A quiet-looking tape with a loud afternoon underneath it. The indices lost half a percent or so, except the Dow, which lost more than twice that, and nobody will remember either number. Oil ran at $100 while the US Navy and the IRGC traded shots over tankers, and that they will.
Brent Went To 98.85
The story today wasn’t the stock market. It was a tanker war.
Brent finished at 98.85, up 1.7%, which puts it $1.15 from a hundred dollars. WTI closed 93.76, up 1.3%. At lunchtime Brent was 97.30 and up a tenth of a percent on the day; almost all of the move happened in the last three hours, and the wire tells you why.
Per First Squawk, in the hour before the bell:
“BRENT CRUDE SURGES TOWARD $100/BBL, HITTING A SIX-WEEK HIGH, WHILE US GASOLINE PRICES SET A LABOR DAY RECORD AT $4.15/GALLON; BANKS WARN PROLONGED CONFLICT COULD PUSH OIL TO $120–$150/BBL.”
“OIL APPROACHES $100 AS NEW HOUTHI ATTACKS ON SAUDI ENERGY INFRASTRUCTURE AND EXPLOSIONS ON IRAN’S KHARG ISLAND HEIGHTEN GLOBAL SUPPLY FEARS.”
“BRENT RISES 1% TO $97.92 AND WTI JUMPS 1.7% TO $93.03 AS HORMUZ FLOWS REMAIN AROUND HALF PRE-WAR LEVELS, RAISING FRESH INFLATION AND REFINED-FUEL SUPPLY RISKS.”
That last one is a snapshot from around a quarter to four, not a closing print. Brent kept going after it and finished at 98.85, another 93 cents higher. We’re quoting the wire as it ran and we’re not going to reconcile its intraday percentages against our closing ones; they’re measured from different points in the day.
On what actually happened out there, again per First Squawk:
“US FORCES ARE REPORTEDLY STRIKING IRANIAN OIL TANKERS, WITH TWO US OFFICIALS CITED AS SAYING AMERICAN DRONES TARGETED TANKERS OFF IRAN’S SOUTHERN COAST: I24 NEWS.”
“IRAN LAUNCHED A SECOND, PREVIOUSLY UNDISCLOSED ATTACK ON US NAVY SHIPS MONDAY, ACCORDING TO US OFFICIALS CITED BY WSJ; NO AMERICAN VESSELS WERE HIT, BUT THE ATTACKS RAISE CONCERNS OVER IRAN’S GROWING MISSILE CAPABILITIES AND POSSIBLE CHINESE OR RUSSIAN ASSISTANCE.”
And the newest one, which landed after the close and is the one that matters into tomorrow morning:
“IRGC NAVY WARNS OIL TANKER CREWS NEAR KUWAITI & BAHRAINI PORTS HOSTING US FORCES TO EVACUATE, THREATENING TO TARGET VESSELS IN RETALIATION FOR US STRIKES ON IRANIAN OIL TANKERS”
Read that slowly. Kuwait and Bahrain are not Iranian waters and those are not military targets. Telling commercial crews to get off their ships is a threat to shipping insurance and shipping routes, and the freight market prices that faster than the oil market does.
Friday’s letter ran a story headlined “Oil Is Back Above $95…” and quoted the CENTCOM wire about three Iranian crude carriers being struck. That was Friday’s close at 96.03. This is the very next session, because Monday was Labor Day, and Brent is $2.82 higher. That story has not finished.
The rest of the table: gold fell 1.1% to 4,357.60, which is the more interesting of the two metals-and-oil facts, because gold usually likes a shooting war. Natural gas at −2.6% is still the largest percentage move on this table in absolute terms, and it is not remotely the story today.
The Call That Was Right At Lunchtime
Most of the time you read someone’s market post, you nod, and you never find out whether they were right. Today we found out twice. Once at lunchtime, when he was right. Once at the bell, when he was half right.
Around 12:20, amit, who posts as @amitisinvesting, put this up. The page cut him off mid-sentence, so this is exactly where the quote ends:
“software is down because semis are gapping up heavily today / this has become the new normal, so it shouldnt surprise anyone / however, id much rather deal with a software dump from 52-week highs vs making new lows almost every week a few months ago when the semis would pump”
At 12:50 that is precisely what the board showed. Intel was up 10.33%, AMD 6.83%, Marvell 2.48%, Sandisk 2.47%, Broadcom 2.38%, and the big software and platform names were red. Then the afternoon happened.
Midday and the close, September 8.
Three of them held their move and three gave most or all of it back. Intel gave back a point and a bit and still finished up 9.03%, which is an enormous day. AMD gave back about a point. Broadcom was the only one that got better, from +2.38% to +2.97%. But Marvell fell from +2.48% to +0.81%, Sandisk went from +2.47% to −0.11%, and Micron went from barely green to −1.62%. Two of the seven closed red that were green at lunch. And NVIDIA, the largest chip name there is, was never in the trade: red at midday, red at the close, −2.01%.
The software half of his sentence aged better than the semiconductor half. CRCL closed −5.74%, worse than it was at lunchtime. Netflix −1.85%. Microsoft −1.15%. Software kept falling all afternoon; the semis stopped leading around one o’clock.
We’re showing you both prints rather than the tidy version. The tidy version would have you believe a rotation ran all day. It didn’t. It ran for a morning and then gave a chunk of itself back, and the difference between those two descriptions is the difference between a story and what happened.
September 8 close. Chip and chip-adjacent names in bold. The 52-week low and high columns are dropped here for width. CBRS has no year-to-date figure, so that cell says “not quoted.” It is not zero.
