Loss given default is a core measure in bond credit analysis. It estimates how much value an investor may lose if an issuer fails to meet its debt obligations and a default event occurs. While probability of default focuses on how likely a borrower defaults, loss given default focuses on what happens after default. For bondholders, this distinction is important because two issuers may have similar default probability but very different loss outcomes.
Loss given default lgd is usually expressed as a percentage of exposure. A 30% lgd means that 30% of the exposure is expected to be lost, while 70% is expected to be recovered through restructuring, asset sales, collateral enforcement, or other recoveries. In simple terms, lgd measures the part of outstanding debt that is not expected to come back to lenders or bond investors after default.
The basic calculation can be shown by the following formula:
LGD = (Exposure at Default - Recovery) / Exposure at Default
For example, if a bond investor has €1,000 of exposure at default and expects to recover €600 after restructuring costs, the lgd is 40%. The recovery rate is 60%, and the loss given default is the remaining percentage not recovered. This makes lgd the mirror image of the recovery rate in many practical credit models.
Loss given default lgd is critical because default risk alone does not determine bond downside. A senior secured bond issued by a company with valuable collateral may have a lower lgd than an unsecured or subordinated bond issued by the same borrower. The probability of default may be similar across the issuer’s instruments, but the recovery rate can differ significantly across the capital structure.
Senior debt typically has a lower lgd than subordinated debt because it is repaid first in bankruptcy or restructuring. This hierarchy is one reason why senior bonds usually trade at lower yields than junior bonds from the same issuer. Investors accept a lower spread because the expected loss given default is lower. Subordinated bonds, by contrast, may offer higher yields, but they usually absorb losses earlier if the borrower defaults.
In bond portfolio management, lgd helps determine whether a credit spread adequately compensates for risk. A high-yield bond with a high coupon may look attractive, but if expected recoveries are weak, the potential loss in default may be substantial. For this reason, professional credit analysis does not stop at yield, rating, or spread. It also looks at capital structure, collateral value, asset coverage, legal ranking, and restructuring outcomes.
Loss given default lgd is one of the three main components used for calculating expected loss. The standard relationship is:
Expected Loss = EAD x PD x LGD
EAD means exposure at default. PD means probability of default. LGD means loss given default. This formula shows why lgd is not a secondary detail. A bond with low probability of default but very high lgd can still create meaningful downside. A bond with higher probability of default but strong recoveries may have a more balanced expected loss profile.
For instance, assume two bonds each have €1,000 exposure. Bond A has a 2% pd and 70% lgd. Bond B has a 4% pd and 25% lgd. Bond A has an expected loss of €14, while Bond B has an expected loss of €10. The example shows that probability and given default loss must be analysed together.
This is especially relevant for portfolio construction. If a portfolio has many bonds with high lgd values, diversification may not fully protect investors during stress. Defaults can cluster during downturn periods, and recoveries can fall at the same time. The combined effect can increase portfolio loss more than a simple average default calculation would suggest.
LGD is influenced by several factors. The most important are debt seniority, collateral quality, recovery costs, legal environment, issuer asset value, and macroeconomic conditions. These variables determine how much value can be recovered after default and how much is lost during the process.
Collateral is one of the most direct drivers. Loans or bonds backed by high-quality collateral generally have lower lgd because assets can be sold to recover money. A secured bond backed by infrastructure assets, receivables, ships, aircraft, real estate, or cash-generating operating assets may have better recovery prospects than an unsecured bond with no specific collateral claim.
Loan-to-value is also important. A lower loan-to-value ratio means more collateral supports the loan or secured bond, reducing potential lgd. If a borrower has €100 million of debt secured by assets worth €200 million, expected recoveries may be stronger than if the same debt is secured by assets worth only €110 million. However, asset values can change, especially during economic downturns.
Seniority is another key factor. Senior secured debt usually has the best recovery profile. Senior unsecured bonds rank below secured debt but above subordinated instruments. Subordinated bonds and hybrid securities often have higher lgd because they are paid only after more senior creditors. This ranking becomes especially important when enterprise value is insufficient to repay all creditors.
