The corporate credit curve ends at ten years. The muni curve doesn't.
The observation
Investment grade option-adjusted spread this morning, by maturity bucket:
1–3 years — 4.70% / 47
3–5 years — 5.04% / 69
5–7 years — 5.25% / 82
7–10 years — 5.56% / 97
10–15 years — 5.81% / 93
15+ years — 6.24% / 101
Read it twice. The spread curve rises to 97 at seven-to-ten years, inverts to 93 at ten-to-fifteen, and finishes at 101 in fifteen-plus. Total compensation for extending from an eight-and-a-half year average maturity to a twenty-two year average maturity: four basis points.
What you are actually buying
Index yields across those same two buckets are 5.56% and 6.24%. Sixty-eight basis points of pickup. Four of them are credit. The other sixty-four are the government curve — 10s30s closed at 58.6bp, 20s30s at 0.8bp, and the long end has done all of the work.
Long corporates are not a credit position. They are a levered Treasury position wearing a credit costume, and the costume costs four basis points.
The calibration
Hold maturity constant and move down the stack instead. AAA 40, AA 58, A 66, BBB 98. One ratings band, from A to BBB, pays 32bp. Twelve additional years of spread duration pays 4.
You are compensated roughly eight times more for a single notch band than for the entire back half of the maturity spectrum. There is no framework in which that is a coherent relative price. It is also not a credit judgment — it is an artifact of who owns the long end, which is the subject of the second half of this letter.
Where it breaks
Spread duration is the risk that does not show up in the OAS print. Call the 15+ bucket thirteen years of spread duration and the 7–10 bucket six and a half. A ten basis point index widening costs the first one about 130bp of price and the second one about 65bp. Same widening, twice the damage, four basis points of upfront compensation for taking it.
The conclusion is not "avoid long corporates." It is narrower than that. If you want the term premium, buy the term premium directly — own the government curve, or own it in a market where the long end still prices credit as credit. What you should not do is pay a credit-instrument liquidity premium and a credit-instrument spread-duration profile to source a government-curve return.
Where the curve still prices credit
Four basis points is not a credit judgment. It is a sponsorship outcome. The long end of the IG market is owned by liability-matched buyers — insurance general accounts, corporate and public LDI programs — who are duration-constrained and close to spread-indifferent, and long-dated IG issuance is scarce enough that those buyers clear it regardless of where it prints. Nobody in that clearing process is expressing a credit view. The spread is a residual, not a decision.
The municipal long end is not sponsored that way. Its marginal buyer is an SMA program, a private client account, a crossover buyer — someone whose mandate is written in credit and after-tax yield rather than in liability duration. When that buyer declines a name, the deal reprices. The municipal curve slopes because somebody is being paid to hold the risk, not because somebody is required to hold the duration.
That distinction is structural and durable, and it is currently showing up in the return series.
Nearly the entire high grade universe is stretched on a short-horizon basis. Agencies sit at 2.15 standard deviations above trailing mean, one-to-three year Treasuries at 1.60, IG corporates at 1.21, MBS at 1.16. Municipals are the only duration-bearing high grade sector that is not — broad muni at 0.90, short muni at 0.26 — and muni is simultaneously the trailing-year total return leader in that cohort at +5.17%, against +2.29% for the aggregate, +1.10% for IG corporates, and −1.48% for long Treasuries.
Outperformed, and less extended for having done it. Nothing else in the cohort occupies that quadrant. The floating-rate and ultra-short sectors sit nearby on the chart, but they arrived there by owning almost no duration, which is a different statement about a different risk.
Same asset, different CUSIP
Put the two halves together, because this is the point at which four basis points stops being an academic observation.
The two largest incremental credit stories in this market — hyperscale data center and AI capex financing, and long-dated natural gas prepayment — are both being funded simultaneously in the corporate market and in municipal structure. Frequently the same underlying asset, the same operating counterparty, and the same guarantor complex, financed twice, into two investor bases, at two clearing spreads that have nothing to do with each other.
In the corporate market that risk arrives as a long-dated tenant credit or a utility bond sitting in the 15+ bucket at 101 over, cleared by a buyer sizing duration. In municipal structure the same economics arrive as a prepay or a project financing where you are explicitly compensated for guarantor credit, for the commodity and investment agreement counterparties, and for the extraordinary redemption optionality — because the buyer on the other side is underwriting precisely those things and will not take the paper otherwise.
Same asset. Two CUSIPs. One of them pays you for the analysis. The other pays you four basis points for a maturity extension and assumes you were never going to do the analysis.
The knowledge gap is the trade
Which is worth saying plainly. There is no shortage of desks that can price a corporate credit curve from memory. There is a genuine shortage of desks that can tell you who the guarantor is on a gas prepay bond, what happens to the structure when the commodity swap provider is downgraded, or what actually triggers the extraordinary redemption at par. That asymmetry does not exist because municipals are cheap. It exists because they are unread.
The corporate curve is fully arbitraged and pays four basis points for the last twelve years of it. The municipal curve is under-covered, sponsored by buyers who are making credit decisions, and still pays for work. That is why the CUSIP File exists, and the queue behind Files No. 1 and No. 2 keeps getting longer as the AI financing complex pushes more of this paper into municipal structure.
The second opinion
One item filed rather than concluded. The 10–15 year IG bucket — the thinnest, least-traded, most model-dependent point on the corporate curve, and the one that inverts — printed a single-session OAS move several multiples larger than any other bucket in the index. Every other bucket moved inside two basis points. There was no issuance event, no ratings event, no macro print.
A spread move that size, in that bucket, on a session when the bucket barely traded, is not a market clearing. It is a model updating. Which is the whole of the franchise thesis: the evaluated print and the executable print are two different numbers, and the gap is widest exactly where the observation is thinnest. We will come back to this one with the tape.
The Board
Summer Friday. The levels did not move. But the shape of the corporate curve is telling you where the compensation has already been arbitraged out, and the shape of the municipal market is telling you where it has not.
Source: Koyfin, market data as of August 14, 2026. Index-level yields and option-adjusted spreads are provider-computed composites; Treasury levels are constant-maturity. Spread-duration figures are estimates.
Produced with AI assistance. All data selection, analysis, conclusions, and final editorial judgment are the author's. All content is reviewed and approved by Positive Carry LLC prior to publication.
The Bond Bro Dispatch is published by Positive Carry LLC. All content is general market commentary provided for informational and educational purposes only and does not constitute investment advice, a recommendation, an offer, or a solicitation to buy or sell any security. Nothing herein is tailored to the circumstances of any recipient. Data are drawn from sources believed reliable; accuracy and completeness are not guaranteed. [email protected] · © 2026 Positive Carry LLC. All rights reserved.