2Y 4.760 ▼5Y 4.860 ▼10Y 5.040 ▼30Y 5.400 ▼VIX 28.44 ▲MOVE 96.04 ▼DXY 103.82 ▼USD/JPY 159.23 ▼GOLD 3,557 ▼COPPER 5.53 ▼WTI 83.10 ▼BTC 62,082 ▼ETH 1,744 ▼S&P 7,176 ▼NDX 26,517 ▼MSTR 86.13 ▼
Rates · Liquidity · Scarcity · Credit · Power · 2026-11-19

84.7 Cents of Every Investment-Grade Dollar Is the Government's Price. The Quality Gap Is Where Credit Actually Trades.

Read the plain-English version

Every corporate bond carries 2 prices stacked on top of each other, and only one of them is about the company. The bottom layer is the risk-free rate: what the US government pays to borrow for the same number of years, which no treasurer can negotiate. The top layer is the spread - the extra yield a lender demands because the borrower might not pay - and that is the only part a treasurer can influence by running the business better. When the spread is 92 basis points on top of a 5.10% Treasury, 84.7% of the yield the treasurer pays is not about the treasurer at all. That arithmetic is usually a footnote. This month it is the whole story, because the supply of the bottom layer changed. The Treasury cut its coupon settlements by more than a third in 90 days and funded itself at the bill line instead, which is the one part of the curve that reprices in weeks rather than decades. Take duration supply away and the price of duration should fall. It rose 11.7 basis points instead. So the question this piece has to answer is not what credit is doing - it is who is now setting the price of the government's own duration, and why that price ends up on a corporate treasurer's desk.

The term premium gave back 11.7 basis points in 6 sessions, in the window the coupons shrank

The 10yr term premium is the extra yield a buyer demands for holding a 10yr bond instead of rolling short bills - the price of duration risk, as estimated by the ACM model the New York Fed publishes. It printed 0.9864% on 2026-11-13, from 0.8693% on 5 November, 11.7 basis points in 6 sessions. Of the 251 sessions in the last year, 19 closed higher, which puts it at the 92.4th percentile.

For context, the same series printed 1.0844% on 6 October. October took 21.5 basis points out of the term premium and November has put 11.7 back, almost all of it after the 5th. That is the same 6-session window in which the Treasury's 90-day coupon settlements were running -34.0% below the prior window, so the two facts are not independent: duration got scarcer and dearer at the same time.

MACROTAPE 10Y TERM PREMIUM · ACM · PERCENT TERM PREMIUM 0.5 0.7 0.9 1.1 03/26 04/26 05/26 06/26 09/26 11/26 6 OCT 1.0844 5 NOV 0.8693 13 NOV 0.9864 251 sessions. Source: Federal Reserve Bank of New York, ACM series.

One caution on the arithmetic. The 10yr par yield fell from 5.15% on 13 November to 5.07% on 18 November, 8 basis points, and the 2yr fell from 4.88% to 4.80%, also 8. 2s10s closed at 27 basis points on both dates, so the curve did not change shape. A nominal yield that falls while its term premium rises means the expected-policy component fell by more than 8 basis points. The long end is not pricing a growth scare or a policy cut. It is pricing the cost of holding the bond.

The Treasury sold 12.4 cents of duration for every dollar it borrowed, and the long end got dearer anyway

Over the 90 days to 16 November the Treasury settled 6,557.3bn of gross auctions across 82 sales: 5,742.5bn of bills, 814.8bn of coupons. Coupons were 12.4% of the total. In the prior 90 days to 17 August the same lines were 8,063.7bn gross, 6,829.4bn bills and 1,234.2bn coupons, a 15.3% coupon share.

Coupon settlements fell 419.4bn, 34.0%, while bills fell 1,086.9bn. The share of the borrowing that carries duration - anything from 2yr out - dropped 2.9 percentage points. If the term premium were a supply price, this is the tape that should have driven it down: the marginal seller of long duration walked away, and the buyers who remained should have charged less for the scarcity.

MACROTAPE GROSS SETTLEMENTS · BILLS VS COUPONS · USD BN BILLS COUPONS 90D TO 17 AUG 1,234 COUPON 15.3% 8,064 90D TO 16 NOV 815 COUPON 12.4% 6,557 Gross settlements, Treasury Fiscal Data auctions_query. Bills incl. all <=52wk.