Recovery costs can also raise lgd. High legal fees, administrative expenses, advisor costs, and slow legal proceedings reduce the value ultimately recovered by lenders and bondholders. A long restructuring can also damage the borrower’s operating business, lowering asset value and reducing recoveries.
| Instrument type | Typical ranking | Typical recovery profile | Main LGD driver |
|---|---|---|---|
| Senior secured bond | High priority with collateral claim | Lower LGD in many cases | Collateral value and legal enforceability |
| Senior unsecured bond | Senior claim without specific collateral | Moderate LGD depending on asset coverage | Enterprise value and amount of secured debt ahead |
| Subordinated bond | Below senior creditors | Higher LGD in many cases | Residual value after senior creditor recoveries |
| Hybrid bond | Often deeply subordinated | Potentially high LGD | Equity-like features and issuer discretion |
| Bank senior preferred debt | Senior bank claim | Depends on resolution framework | Regulatory treatment and bail-in hierarchy |
| Bank subordinated debt | Below senior bank debt | Higher LGD under stress | Capital structure ranking and resolution losses |
LGD can be calculated using different methods. A common method is gross lgd, where total losses are divided by exposure at default. This method looks at the full exposure and measures the share not recovered after default. It is intuitive and widely used in credit risk models because it links directly to total economic loss.
Another method is sometimes referred to as Blanco lgd. This approach considers only the unsecured portion of a credit line or exposure. In capital markets analysis, the distinction matters because a secured instrument may have both covered and uncovered components. If collateral fully covers part of the exposure, the loss calculation may focus on the portion not protected by collateral.
For bond investors, the practical question is less about terminology and more about what the lgd model is trying to capture. A model based on historical bond recoveries may produce different lgd estimates from a collateral-based model. A model using market prices after default may also produce different values from a final workout recovery method, because recoveries can change over time as restructuring terms become clearer.
Loss given default lgd varies across the economic cycle. In normal conditions, asset values may be stable, credit markets may remain open, and distressed issuers may have access to refinancing or asset buyers. In a downturn, the same borrower may face lower asset prices, weaker earnings, reduced refinancing options, and slower recoveries.
This is why analysts often distinguish between long-run lgd and downturn lgd. Long-run lgd is usually based on average recovery experience across a broad period. Downturn lgd represents lgd at the worst time of the economic cycle or under stressed macroeconomic conditions. Downturn lgd is usually higher for corporate borrowers because economic downturns can reduce asset values and lower the recovery rate.
For example, the collateral value of real estate, commodities, ships, or industrial assets may fall exactly when default risk rises. In such a situation, both pd and lgd deteriorate at the same time. This procyclicality is important for credit risk models because the expected loss can rise sharply even if the initial lgd estimates looked manageable during benign market conditions.
However, the relationship is not identical across all segments. During economic downturns, lgd values can sometimes decrease for defaulting financial institutions because governments and central banks may intervene to maintain financial stability. Support measures, resolution tools, liquidity facilities, or forced mergers can improve recoveries for some creditors. This makes financial institutions different from many corporate issuers, where downturn recoveries are often weaker.
Banks and other financial institutions use lgd models for credit risk management, pricing, provisioning, and regulatory capital. Under Basel II, banks must estimate lgd for corporate, sovereign, and bank exposure to determine required capital. They can use a foundation approach with fixed lgd ratios or an advanced approach based on internal data, subject to regulatory standards.
In the advanced approach, institutions use historical data, borrower characteristics, collateral information, seniority, recovery timing, and macroeconomic variables to estimate lgd. These models are important because lgd directly affects capital requirements. Higher lgd means higher expected and unexpected loss, which can lead to higher required capital.
There are practical challenges. Financial institutions often face difficulties estimating downturn lgd because historical data may be limited. Definitions of default can also vary across institutions, making lgd parameters less comparable. Recovery data may be incomplete, especially when workouts take many years or when defaulted exposures are sold before final recovery is known.
For bond investors, an lgd model does not need to be identical to a bank regulatory model. A market-oriented lgd model may use bond pricing, historical recoveries by rating, sector, seniority, and jurisdiction. It may also include issue-level features such as collateral, covenants, guarantees, maturity, and structural subordination. The purpose is to determine whether the bond spread compensates for expected loss and downside risk.
The effectiveness of debt restructuring affects lgd in corporate loans and bonds. A fast and coordinated restructuring can preserve operating value, reduce uncertainty, and improve recoveries. A disorderly default can destroy value through customer losses, supplier pressure, employee departures, legal disputes, and asset fire sales.
