FI 14 Credit Risk

Credit risk is the danger of economic loss that arises when a borrower does not pay interest and principal in full and on schedule. A borrower that misses a promised payment under a bond or loan contract is in default, and the fixed-income investor wants to be paid for the loss a possible default could inflict across the life of the contract. It is a form of performance risk inside a contractual relationship, and it shifts over time with both the borrower itself and the wider economy. Note that credit risk exists before any actual loss occurs: an expected loss is priced into the bond long before a borrower ever defaults.

An inability to pay on time ultimately comes down to a shortage of cash on the payment date. It helps to separate two situations. An illiquid borrower cannot raise the funds needed to meet an obligation as it comes due, because it cannot tap credit lines, sell assets, or otherwise secure cash in time. An insolvent borrower is different: its assets are worth less than its liabilities. A borrower can be illiquid without being insolvent. When a lender is not repaid, part or even all of the principal and interest due can be lost, producing a capital loss or an income shortfall, and that disruption can ripple into the lender’s own ability to service its debts, on top of added legal and collection costs.

The Cs of credit analysis

A traditional way to organize creditworthiness is the set of criteria known as the Cs of credit analysis. Five are bottom-up factors tied to the individual borrower, and three are top-down factors bearing on every borrower to some degree.

Capacity and capital are mostly quantitative, read straight from the financial statements. The other three, collateral, covenants, and character, are more subjective, judged from a borrower’s history, its credit relationships, and the standing of its management.

Where repayment comes from

How credit risk changes depends on the source of repayment, which differs by issuer type. A corporation repays chiefly from the cash its operations, investments, and financing activities generate; if that cash falls short, default risk rises. For unsecured debt the cash flow of the business is the primary source of repayment. For secured debt the firm’s cash flow is still primary, but specific collateral pledged by the company acts as a secondary source. A sovereign government instead relies on its power to tax economic activity inside its borders, so its repayment comes from income taxes, sales and value-added taxes, tariffs, fees, and similar revenue.

The two components of credit risk

The probability-weighted shortfall an investor expects over a given period is the expected loss (EL), and it has two components. The probability of default (POD) is the chance that the issuer will not pay principal and interest in full and on time, and it is usually stated as an annualized figure. The loss given default (LGD) is the investor’s loss conditional on a default having occurred. LGD combines the size of the claim with how much is recovered: the expected exposure (EE), also called exposure at default, is the claim outstanding at the time of default (roughly face value plus accrued interest less the market value of any available collateral), and the recovery rate (RR) is the percentage of that claim recovered, so the loss severity is the unrecovered fraction, (1 − RR).

The expected loss for a period can be compared with the compensation the investor collects for bearing credit risk, which is the credit spread. Recall that the G-spread is measured in basis points against an actual or interpolated government bond of matching maturity. As an approximation:

An investor is fairly compensated when the credit spread earned at least matches the expected loss over the period. If the spread exceeds POD times LGD, the investor is more than fairly paid for the risk.

What drives each component? The POD reflects an issuer’s ability to service debt, captured by profitability (stable, predictable cash flows and profits, for example EBIT margin), coverage (enough cash flow to cover payments, for example EBIT to interest expense), and leverage (reliance on debt, for example debt to EBITDA or cash flow to net debt). Higher profitability and coverage together with lower leverage point to a lower POD and higher credit quality. The LGD, by contrast, turns mainly on the nature and seniority of the claim: secured debt ranks higher and carries a lower LGD, while unsecured debt sits lower and carries a higher LGD. Investors in unsecured investment-grade debt therefore worry most about a rise in POD, whereas high-yield investors, already braced for default, try to hold down LGD through covenants or security.