It rose 11.7 basis points instead. That leaves 2 readings. Either demand for duration fell faster than supply did - foreign official buyers, pension de-risking, bank duration limits - or the term premium is not being set by Treasury supply at all, in which case the coupon mix is a symptom rather than a cause. This piece takes the second reading, and the rest of it is the evidence.

The mechanism: a unified government funds itself at the bill line and hands the duration decision to the private sector

Here is how the political fact reaches a price. On 3 November the Democrats took the Senate and were projected to take the House, which ended the divided Congress every piece in the midterm series had been priced against. A unified government with an open fiscal channel has a specific reason to fund short: a bill reprices in 13 or 26 weeks, so the sovereign does not lock a coupon into a policy path it is about to change. A 30yr bond is a 3-decade bet on inflation and on the political settlement that produces it, and a government that intends to legislate is the configuration least willing to make that bet.

So duration supply does not disappear - it migrates. Every unit of long-dated exposure in the system now has to be manufactured by someone who is not the sovereign: a corporate issuing a 10yr, a household taking a 30yr mortgage, a bank holding asset duration against deposits. Scarcity of the government's duration does not make the private sector's duration cheap, because they are not substitutes at the point of sale. A treasurer cannot issue Treasury bonds.

That is why the scarcity shows up as a higher price rather than a lower one. The term premium is the clearing price for the duration that no sovereign is supplying, and it is now 92.4th percentile with the supply side withdrawn. And it is why the corporate all-in is 84.7% government: the corporate curve is priced off the Treasury curve, and the Treasury curve's long end is priced off a term premium whose supply anchor just got smaller.

In investment grade, credit is 92 of 602 basis points

Investment grade OAS printed 0.92% on 17 November against a 5.10% 10yr par close the same day: 6.02% all-in. The sovereign leg is 84.7% of that number. High yield at 3.43% is 8.53% all-in, 59.8% sovereign, so the credit component is 343 basis points against 92 - a ratio of 3.7 to 1 for a default risk that no one thinks is 3.7 times higher.

This is where the AI capex argument actually lives, and it is not the argument usually made. A datacenter is underwritten on an all-in cost of capital and a depreciation schedule: 15 years of useful life against a 10yr funding cost of 6.02% for an investment-grade issuer. Of that 6.02%, 5.10 points are a judgment about the US government's duration, not about the building, the chips, or the tenant. When the term premium moves 11.7 basis points, the hurdle rate on every project in the pipeline moves with it, and no amount of demand for compute changes that arithmetic.

MACROTAPE ALL-IN 10YR COST · SOVEREIGN LEG VS CREDIT LEG · PERCENT TREASURY 5.10 CREDIT SPREAD INVESTMENT GRADE 5.10 92bp 6.02 SOVEREIGN SHARE 84.7% · CREDIT SHARE 15.3% HIGH YIELD 5.10 343bp 8.53 SOVEREIGN SHARE 59.8% · CREDIT SHARE 40.2% Par 10yr 5.10 (17 Nov) plus OAS of the same date. No curve interpolation.

Read the 2 bars together and the asymmetry is the point. A 10 basis point move in the term premium adds 10 to both all-in numbers, which is 1.7% of the investment grade cost and 1.2% of the high yield cost. A 10 basis point move in the investment grade spread adds 10 to one number only. The sovereign leg is the larger, the more volatile, and the less diversifiable of the 2.

The quality gap made a 1-year high, retraced 24 basis points, and has given back 20

The quality gap is high yield OAS minus investment grade OAS: what the market charges for the difference between a borrower likely to repay and one that might not. It printed 251 basis points on 17 November, from 231 on 12 November, 20 basis points in 3 sessions. That is the 27th largest 3-session widening in the 781 the file holds since November 2023, the 97.1th percentile - large, and 26 larger moves exist, clustered in April 2025 and August 2024.

The level is the story. A 99.2th percentile print means 261 sessions of the last 264 closed lower and exactly 2026-03-30 and 2026-11-05 closed higher: 253 on 30 March and 255 on 5 November. The median is 205. The sequence is a full round trip - a 1-year high 2 sessions after the vote, a 24 basis point retracement into 12 November, and 20 of those 24 given back in 3 sessions, 83.3%.

