Credit Risk

Credit risk is quantified as an expected loss built from three inputs. The probability of default measures the likelihood that the obligor fails to perform over a stated horizon. Loss given default is the fraction of exposure not recovered, equal to one minus the recovery rate. Exposure at default is the amount at risk when the failure occurs. Expected loss is the product of these three terms, and it is the amount a lender should expect to lose on average, not the loss in any single outcome.

Because a bond pays over many periods, credit analysis discounts the expected loss of each period back to today. The sum of those present values is the credit valuation adjustment. The value of a risky bond is therefore the value of an otherwise identical default-free bond less the credit valuation adjustment, which is the cleanest way to see that credit risk is a deduction from price rather than an add-on to yield alone.

Default is an event at the issuer level, but loss severity is determined at the instrument level. Bankruptcy regimes broadly apply absolute priority: secured creditors are satisfied from collateral first, then senior unsecured claims, then subordinated debt, then preferred and common equity. Instruments of the same rank are pari passu and share proportionally in whatever value the estate produces. In practice, negotiated reorganisations often deviate from strict priority, so junior classes sometimes recover more than the legal hierarchy implies.

This structure explains why one issuer carries several ratings. Agencies assign a corporate family or issuer rating reflecting default probability, then notch individual issues up or down for expected recovery. Secured obligations of a weak issuer can be rated above the issuer itself, while deeply subordinated or holding-company debt is notched down. Structural subordination, where debt sits at a parent above operating subsidiaries, produces the same effect without any contractual subordination clause.

In the market, credit risk is expressed as the yield spread over a benchmark curve, measured as a G-spread, I-spread, Z-spread or option-adjusted spread depending on the reference curve and embedded optionality. That spread compensates for expected loss, for the uncertainty around it, and for liquidity. Credit default swap premiums provide a parallel and often more liquid quotation of the same risk, with the upfront payment and standardised coupon reconciled through the recovery assumption used at auction.

Most realised pain arrives before default. Downgrades, deteriorating leverage and coverage ratios, or sector-wide risk aversion widen spreads and mark positions down immediately. The approximate return effect is minus the spread duration multiplied by the change in spread, plus a convexity term for larger moves. Credit risk also appears outside bond portfolios entirely: in bank loan books, trade receivables, and as counterparty risk on derivative exposures, where collateral and netting agreements reduce exposure at default.

Expect to compute expected loss from a stated default probability and recovery rate, and to convert between loss given default and recovery. Expect to value a risky bond as a default-free bond less a credit valuation adjustment built from period-by-period expected losses. Expect to translate a spread change into a price or return impact using spread duration, and to rank instruments by expected recovery given their position in the capital structure.

Qualitative command matters equally. Be able to describe the traditional credit factors, capacity, collateral, covenants and character, and to interpret leverage, coverage and profitability ratios in a credit rather than equity frame. Know the investment-grade boundary, the meaning of issuer versus issue ratings, and the limitations of ratings: they are through-the-cycle opinions on default risk, they lag market prices, and they say nothing about liquidity or the volatility of spreads along the way.