Capital structure complexity also matters. If the borrower has many different types of debt, including secured loans, senior unsecured bonds, subordinated bonds, leases, guarantees, and local operating company debt, recoveries may be difficult to estimate. Investors must determine where each instrument sits in the legal structure and which creditors have access to which assets.
Structural subordination is a frequent issue in bond analysis. A bond issued by a holding company may be structurally subordinated to debt at operating subsidiaries. Even if the bond is labelled senior unsecured, its practical recovery rate may be weaker if operating assets are pledged to subsidiary lenders. This is why lgd estimates should account for legal entity structure, not only headline ranking.
Covenants may also influence lgd. Strong covenants can restrict additional secured debt, asset sales, dividend payments, or aggressive refinancing transactions. Weak covenants can allow the issuer to move value away from bondholders before default. In such cases, the recovery rate may be lower than initially expected.
Bond markets price default and recovery expectations continuously. A widening credit spread can reflect higher pd, higher lgd, or both. Distressed bond prices often provide an implied recovery estimate. If a defaulted bond trades at 35 cents on the euro, the market may be signalling an expected recovery rate near 35%, although prices also include timing, uncertainty, liquidity, and investor required return.
Calculating implied lgd from market prices requires caution. A distressed bond price is not always equal to final recovery value. Investors may demand a discount for uncertainty, legal complexity, or the time needed to obtain recovery. Conversely, a bond may trade above expected recovery if investors expect a restructuring that includes new debt, equity, or warrants with upside potential.
Credit analysts therefore combine market tools with fundamental models. They assess asset value, debt ranking, collateral, business viability, jurisdiction, and restructuring incentives. The best lgd estimates are usually not derived from a single formula alone. They require a combination of data, judgement, and scenario analysis.
Consider a company with €500 million of senior secured debt, €400 million of senior unsecured bonds, and €200 million of subordinated bonds. After default, the estimated enterprise value available to creditors is €650 million before recovery costs. If secured lenders have first claim and recover €500 million, only €150 million remains for senior unsecured creditors. Subordinated creditors may receive little or nothing.
For the senior secured debt, lgd may be close to 0% before costs if the claim is fully covered. For the senior unsecured bonds, the recovery rate may be 37,5%, calculated as €150 million recovered on €400 million exposure. The lgd would therefore be 62,5%. For subordinated bonds, the recovery rate may be 0%, implying 100% lgd.
This example shows why given default loss can differ significantly across debt instruments issued by the same borrower. The default event is the same, but the loss percentage is not. This is one of the reasons why bond investors analyse capital structure in detail before comparing yields.
LGD estimation is inherently uncertain. It depends on future asset values, restructuring terms, legal processes, market liquidity, creditor behaviour, and macroeconomic conditions. Even sophisticated models can be wrong if they rely on weak data or if the next default cycle differs from previous cycles.
Historical recovery data can be useful, but it may not capture new risks. Sector changes, regulation, inflation, interest rates, and financing conditions can alter recoveries. For example, an asset-heavy issuer may appear well protected, but if the assets are difficult to sell during stress, the actual recovery rate may disappoint. Likewise, a borrower with few tangible assets may still deliver reasonable recoveries if the operating franchise remains valuable.
This is why lgd estimates should be periodically re-evaluated. Both lgd and pd are influenced by economic conditions and can vary significantly during economic cycles. A model that was reasonable in a low-rate, liquid market may understate risk in a downturn with weak refinancing access and falling collateral values.
Loss given default lgd is one of the most important measures in bond credit risk analysis. It quantifies how much a lender or bond investor may lose if a borrower defaults, while pd estimates the probability that the default will occur. Together with exposure at default, lgd and pd determine expected loss.
For bond investors, lgd is especially important because recoveries differ across seniority, collateral, issuer type, jurisdiction, and economic cycle. Senior secured bonds usually have lower lgd than subordinated bonds, but this depends on collateral quality, asset value, recovery costs, and legal enforceability. Downturn lgd is particularly relevant because default rates and loss severity can rise at the same time.
A disciplined bond analysis should therefore combine probability of default, recovery rate, capital structure analysis, and lgd model outputs. Yield alone is not enough. The key question is not only whether a default may occur, but also how much value investors may lose given default.