MACROTAPE QUALITY GAP · HY OAS MINUS IG OAS · BASIS POINTS QUALITY GAP 1Y MEDIAN 180 200 220 240 260 MEDIAN 205 03/26 04/26 05/26 06/26 07/26 09/26 10/26 11/26 5 NOV 255 12 NOV 231 17 NOV 251 264 sessions. OAS from ICE BofA via FRED; the gap is their difference.

The 2 legs did the work at very different rates. High yield widened from 3.20% to 3.43%, 23 basis points or 7.2%. Investment grade widened from 0.89% to 0.92%, 3 basis points or 3.4%. The bottom of the stack moved 7.7 times as far in absolute terms and 2.1 times as far in relative terms.

That is what makes this the political instrument and not the gauge. The gauge that ran today moved 0.588 on a VIX leg that supplied 67.5% of it - a volatility repricing, which is a statement about the price of insurance. The quality gap is a statement about which borrower the market is repricing, and it is the only line on the credit tape where the answer changed twice in 8 sessions.

What cuts against this: the insurance market turned first, and the level is not unprecedented

Two measurements argue with the read, and both are in this file. First, the quality gap is a 17 November print and the vol tape is one session ahead of it: VIX closed at 24.56 on 18 November from 25.46, minus 0.90, and MOVE at 98.15 from 101.88, minus 3.73. If the insurance market is right, cash credit should follow it down, and the next spread prints should land inside 244 - the 16 November gap - rather than above 251. A gap that holds above 244 while vol falls is a credit repricing. A gap that slides back to 231 is a flow that briefly ran.

Second, the 99.2th percentile is a 1-year statement, not a history. The full 784-session file runs back to November 2023 and its maximum is 341 basis points on 2025-04-07, the week of the tariff shock. 69 of 784 sessions have printed at or above 251. On the 3-year record this is a high zone, not an extreme, and the piece should not be sold as a crisis call. It is a claim about who is setting the price of duration, tested on an instrument that is expensive but not broken.

Third, and least comfortable: the 20 basis point give-back coincides with a broad risk move, so a share of it is beta, not politics. The defensible version of the claim is narrower than the headline - the quality gap is where the political variable would have to show up first, and it has now failed twice to hold a retracement, which is evidence about the direction of pressure rather than proof of its cause.

The instrument, and the levels at which this is wrong

The instrument is the quality gap itself - high yield OAS minus investment grade OAS - at 251 basis points, 99.2th percentile of 1 year, against a 1-year median of 205. The second instrument is the investment grade all-in at 6.02%, of which 5.10 points are the sovereign leg and 84.7% of the total. The claim is that credit is a small, slow component of a number the government sets, and that the bottom of the credit stack is where the political variable is priced.

Wrong if the quality gap closes below 231 basis points - the 12 November low of the retracement - with high yield at or below 3.20%. That combination says the 20 basis point give-back was flow and the retracement resumes, and it takes the political read with it.

Wrong on the breadth if investment grade OAS prints above 0.94% and high yield above 3.48% on the same date: both legs at 1-year highs at once is a repricing of credit, not of quality, and the 'credit is 92 of 602 basis points' framing stops being the point.

Wrong on the all-in if investment grade OAS holds at or above 0.92% for 3 consecutive prints while the 10yr par falls to 4.90% or below. That would mean credit, not duration, is the binding constraint on the corporate number, and the sovereign-share argument is describing the wrong leg.

Wrong on the supply read if the term premium falls back below 0.8693% while the coupon share stays at or under 12.4%. That makes the 11.7 basis point rise a positioning event, not the price of withdrawn duration, and the mechanism section above is then a story about nothing.

Notes

Term premium: ACM 10yr, Federal Reserve Bank of New York via FRED series THREEFYTP10, last print 2026-11-13 at 0.9864%. The series is 5 days behind the cash curve and every use of it here is dated to the print, not to today.

Credit: ICE BofA US High Yield Index option-adjusted spread (BAMLH0A0HYM2) and ICE BofA US Corporate Index option-adjusted spread (BAMLC0A0CM), both as of 17 November: 3.43% and 0.92%. The quality gap is their difference in basis points, 251. All-in numbers use the 17 November par 10yr close of 5.10%, the same date as the spreads, not the 18 November close of 5.07%.

Curve: Treasury daily par yield curve, 18 November close - 4.62 / 4.80 / 4.84 / 4.89 / 4.96 / 5.07 / 5.44 / 5.41 at 1, 2, 3, 5, 7, 10, 20 and 30 years; 2s10s 27 basis points. The par file was refetched from the Treasury's 2026 annual CSV on 19 November and merged newest-first.

