Note: I am leaving this morning for the weekend wedding of my dear friends’ daughter in Newport (oooooh…), so there will be no note tomorrow.
"Those who cannot remember the past are condemned to repeat it." — George Santayana
Morning Musings, daily, at morningmusing.com.
**THE SETUP.**Credit already broke. It broke in CCC, where the index does not look. 1,019bp against a 273bp index.**THE TELL.**CCC/HY at 3.733, the highest in three years of daily data. Exactly one observation has ever been this high.**THE BOOK.**Short private credit against JPM. Long vol into payrolls. Duration short cut, not closed.**WRONG IF.**The ratio breaks back under 3.55.Eighty-one years ago today a B-29 flew over Hiroshima. The date sits oddly next to a tape that has decided a US-Iran deal covering nuclear enrichment and sanctions is a solved problem, priced and filed. Consensus has the Hormuz corridor done and has moved to payrolls. That is the thing markets do best and worst at once: they finish with a subject.
Which brings me to a smaller act of forgetting, one that actually costs money this week.
The high yield spread in circulation this morning is 281 to 285 basis points, described everywhere as benign and in the richest decile of its history. That number is from 31 July. It has been stale for four sessions. The current print is 273.
Being twelve basis points out of date is not the problem. The problem is that the index is the wrong instrument, and the whole market is reading it anyway.
Santayana gets quoted at people who started a land war or ran the same bad trade twice. The version that costs money is duller and closer to hand. Nobody looked at the number again. The past being repeated here is four days old.
Wednesday’s tape was sold to you as a rotation. The Dow set a record close at 54,349.12, up 0.49%, while the S&P slipped 0.17% and the Nasdaq fell 0.83%, and the standard read is money moving from expensive AI hardware into healthcare and cyclicals.
Check the equal-weighted tape and that story falls apart: the average Nasdaq-listed stock fell 1.30% with two of eleven sectors green, the average NYSE-listed stock fell 0.33%, and equal-weight underperformed the cap-weighted index on the exact day breadth was supposedly leading.
It was not a rotation. It was a narrow bid inside a broad decline.
The real signal is in credit, and it is not where anyone is looking. Everyone is watching the high yield index, which is tight and getting tighter. Underneath it the CCC tier is at 1,019bp, near the widest of its year, and the dispersion between the two has never been higher in the three years of daily data I can pull. The fundamental data corroborates it from the other direction: first-lien recoveries under thirty cents have gone from roughly 7% of cases to 21%, distressed loans mark near 68 against 93 for performing paper, Ares pulled a billion-euro continuation vehicle because buyers would not accept the marks, and non-traded BDC redemption requests are running 12.1% against a 5% door.
Oil is up this morning, not down. Brent 80.54, plus 1.37%. Gold is not at a record, it is 23% below its 52-week high and below its 200-day. Silver is falling while gold rises. The miners already ran 7.39%, three and a half times spot, in a single session.
Here is the mechanism, because the conclusion is worthless without it.
A high yield index spread is a market-weighted average. When the BB and single-B tiers rally hard, the index tightens, and the index is what everybody quotes. The CCC tier is a small share of index weight and a large share of the actual default risk. So you can have the headline number printing 273bp, sixteenth percentile of its own year, sitting on top of a bottom tier at 1,019bp, ninety-eighth percentile of its own year. The average conceals the split. That is not a subtle statistical point. It is the entire early warning system, switched off.
The ratio of the two is 3.733. I pulled the full daily history back to August 2023 and there is exactly one observation at or above that level: this one. Five months of near-uninterrupted widening in that ratio, and it has not shown up on a wire all week.
The consensus read is that credit is fine and offers no cushion if something goes wrong later. My variant view is that the thing everyone is waiting for has already started; it started at the bottom of the capital structure where it always starts, and the index will be the last place it shows up rather than the first.
The recovery data your charts carry is the confirmation from the fundamental side: if first-lien lenders are taking sub-thirty-cent recoveries in a fifth of cases, up from seven percent, then the loss-given-default assumption underneath every leveraged loan mark, every BDC NAV and every private credit valuation is stale by a wide margin.
Ares could not clear a continuation fund at its own marks. That is the market telling you the marks are wrong, in the only language that counts: a failed transaction.
Now hold that next to the vol tape. VIX at 16.02, having fallen 4.2% on a down session, twenty-three percent below where it stood on 29 July, into a payroll print with a survey of 80k and three FOMC dissenters who have said publicly they want to raise rates. Kashkari says it is time to start moving up. Cook says she is prepared to act. Schmid says policy is not restrictive. That is a genuinely two-sided event, and the option market is charging nothing for it.