Three agencies dominate the market: Moody’s Investors Service, Standard and Poor’s (S&P), and Fitch Ratings. Each independently and on a forward-looking basis assesses issuer credit risk using both quantitative and qualitative analysis, and for most sizable corporate and sovereign issues, ratings come from at least two of them. A credit rating is a symbol-based gauge of the potential default risk carried by a specific bond issue or its issuer. Investors use ratings to compare creditworthiness quickly across issuers, industries, and bond types, and rating changes give a broad read on shifting credit conditions. In specific bonds a change can trigger covenants or reset the pricing of step-up bonds. The risk that an issuer’s creditworthiness deteriorates and its rating migrates lower, prompting investors to see a higher chance of default, is credit migration risk, also called downgrade risk.

Ratings are commissioned on behalf of the issuer, and to produce one the agency often meets management and may receive material non-public information such as internal projections. After a rating is assigned the agency keeps monitoring, moving it up as credit risk falls or down as default looks more likely, and it may attach a positive or negative outlook when creditworthiness is drifting but a change is not yet warranted.

The rating scale

The three agencies use similar symbol-based scales for long-term debt (maturities beyond one year), ranked from highest quality to lowest. The single most important dividing line is between investment grade and non-investment grade.

Bonds rated Baa3 or BBB− and above are investment grade (IG). A rating of Ba1 or below at Moody’s, or BB+ or below at S&P and Fitch, marks non-investment grade, known variously as speculative grade, below investment grade, high yield (HY), or, more bluntly, junk. Investment-grade issuers can generally issue debt more consistently and borrow at lower rates. Compared with high-yield bonds, investment-grade bonds carry a lower risk profile, are less hurt by adverse economic and market conditions, and suit institutional portfolios that face quality restrictions.

Why ratings are not enough on their own

Ratings are useful, but market participants normally run their own analysis rather than lean solely on them. Three pitfalls stand out.

First, ratings tend to be sticky and lag the market pricing of credit risk. Prices and spreads move daily and often faster than agencies revise ratings, and even over long stretches ratings can trail prices. Two speculative-grade bonds with the same rating may trade at very different spreads, because ratings aim mainly at expected loss while distressed pricing focuses more on the timing of default and expected recovery. Investors who wait for a rating change before acting may lag those who move ahead of it.

Second, some risks are hard to capture in a rating: litigation, environmental hazards, natural disasters, and leveraged events such as debt-financed acquisitions or large stock buybacks, which are difficult to anticipate. Agencies can also read complex risks very differently, producing split ratings. When WeWork Inc., an asset-light US real estate start-up with negative cash flow, first issued unsecured debt in 2018, Fitch assigned BB−, S&P assigned B+, and Moody’s assigned Caa1.

Third, ratings can rest on miscalculations or unforeseen changes. Agency models did not anticipate the housing collapse that pushed highly rated subprime mortgage bonds into default during the 2008 to 2009 Global Financial Crisis, and complex accounting frauds sustained strong ratings at the US firms Enron and WorldCom, and at Germany’s Wirecard AG, until each collapsed.

Corporate debt and other credit-risky instruments change hands at a yield premium, or spread, above securities treated as default-risk-free, for example US Treasury or German government bonds. Spreads, quoted in basis points, widen on issuer-specific news such as deteriorating creditworthiness (credit migration or downgrade risk) and on market-wide forces such as heightened risk aversion during stress. Credit spread risk captures the threat of a larger expected loss when credit conditions move for macroeconomic, market, or issuer-related reasons. The whole yield spread over the government benchmark bundles together the credit, liquidity, and tax premiums an investor demands.

Macroeconomic factors

The credit cycle tends to move with the business cycle. As the economy improves, spreads narrow and investors take on more credit risk; as it deteriorates, spreads widen. Put precisely, spreads reach their tightest point at or close to the top of the credit cycle, where perceived credit risk bottoms out, and their widest at or close to the bottom, where markets judge credit risk to have peaked. Under stress investors sell risky assets and buy default-risk-free ones, a flight to quality that hits high-yield hardest, and thin liquidity widens bid-ask spreads further. Systematic forces also matter: tighter regulation raises the cost for dealers of holding credit-risky inventory, funding stresses feed straight into wider spreads, heavy new-issue supply widens spreads when demand lags, and strong demand pulls them tighter.

