Loss given default, usually shortened to LGD, is the share of a loan or credit exposure a lender actually loses when a borrower fails to pay, expressed as a percentage. If a bank lends $1 million and recovers $600,000 after the borrower defaults, the LGD is 40%. Banks use the figure to price loans, set aside reserves, and calculate the regulatory capital they must hold under the Basel framework.
The Basic Formula
LGD begins with a simple relationship: LGD equals one minus the recovery rate. A lender that gets back 55 cents on the dollar from a defaulted loan has a 55% recovery rate and a 45% LGD. The recovery rate captures everything the lender collects through collateral sales, restructured payments, or selling the defaulted debt to a third party, net of the costs of collecting it.
The complication is that “recovery” is not just “principal minus whatever came back.” A proper LGD calculation adjusts for the direct costs of running a workout, the indirect economic burden of holding a non-performing asset, and the time value of money across a recovery period that often stretches for years. The denominator is exposure at default (EAD), which itself takes work to estimate for anything more complicated than a fully drawn term loan.
What Counts as an Economic Loss
Three categories of cost feed into the numerator.
Direct Recovery Costs
These are out-of-pocket expenses tied to the recovery effort: legal fees for foreclosure or litigation, court filing costs, appraisal fees, and the allocated salaries of internal workout staff. On a secured loan, these costs come straight out of whatever the collateral produces. A property that sells for $400,000 but costs $30,000 in legal and liquidation fees delivers only $370,000 in net recovery.
Indirect and Opportunity Costs
Capital tied up in a non-performing loan cannot be deployed elsewhere, and management attention diverted to the workout is not free either. These costs are harder to quantify but matter for capturing the real economic damage of a default.
Discounting for Time
Recovery cash flows often arrive over several years. A lender that recovers $500,000 three years after default has not truly recovered $500,000 in economic terms, because the money sat unavailable in the meantime. Workout LGD calculations discount all future recoveries back to the date of default.
The choice of discount rate meaningfully affects the final number. Three approaches dominate. One uses the original contractual loan rate, on the logic that the lender’s opportunity cost is the yield it expected to earn. A second uses the risk-free rate, which produces a lower discount and a lower LGD, but arguably understates true economic cost. A third uses a risk-adjusted rate estimated from market data, which is theoretically cleanest but demands more modeling. No single approach is universally required; the choice depends on the institution’s internal methodology and its supervisor’s guidance.
Exposure at Default: The Denominator
EAD is the total amount at risk when the borrower defaults. Get it wrong and the whole LGD ratio skews, whether the result feeds regulatory capital, loan pricing, or provisioning.
Term Loans
For a fully drawn term loan, EAD is the outstanding principal balance plus accrued but unpaid interest. If you lent $2 million and the borrower has paid down $400,000, the EAD is roughly $1.6 million plus accrued interest.
Revolving Facilities
Revolving lines are harder because the borrower can draw more right up to the moment of default, and distressed borrowers often do exactly that. EAD for a revolver depends on a credit conversion factor (CCF), which estimates how much of the undrawn commitment will be tapped by the time default happens.
Under the Basel standardized approach, CCFs run from 10% for commitments the bank can cancel at any time without notice up to 100% for firm commitments like standby letters of credit and forward purchases. General commitments that don’t qualify for a lower factor carry a 40% CCF, and trade-related contingencies and note issuance facilities receive 50%.1Bank for International Settlements. CRE20 – Standardised Approach: Individual Exposures
The formula is: amount drawn, plus the undrawn commitment multiplied by the CCF. A borrower with a $100,000 line who has drawn $30,000, where the remaining $70,000 carries a 40% CCF, has an EAD of $58,000.
Derivatives
Derivatives have no loan balance in the usual sense; their value moves with the market. Under the Standardized Approach for Counterparty Credit Risk (SA-CCR), EAD equals 1.4 times the sum of replacement cost and a potential future exposure add-on, with the add-on scaled by asset class, notional, and maturity.2Bank for International Settlements. CRE52 – Standardised Approach to Counterparty Credit Risk
How Banks Estimate LGD
Banks do not calculate LGD one default at a time and stop there. They use systematic methods across portfolios, and the method depends on data availability, exposure type, and whether the estimate feeds regulatory capital or internal risk decisions.
Workout LGD
The workout method is the most granular. The bank tracks every cash flow from each defaulted loan, subtracts direct and indirect costs, discounts the net recoveries back to the default date, and expresses the loss as a fraction of EAD. It relies entirely on internal data, so it is institution-specific.
The trade-off is time. Workouts commonly run five to seven years, so a large share of the defaulted portfolio is still in progress at any given moment. Banks use statistical techniques to project final recoveries on incomplete cases. And by the time a workout finishes, the economic conditions it reflects may no longer resemble today’s.
Market LGD
When defaulted debt trades on a secondary market, the trading price provides a real-time recovery estimate. If a defaulted bond trades at $35 per $100 of face value, investors are collectively pricing recovery near 35%, implying an LGD of roughly 60% once accrued interest and trading costs are considered. The method is fast and forward-looking. It only works, though, for debt that actually trades. Publicly issued corporate bonds and syndicated loans have observable prices; bilateral loans to mid-market borrowers do not. Market prices also reflect sentiment and liquidity, which can drift from fundamental recovery value in stressed periods.
Statistical Modeling
Rather than lean on portfolio averages, statistical models predict LGD for a specific exposure from its characteristics: seniority, collateral type and appraised value, borrower industry, loan-to-value, and macroeconomic variables such as unemployment and property price indices. The output is a predicted LGD for each individual facility, allowing sharper risk differentiation than a single portfolio-wide number.
What Drives the Number
Three factors explain most of the variation in realized LGD.