And the political overlay, because it is the same shape. A senator on the Judiciary Committee said this week there is nothing they could do to rein in the President, which is a description of a constraint that exists on paper and does not bind (
Dean Blundell). The market version is a kill switch nobody pulls. Everyone’s credit trigger is set at 350bp on the index. The index will not get there, because the index is not where the damage is. The trigger is decorative.I’m wrong ifthe CCC/HY ratio falls back through 3.55 with CCC inside 950bp while the index holds under 285. That would say the widening was a composition effect in the CCC bucket rather than a repricing, and the whole read collapses. I’m also wrong, more expensively, if a hot payroll tomorrow tightens everything together and the ratio compresses from the top down. Watch the ratio, not the level.
**The payroll setup is more two-sided than the vol market thinks.**Survey 80k. Challenger job cuts came in at minus 46.1% year over year with 33,429 announced, a sharp deceleration in layoffs. ISM services gave a genuinely split print: business activity 59.1, new orders 57.2, prices paid 70.3, employment 47.4 and back under fifty. Hot prices, hiring contracting. Both camps get to keep their thesis for one more day, which is precisely why the print matters.**Oil is bid, not bleeding.**Brent 80.54, up 1.37%, on an agreement that Iran and Oman have reportedly reached on a partial corridor, active two to four months, joint statement still under review and no US response yet. This is not a reopening. OPEC+ added a final 188,000 barrels for September and intends to pause. The premium unwind story that everyone traded on Tuesday is not what the tape is doing this morning.**The Fed is louder than the curve.**Three dissenters at the July hold. Effective funds around 3.63%. The curve at 2s10s plus 45bp did not move on shape at all over the last three sessions, only level, with the 2Y, 10Y and 30Y all down ten to twelve basis points from the 31 July high. Banks needed shape and got level. That distinction is the whole regional bank argument and almost nobody draws it.**Policy independence is quietly repricing.**The ‘sell America’ debate is back on the wires, tied to the chair’s communication style and reports of repeated conversations with the White House. Separately, Treasury is leaning on the Fed to expand the foreign repo facility so Japan can defend the yen without selling Treasuries. Nobody prices that until the week it matters.
**AMD minus 7.04% to 482.05, Nvidia plus 3.43% to 219.22.**SpaceX disclosed first-half capex of 28.5 billion dollars, more than four times a year ago, and confirmed it will use Nvidia silicon exclusively. Nvidia’s market capitalization is now 5.31 trillion.**So what:**this is a 10.5 percentage point single-session gap between the two largest merchant AI silicon vendors on a procurement decision, not an earnings miss. The AI trade has stopped being a sector and become a single-name concentration bet with a supplier list. That is a different risk than the one most books are sized for.**The memory tier broke on guidance, not results.**Sandisk beat on both lines, revenue 8.97 billion against 8.39 expected, adjusted EPS 39.25 against 34.37, and guided first-quarter revenue to 10.30 to 10.80 billion against an 11.16 billion consensus. Down nine percent pre-market. Western Digital beat and guided in line at the midpoint. Down fifteen. Korea took it hardest, the Kospi off 4.6% with SK Hynix printing a roughly thirty percent flash crash before recovering to close down ten.**So what:**beats are no longer sufficient, and the punishment is landing on the commodity tier of the AI supply chain rather than the compute tier. That is a cycle rolling over in memory, not an AI demand problem, and the market is currently pricing them as the same thing.**Fiserv cut full-year adjusted EPS to 7.20 to 7.40 from 8.00 to 8.30, with both lines missing.**Down ten percent.**So what:**this is the largest guidance cut in payments this season and it lands in the one financials subsector everybody treats as a defensive compounder. Payment processors are sold as toll roads on nominal spend. Toll roads do not cut guidance eleven percent at the midpoint.
The rest is for paid subscribers
FINANCIALS, FLOWS & POSITIONING, the full book with five expressions, decision tree, conviction ranking and watchlist.morningmusing.com
Disclaimer: Opinions are mine alone and may change. It does not constitute an offer to buy or sell or a solicitation of an offer to buy or sell any security, loan or asset or to participate in any trading strategy. It is not intended to form the basis of any investment decision, should not be considered a recommendation, and does not constitute an offer or solicitation with respect to the purchase or sale of any investment, nor is it a confirmation of such terms.
