Credit Spreads: What the Extra Yield Is Paying For
5 steps · one page
In short
A credit spread is the difference between a bond's yield and the yield on a comparable government bond — the extra return demanded for lending to someone other than the sovereign.
It is the number credit professionals actually watch, because it isolates the part of a bond's yield that reflects the borrower from the part that reflects the general level of interest rates. The pricing article made this decomposition briefly: a bond's required yield combines the rate level for that maturity with compensation for default risk. This article takes the second component apart — what it contains, how it is quoted and measured, and how it behaves — with the standing discipline this pillar applies to every market-implied figure: a spread level is a description of what the market is currently charging, never a signal about what to do.
What a spread contains — and it is not only default risk
Spreads are quoted in basis points (hundredths of a percentage point), so a corporate bond yielding 4.4% against a 3.0% government bond of the same maturity trades at a 140 basis point spread. The naive reading is that those 140 points are pure default compensation. They are not, and the decomposition matters. Expected credit loss — the probability of default times loss given default, the arithmetic the previous articles established — is one component, and typically a smaller one than people assume. A risk premium for uncertainty is another: investors demand compensation not merely for expected losses but for the variance around them, and for the fact that credit losses cluster in bad times, which makes them worse than an equivalently sized independent risk. A liquidity premium is a third — corporate bonds trade less readily than government benchmarks, and holders charge for that. Compensation for embedded options is a fourth where relevant, since a callable bond's quoted spread partly reflects the option the holder has written. And supply, demand, and technical factors supply the residual: fund flows into and out of credit, dealer balance-sheet capacity, index rebalancing, and forced selling can all move spreads without any change in fundamentals. The persistent empirical observation that spreads exceed what historical default and recovery experience alone would justify is known as the credit spread puzzle, and it remains an active area of academic debate — with explanations ranging from the correlation and liquidity premia above to the possibility that historical default samples understate tail risk. This portal reports the puzzle rather than resolving it; the practical takeaway is simply that a spread is a bundle, and reading it as a pure default estimate misreads it.
How spreads are measured: the four common quotations
Four measures appear in practice, and knowing which one a screen shows matters as much as it did for yield measures. Nominal spread is the crude version: the bond's yield minus the yield of a single chosen government bond of similar maturity. Simple, and imprecise, because the two bonds' cash-flow timings differ. Spread to the curve (interpolated, sometimes called I-spread) improves on this by comparing against a point on the government yield curve matched to the bond's maturity rather than to whichever benchmark happens to be nearby. Z-spread is the more rigorous construction: the constant number of basis points that, added to every point on the government spot curve, makes the discounted value of the bond's cash flows equal its market price — a whole-cash-flow measure rather than a single-point comparison. And option-adjusted spread (OAS) takes the Z-spread and strips out the value of embedded options, which is the only sound way to compare a callable bond with a non-callable one: without that adjustment, a callable bond looks cheap when it is merely optioned. For high-yield bonds and any bond with a call feature, OAS is the professional standard and a nominal spread is close to useless. Two conventions complete the picture: spreads may be quoted against government curves or against swap curves depending on market and instrument, and the reference matters because the two are not identical; and credit default swap spreads represent the cost of insuring against a specific issuer's default, which is a related but distinct market whose quotes can diverge from cash-bond spreads — a divergence professionals watch and this portal simply notes exists.