Beyond their higher coupons, high-yield bonds attract investors for three reasons. They offer portfolio diversification, since HY often has lower correlation with IG bonds and with default-risk-free rates. They offer capital appreciation, because an economic recovery or an issuer-specific improvement (a rating upgrade, a merger, a favorable management change) lifts HY prices more than IG prices. And they can deliver equity-like returns with lower volatility, because a larger income component steadies HY returns relative to stocks; some studies find HY offers a better long-run risk-return profile than equities, which appeals to yield-seeking investors who cannot stomach full equity volatility.

Market factors

For the most liquid, lowest-default debt, such as developed-market sovereign bonds, the yield at a given maturity is essentially a real interest rate plus an expected inflation premium. A corporate yield adds a further premium for credit risk, liquidity risk, and the potential tax treatment of the specific bond, and a change in any component moves the yield, price, and return. Market liquidity risk is the cost of actually transacting: the price at which an investor can trade may differ from the indicated price, a gap shown in the bid-ask spread. Two issuer-specific traits drive it. Issuer size, meaning the aggregate value of publicly traded debt outstanding, matters because more debt usually trades more often; the less an issuer has outstanding and the less it trades, the higher the liquidity risk and the wider the bid-ask spread. Credit quality matters too, since lower quality raises liquidity risk. The most liquid government bonds can show only a fraction of a basis point separating purchase and sale prices, while less liquid corporates leave much wider gaps, and in a crisis liquidity can dry up, pushing prices down, widening spreads, and spilling stress from one market segment into others.

Issuer-specific factors

The expected financial performance of the individual issuer strongly shapes both the level and the volatility of its yields and spreads. Two factors are common to all issuers: debt coverage, the sufficiency of resources and cash flow to meet interest and principal, and leverage, the reliance on debt versus other financing. The use of proceeds and source of repayment differ by type, with corporate issuers investing in long-term assets and repaying from operating cash flow, and sovereigns running fiscal policy and repaying from tax revenue. Analysts typically judge a bond’s yield and spread against peers in the same rating category, the same sector, or with a similar business model, which helps separate market-wide moves from issuer-specific ones.

A corporate bond’s yield-to-maturity equals a government benchmark yield plus a spread, so a shift in the yield can originate in either piece. What matters is that on an option-free, fixed-rate bond, the modified duration and convexity that turn a benchmark-yield move into a price move will equally turn a spread move into a price move, whatever its source. In practice the pieces interact: a flight to quality, for example, can pull benchmark yields down at the same time credit spreads widen and liquidity thins for low-quality issuers.

Decomposing the spread

The total yield spread over the benchmark can be split into a liquidity spread and a credit spread. The idea is to price the bond at its mid-market level for the total spread, then use the bid and offer prices to isolate the liquidity portion.

Approximating the price change

For a small, instantaneous change in spread the price impact is approximately the modified duration times the spread change, with the sign reversed, so a narrower spread lifts the price and a wider spread lowers it.

For larger changes the convexity term improves the estimate.

Convexity must be scaled to match how the spread change is expressed. For an option-free bond, rescale the reported convexity to roughly the size of duration squared, and express the spread change as a decimal. For instance, a bond with duration 5.0 and reported convexity 0.235 uses a rescaled convexity of 23.5; for a 1% rise in spread the estimate is (−5.0 × 0.01) + ½ × 23.5 × (0.01)² = −0.048825, or −4.8825%.

Longer-duration corporate bonds carry higher spread sensitivity, so their prices and returns swing more for a given spread change. Investors also demand more for the greater uncertainty about an issuer’s distant creditworthiness, since factors such as strategic missteps, technological obsolescence, disasters, or leveraging events are harder to foresee the further out one looks.

The duration-only estimate is close for small spread changes, but the convexity term visibly improves the approximation as the spread change grows.

Estimating duration and convexity from prices

When duration and convexity are not given, they can be approximated by repricing the bond after raising and lowering the yield by the same small amount, ΔYield, to obtain prices PV− (yield down) and PV+ (yield up) around the initial price PV₀.