Seniority
Where a claim sits in the repayment hierarchy is the strongest single predictor of recovery. Senior secured creditors are paid first from the borrower’s assets; senior unsecured holders come next; subordinated holders are last. Long-run data on defaulted corporate bonds shows senior secured issues recovering roughly 50 cents on the dollar, senior unsecured about 33 cents, and subordinated about 27 cents.3Moody’s Investors Service. Recovery Rates on Defaulted Corporate Bonds and Preferred Stocks Individual outcomes vary widely around those averages, but the ranking is consistent.
Collateral Quality and Perfection
Collateral only helps if the lender’s claim is legally enforceable and if the asset holds value through the workout. A properly filed security interest in commercial real estate provides real downside protection. An improperly documented lien on depreciating equipment may provide almost none. Real estate and financial assets tend to keep value and have liquid resale markets. Specialized industrial equipment, perishable inventory, or intangibles like intellectual property are harder to move and typically produce lower recoveries. Conservative, frequent appraisals matter, because models that rely on stale valuations will understate loss when the underlying asset has deteriorated.
The Economic Cycle
LGD is procyclical. In a downturn, default rates rise at the same time recovery values fall: property prices drop, buyers for distressed assets disappear, and workout timelines lengthen. That correlation is why regulators require “downturn LGD” rather than through-the-cycle averages.
Expected Loss: Where LGD Fits
LGD is one of three inputs in the expected loss formula that anchors credit risk management: EL = PD × LGD × EAD, where PD is the probability of default.4Bank for International Settlements. CRE35 – IRB Approach: Treatment of Expected Losses and Provisions
Each component answers a different question. PD asks how likely default is. EAD asks how much is at stake. LGD asks how bad it gets when default actually happens. A loan with a 2% PD, $1 million EAD, and 40% LGD carries an expected loss of $8,000. That figure drives loan pricing, provisioning, and portfolio-level capital allocation.
Expected loss covers only the average outcome. Regulatory capital exists to cover unexpected losses beyond the average, which is why Basel runs the same PD, LGD, and EAD inputs through a risk-weight function calibrated to a high-confidence tail.
LGD Under Basel
Under the Internal Ratings-Based (IRB) approach, banks choose between a Foundation method, where regulators prescribe LGD values, and an Advanced method, where banks estimate LGD from their own data with supervisory approval.5Bank for International Settlements. CRE36 – IRB Approach: Minimum Requirements to Use IRB Approach
Foundation IRB
Foundation banks use fixed LGD values. Senior unsecured claims on most corporates get 40%. Senior claims on banks, sovereigns, and other financial institutions get 45%. Subordinated claims receive 75%. Eligible collateral pulls the number down. Exposures fully secured by eligible financial collateral can reach 0%. Real estate and receivables collateral bring the secured portion to 20%, and other physical collateral to 25%.6Bank for International Settlements. CRE32 – IRB Approach: Risk Components
Advanced IRB
Banks approved for Advanced IRB estimate their own LGD for each exposure. That means more risk sensitivity but also more scrutiny: robust data, validated models, and governance are all required, and each estimate must be based on the bank’s own loss experience over a minimum observation period.5Bank for International Settlements. CRE36 – IRB Approach: Minimum Requirements to Use IRB Approach
Downturn LGD
Fair-weather loss estimates are not allowed for capital. LGD estimates must reflect downturn conditions where relevant, and cannot fall below the long-run default-weighted average loss rate.5Bank for International Settlements. CRE36 – IRB Approach: Minimum Requirements to Use IRB Approach European regulators have added that a downturn LGD estimate, including a margin of conservatism, should sit at least 15 percentage points above the long-run average LGD, capped at 105%.7European Banking Authority. Guidelines for the Estimation of LGD Appropriate for an Economic Downturn The conservative calibration reflects the point of capital buffers: they exist to absorb losses when conditions are worst, not average.
LGD Under CECL
The Current Expected Credit Losses (CECL) standard, codified in FASB’s Topic 326, changed how U.S. institutions use LGD for accounting. The previous “incurred loss” model recognized losses only after a trigger event. CECL requires estimating expected credit losses over the full remaining life of each financial asset, moving loss recognition into current-period provisions.8National Credit Union Administration. CECL Accounting Standards
The lifetime horizon differs sharply from the one-year window that dominates Basel IRB. Under CECL, LGD estimates must cover the contractual term and incorporate a reasonable and supportable forward-looking forecast; beyond that horizon, the estimate reverts to historical loss experience. Accounting LGDs therefore tend to be more forward-looking and scenario-dependent than the LGDs used purely for regulatory capital.
In July 2025, FASB issued ASU 2025-05, which eases the forecasting burden for accounts receivable and contract assets by allowing entities to assume current balance-sheet-date conditions persist for the remaining life of those assets.9Financial Accounting Standards Board. FASB Issues Standard that Improves Measurement of Credit Losses for Accounts Receivable and Contract Assets The simplification does not reach loan portfolios, where full lifetime LGD estimation still applies.
Tax Treatment When the Loss Is Realized
When an estimated LGD turns into a realized loss, tax rules take over. Under federal tax law, a business debt that becomes wholly worthless during the tax year is deductible in the full amount of the loss. A debt that is only partially worthless can be deducted to the extent charged off during the year, provided the IRS is satisfied that full recovery is not possible.10Office of the Law Revision Counsel. 26 USC 166 – Bad Debts
The deduction is measured against the lender’s adjusted basis in the debt, not the original face value or the EAD used for regulatory purposes. Debts evidenced by a security, such as a bond, fall under separate rules rather than the general bad debt provision, and banks have their own parallel treatment for worthless securities. The distinction affects timing and character, so the same loss can look different on the tax return depending on whether it came off a loan on the books or a security in the portfolio.