How spreads behave — and why that is not a trading signal
Three behavioural regularities, each stated as observation. Spreads widen in stress and compress in calm. When economic conditions deteriorate or risk appetite falls, investors demand more compensation for credit risk and spreads widen — often sharply, and often across the whole market at once rather than issuer by issuer. In benign conditions they compress. This is why aggregate spread indices are watched as barometers of financial conditions, and it is why credit is described as a risk asset despite being debt. Widening hurts existing holders twice over. A spread that widens lowers a bond's price (higher required yield, per the pricing arithmetic) — and it typically widens when the holder's other risk assets are also falling and when the bond has become harder to sell. The three problems arrive together, which is the correlation point the high-yield article emphasised, appearing here in price terms. And spreads move for two distinguishable reasons that a reader should try to separate: issuer-specific deterioration (this borrower is in more trouble) versus market-wide repricing (all credit is being charged more). A bond whose spread widened 80 basis points while the whole market widened 80 basis points has told you nothing about its issuer. Now the discipline. Spread levels are routinely presented in commentary as buy or sell signals — "spreads are historically tight, therefore credit is expensive" — and this portal does not endorse that inference. Historically tight spreads have persisted for years; historically wide spreads have widened further. The mean-reversion intuition is intuitive, the evidence for reliably trading it is contested, and the position taken here is the same one taken for breakeven inflation and every other market-implied figure in this portal: it tells you what the market is charging today, which is information, not instruction. Anyone considering acting on a spread level is forming a view on credit cycles and should do so with a licensed adviser.
Worked example
Worked example (fictional). Two dates, one bond. Fictional Cadera Power's ten-year note yields 4.4% when the comparable Republic of Meridia government bond yields 3.0% — a spread of 140 basis points. Six months later Cadera's bond yields 5.6%. Has Cadera's credit deteriorated? It depends entirely on the government yield, and there are three quite different possibilities. Case one: the government bond now yields 4.2%, so the spread is 140 basis points — unchanged; Cadera's yield rose purely because rates rose, and the market's view of Cadera is identical. Case two: the government bond still yields 3.0%, so the spread has widened to 260 basis points — this is a genuine credit event, the market charging Cadera substantially more. Case three: the government bond yields 3.0% and every comparable corporate spread has also widened by 120 points — Cadera's spread widened, but market-wide, so this is a repricing of credit generally rather than a judgment about Cadera. Same 120 basis points of yield increase, three entirely different meanings — which is precisely why credit is watched in spread terms rather than yield terms. And note the holder's position in cases two and three: the price fell, the bond became harder to sell, and other risk assets were probably falling at the same time. (All names and figures fictional.)
Frequently asked
6 questions
What is a credit spread?
The difference between a bond's yield and that of a comparable government bond, quoted in basis points — the extra return demanded for lending to a non-sovereign borrower. It isolates the borrower-specific part of a yield from the general interest-rate level, which is why credit is analysed in spreads rather than absolute yields.
Is the spread just compensation for default risk?
No — that's a component, and often a smaller one than expected. A spread also contains a premium for uncertainty (credit losses cluster in bad times, making them worse than independent risks), a liquidity premium, compensation for any embedded options, and supply-and-demand effects from fund flows and dealer capacity. The persistent gap between spreads and historical default experience is known as the credit spread puzzle and remains debated.
What's the difference between Z-spread and option-adjusted spread?
Z-spread is the constant addition to every point on the government spot curve that makes the bond's discounted cash flows equal its price — a whole-cash-flow measure. OAS takes that and removes the value of embedded options, which is the only sound basis for comparing a callable bond with a non-callable one. Without the adjustment, a callable bond looks cheap when it's merely optioned.
What does it mean when spreads "widen"?
Investors are demanding more compensation for credit risk, so corporate yields rise relative to government yields and corporate bond prices fall. It typically happens across the market at once in deteriorating conditions — and it hurts holders on three fronts simultaneously: prices fall, the bonds become harder to sell, and other risk assets are usually falling too.
If a bond's yield rises, has its credit worsened?
Not necessarily — you can't tell without the government yield. The yield may have risen purely because rates rose across the market, with the spread unchanged and the market's view of the borrower identical. That's exactly why the spread, not the yield, is the credit measure.
Are tight spreads a sign to sell credit?
This portal doesn't offer that inference. Historically tight spreads have persisted for years, and historically wide ones have widened further — the mean-reversion intuition is appealing but the evidence for reliably trading it is contested. A spread level tells you what the market is charging today. That's information, not instruction, and acting on it means forming a view on credit cycles, which belongs with a licensed adviser.
References
Educational and informational only — not investment advice, a recommendation, or an offer to buy or sell any security. Investing involves risk, including the possible loss of principal. Worked examples use fictional companies and figures.