| Borrower | Exposure | EAD | PD | LGD | Expected Loss | Grade |
|---|---|---|---|---|---|---|
| Retail SME | term | 257,000.00 | 77.41% | 26.23% | 52,187.44 | E (Very High) |
| Manufacturing | term | 1,240,500.00 | 46.88% | 12.92% | 75,114.91 | E (Very High) |
| Revolving Trader | revolving | 860,000.00 | 91.56% | 5.00% | 39,370.76 | E (Very High) |
Formula used
How to use this calculator
- Enter borrower details and choose exposure type.
- Provide income, debt, collateral, and behavioral indicators.
- Select a confidence level for the VaR estimate.
- Click Calculate to see results above the form.
- Adjust inputs to compare scenarios and risk controls.
- Use CSV/PDF downloads to keep an audit trail.
Risk inputs and benchmark ranges
This calculator combines borrower behavior, affordability, and collateral strength to estimate risk. Debt-to-income (DTI) is computed from monthly debt divided by monthly income, while loan-to-value (LTV) compares loan size to collateral value. Higher utilization on revolving lines, more delinquency days, and a higher industry risk score (1–5) increase the risk score. Higher credit scores and longer operating tenure reduce the risk score for stable borrowers. Many lenders flag DTI above 40%, LTV above 90%, utilization above 80%, or delinquencies of 30+ days as early-warning triggers for review.
Reading PD and risk grades
Probability of Default (PD) is produced by a logistic score that maps the risk score into a percentage. As a rule of thumb, PD near 1% suggests low-frequency default, while PD above 10% signals elevated risk and tighter monitoring. The risk grade summarizes PD into bands for quick screening. For planning, interpret PD as an annualized likelihood over a one-year horizon: a 2% PD implies about two expected defaults per 100 similar exposures.
Collateral impact on recovery and LGD
Loss Given Default (LGD) reflects how much exposure may be lost if default occurs. The calculator estimates recovery from collateral coverage, then sets LGD as one minus recovery. A coverage ratio above 1.0 generally improves recovery, but higher industry risk and longer delinquency reduce it. Recovery is capped between 5% and 95% to avoid unrealistic extremes, making LGD useful for scenario comparisons.
EAD for term and revolving products
Exposure at Default (EAD) depends on product structure. For term lending, EAD is the loan amount with a small buffer for accrued cost. For revolving credit, EAD uses a credit conversion factor (CCF) applied to the undrawn limit: EAD = Balance + CCF × (Limit − Balance). Example: limit 1,000,000, balance 600,000, CCF 75% gives EAD 900,000, capturing potential drawdown risk.
Using EL and VaR for decisions
Expected Loss (EL) is the core planning metric: EL = PD × LGD × EAD. Use EL to compare borrowers, set provisions, and test pricing assumptions. Credit VaR adds a confidence-based cushion around EL using the variance of default, producing a tail-loss estimate at your chosen confidence level. Unexpected Loss equals VaR minus EL and can guide limits, collateral requirements, and stress actions when markets weaken or cashflows tighten.
FAQs
PD is the estimated chance of default for a similar exposure under current inputs. It is a scenario estimate, not a guarantee, and should be calibrated to your portfolio data when possible.
Logistic scoring is nonlinear. When inputs move around common cutoffs, such as higher DTI or added delinquency days, the probability can shift quickly. Compare scenarios with incremental changes to understand sensitivity.
Collateral value feeds an estimated recovery rate, then LGD is calculated as one minus recovery. Legal costs, time-to-recover, and market stress are simplified, so treat LGD as directional for analysis.
Use revolving mode for credit cards, overdrafts, and working-capital limits where borrowers can draw additional funds. Provide limit, current balance, and a conversion factor to reflect likely drawdown before default.
Credit VaR is a single-exposure tail-loss approximation at your confidence level. It adds a statistical cushion around expected loss using default variance. It is best for ranking risk, not for regulatory capital.
Run at least one calculation so it is saved in history. Then download CSV for spreadsheets or PDF for a quick report. Clear history when you want a fresh audit trail for a new borrower.