What HY OAS Actually Measures

It is the number this site quotes every time it says credit has not blinked. High yield OAS is the extra yield the market demands to lend to a company that might not pay it back - and it is the cleanest thermometer credit has.

The number, and what it is made of

OAS stands for option-adjusted spread. In practice it is one subtraction: take the average yield across a basket of corporate bonds, subtract the yield on a Treasury of the same maturity, and express the difference in basis points, where 1 basis point is one hundredth of a percentage point.

The HY part says whose bonds. High yield means below investment grade - issuers rated BB and lower, the ones with weaker balance sheets and a real chance of not paying. The press calls them junk; the index calls them high yield because the yield is the compensation for the risk of owning them.

So when HY OAS prints 2.68% - its reading on 22 September 2026 - it means the average high-yield bond in the index yields 268 basis points more than the Treasury of the same maturity. Put the 5yr Treasury at 4.83% (21 September 2026) and the same bond yields roughly 7.5%. Same page, two different readers: the income buyer sees 7.5%, the risk taker is being paid 2.68% to hold the default risk.

Why this one number is the credit thermometer

Because it subtracts the risk-free rate, it isolates the part of the yield that is not about the price of money. What is left is compensation for two things: the expectation of defaults, and the price of being willing to sit with that expectation.

That is why credit watched Treasury yields tripling over the past few years and barely moving. A rising Treasury yield makes a corporate bond's price fall, but it does not make the borrower worse. The spread is the part that says whether the market thinks the borrower gets paid.

It moves for two separate reasons and the level alone does not tell you which. Expected losses rising - bad. The price of risk appetite rising, because cash is expensive and capital is scarce - also bad, and usually comes first.

Worked example, a level: 2.68% on 22 September 2026

Our series for this spread runs back to September 2023 and holds 786 daily observations. The low is 2.59% (22 January 2025) and the high is 4.61% (7 April 2025, during the tariff repricing).

A reading of 2.68% sits in the 8.7th percentile of that window. Percentiles here are of level, so a low number means a tight spread: the bondholder is charging less for the risk of a default cycle than they have charged in almost 3 years.

Read that against the rest of the board on the same day. Investment grade OAS was 0.77%, BBB 0.95%, BB 1.56% and CCC 10.75%. The rungs are the point: if credit were beginning to price trouble, CCC - the weakest balance sheets, the widest absolute spread - is where it would show first. It was tightening with everything above it.

Worked example, a change: 2.76% to 2.68% while the 10yr went up

The level mattered less than what happened around it. On 15 September 2026 the spread was 2.76%. 7 days later it was 2.68%, 8 basis points tighter.

Over the same week the 10yr Treasury went from 5.01% (16 September, par close) to a tape high of 5.11% on 23 September, with the 5yr auction clearing at 5.033% and a 3.1 basis point tail.

Read those two sentences together and you have a contradiction. The risk-free curve is repricing toward tighter money for longer, and the compensation for corporate default risk is shrinking at the same time. The spread says there is no default cycle coming; the curve says money is getting expensive anyway. Both can trade, but both cannot be right - and that gap is the trade this site keeps coming back to.

What it cannot tell you

It is a model, not a print. The option-adjusted part matters because many high-yield bonds are callable, and the index tries to strip out what the borrower's option to refinance is worth. For bonds that trade rarely, the mark is an estimate, and an index of those estimates is an average of opinions.

Composition drifts. Issuers enter and leave the index, and a company whose credit improves enough leaves high yield entirely - which mechanically tightens the index without anyone getting paid for more risk.

And the largest part of the modern credit market has no OAS at all. Private credit is held at marks a manager produces, roughly quarterly, with no daily quote and no public series. The absence of a spread is not the absence of risk - it is the absence of a measurement, and those are very different things.

The two ways to misread it

The first is treating a tight spread as a green light. A tight spread is not a verdict that borrowers are safe; it is a statement that the market is charging almost nothing for the risk. When you are paid 268 basis points over the risk-free curve to own the weakest balance sheets in the market, the reward for being right is small and the cost of being wrong is not.

The second is reading the spread without the yield. Income buyers and risk takers are looking at the same instrument and seeing different numbers - 7.5% against 2.68% - and forgetting which side of that subtraction you are on is how a safe-looking carry trade becomes an unhedged credit position.

Next: What a Percentile Hides