Issuance: Treasury Fiscal Data auctions_query, gross settlements, 2 rolling 90-day windows to 16 November and 17 August. Bills are every tenor of 52 weeks or less; coupons are notes and bonds. These are gross settlements, not net borrowing, so they describe the mix of what the market had to absorb rather than the deficit.

Volatility: VIX and MOVE for 18 November read from the frozen snapshot data/snapshots/2026-11-19.json. The hourly latest.json refreshed at 13:00Z on 19 November republished a different VIX mark (22.31) against the same 18 November asof; the snapshot is the self-consistent file and is the one used. Not used: the 30yr mortgage series (last print 22 October) and the quarterly median sale price (last print 2026-Q3) - both are dated, both are a month or more behind the claim, and neither is load-bearing here. The housing channel was the obvious third leg of this piece and it is 4 weeks stale, so it is left out rather than dated into an argument it cannot carry.

CALL &amp; INVALIDATION
Levels2026-11-18 par: 2Y 4.80 / 3Y 4.84 / 5Y 4.89 / 7Y 4.96 / 10Y 5.07 / 20Y 5.44 / 30Y 5.41 / 2s10s 27bp
CallThe 10yr term premium is 0.9864% on 2026-11-13, 92.4th percentile of the last year and up 11.7 basis points from 0.8693% on 5 November, in the same 90 days the Treasury cut coupon settlements 34.0% to 814.8bn and a 12.4% share of gross issuance. Withdrawing duration supply raised the price of duration rather than lowering it, which is what happens when the sovereign hands the duration decision to the private sector. Investment grade costs 6.02% all-in against 5.10% on the 10yr, so 84.7% of the corporate dollar is the government's price and 92 basis points is what credit costs; high yield at 3.43% is 8.53% all-in, 59.8% sovereign. The quality gap, the only instrument where credit trades on its own, is 251 basis points at the 99.2th percentile, a 1-year high on 5 November at 255, a 24 basis point retracement to 231 on 12 November, and 20 of it back in 3 sessions.
InvalidationThe quality-gap read dies below 231 basis points with high yield at or below 3.20%: the give-back was flow and the retracement resumes. The breadth read dies if investment grade prints above 0.94% and high yield above 3.48% on the same date. The all-in read dies if investment grade OAS holds at or above 0.92% for 3 prints while the 10yr par reaches 4.90%. The supply read dies if the term premium falls below 0.8693% with the coupon share at or under 12.4%. The confirming case is a gap that holds above 244 while VIX stays under 25.46, with the term premium above 0.8693%.
EvidenceTerm premium 0.9864% (2026-11-13) from 0.8693% (5 Nov), +11.7bp, 19 of 251 sessions higher, 92.4th percentile; 1.0844% on 6 Oct. Quality gap 251bp (17 Nov) from 231 (12 Nov), +20bp, 27th of 781 3-session changes, 97.1th percentile; 255 (5 Nov), 244 (16 Nov); 1-year median 205, 2026-03-30 253 and 2026-11-05 255 the only higher prints; 3-year max 341 (2025-04-07), 69 of 784 sessions at or above 251. HY 3.20% to 3.43% (+23bp, 7.2%), 98.9th percentile; IG 0.89% to 0.92% (+3bp, 3.4%), 97.0th percentile. All-in 5.10+0.92=6.02 (84.7% sovereign) and 5.10+3.43=8.53 (59.8% sovereign). Issuance: coupons 814.8bn vs 1,234.2bn, -419.4bn (-34.0%), share 12.4% vs 15.3%; bills 5,742.5bn vs 6,829.4bn. Par 18 Nov: 4.80/4.84/4.89/4.96/5.07/5.44/5.41; 2s10s 27bp (13 Nov 27bp). VIX 24.56 from 25.46; MOVE 98.15 from 101.88 (18 Nov, frozen snapshot).
DISCLOSURE — NOT INVESTMENT ADVICE
NoticeMacroTape publishes macroeconomic analysis and market commentary for information and education only. Nothing published here is financial, investment, legal or tax advice, or a personal recommendation, and no author is a registered investment adviser. Calls, Levels and invalidation points are analytical views about how a market mechanism works, not instructions to trade. Readers act on their own responsibility and carry their own losses.