk modeling in our FRM study resources.