Four of the five best names on the day are chips. The fifth is Tesla at +3.98%, which has nothing to do with any of this. At the other end, three of the weakest are CRCL at −5.74%, CBRS at −4.87% and NVIDIA at −2.01%, with IBIT and Netflix both at −1.85% just behind them.
The year-to-date column is the one that sticks, and it barely moved on a day like this. Sandisk is up 632.22% on the year and closed down eleven basis points. Micron is up 250.41% and closed down 1.62%. Intel is up 183.06%. We don’t own those, we don’t recommend them, and they aren’t on any list we track. They’re there to tell you the scale this rotation has run to, and to make the point that one afternoon barely dents it.
The Scorecard
Everything sold off further into the bell. The Dow finished down 1.2% against the S&P’s 0.5%, more than twice the decline, and it was the standout on the downside all day.
The VIX closed 15.63, up 2.2%. Volatility rising on a down day is the most ordinary thing on this page and we’re not going to dress it up as a signal. One honest note: Friday’s close was 14.58 in our weekend letter, and the change column here doesn’t reconcile against that base. So take the 15.63 and the direction, and we’ll leave the subtraction alone rather than publish arithmetic that doesn’t work.
The Sector Board
September 8 close.
Three green, eight red. Energy took the lead, which after the section above is not a surprise, and real estate slipped from green to red in the afternoon. Health care was the worst sector all day and got worse into the close at −2.5%, with financials next at −1.4% after deteriorating through the afternoon. Top to bottom the board spans about three and a half points.
Now the inference, labelled as one, because it is not in the data. A 2.5% health care selloff is the obvious candidate for why the Dow fell more than twice as much as the S&P. The Dow is price-weighted and it carries big-ticket health care names. We cannot prove it here, because what we have is sector ETFs, not Dow component moves, so we don’t know which individual stocks did the damage. Treat it as the likely explanation, not as something we checked.
Nothing Happened In Credit
The absence is the story, and it held all day. High yield finished down three cents. Investment grade down two. Treasuries, TIPS and municipals all within a couple of pennies of unchanged. Nothing on that table moved as much as a tenth of a percent except the convertibles.
Which are the odd row again. CWB up 1.1% on a day stocks were lower is a strange print, because convertibles are equity-linked and usually go where stocks go. It was up 1.4% at midday and held most of it. We don’t have the holdings behind that ETF here, so we can’t tell you whether it’s one large position or something broader. We’re flagging it rather than explaining it.
If today were the front edge of something, credit is where the first flinch usually shows up. There wasn’t one, on a day with drone strikes on tankers in it.
The Price Of Money
Friday’s letter argued that the market’s problem isn’t growth, it’s the price of money. Here is the version of that argument that should bother you: the thirty-year did not move at all.
The US thirty-year closed at 5.249%. At lunchtime it was 5.249%. Identical to the thousandth, through an afternoon in which oil went to 98.85 and the IRGC threatened commercial shipping. The front end nudged up, the one-year from 4.131% to 4.142%. The ten-year slipped a hair to 4.795%. The UK thirty-year went from 5.801% to 5.823%. That’s the whole day in the world’s most important bond.
The rest of the shape is unchanged. The US pays 142 basis points more than Germany to borrow for thirty years, and 128 more than Japan. Britain pays another 57 on top of the US. Of the six thirty-years quoted here, the two Anglo economies are the two most expensive, and it isn’t a photo finish. The US curve slopes up the whole way, 4.142% at a year to 5.249% at thirty, and 45 of those 111 basis points sit in the last twenty years of it.
Liz Ann Sonders, Chief Investment Strategist at the Schwab Center for Financial Research, made the same point today alongside a chart. Her post was cut off by the page. Quoting to the cut:
“As U.S. gov’t bond yields have risen, spread between equity earnings yields and bond yields has compressed; not a phenomenon unique to U.S.; spread may be tighter in U.S. than in other markets, but same pattern can be observed across developed markets (selection of which shown”
It ends there, at “shown,” so that’s where we stop. The readable half is the useful half. When the earnings yield on stocks and the yield on bonds converge, what you’re paid for owning the riskier of the two shrinks. Her second clause is the one Friday’s letter didn’t make: this isn’t only happening here. The UK ten-year at 5.171% sits 38 basis points above the US ten-year, and Italy’s thirty-year is 110 above Germany’s.
One more thing on the demand side, per First Squawk about half an hour before the close:
“FOREIGN INVESTORS INCREASE TREASURY BUYING, PURCHASING $9.84B OF 2-YEAR, $8.75B OF 5-YEAR & $6.79B OF 7-YEAR NOTES, UP FROM PRIOR MONTH LEVELS.”
Two, five and seven years. Foreign buying up month over month at the front and the belly, and a thirty-year that would not budge off 5.249% on a day the Gulf caught fire. We’re putting those two facts next to each other because they’re both true and they’re both interesting. We are not telling you one caused the other, and we haven’t seen anything today that would let us.
Two Voices, Opposite Readings
This is the part to sit with, and it isn’t our argument. It’s between two people we read every day.
Liz Ann Sonders, this morning, complete and not truncated:
“August @NewYorkFed median 1-year inflation expectations +3.58% vs. 3.63% prior … 3-year +3.19% vs. 3.26%”
Both tenors fell. The one-year is down five basis points, the three-year down seven. Small moves, and nobody should build a thesis on five basis points of survey data. But the direction is down.
Keith McCullough, CEO of Hedgeye Risk Management, in the last hour of trading, also complete:
“The Global #Quad3 commodity trade keeps paying / WTI: +9.7% last week | +23.0% 1M / Copper pushing to ATHs / Nat Gas: +10.9% 1M / Long OIH, ICOP & UNG. Inflation isn’t slowing down- it’s broadening”
Those return figures are his and we’re passing them on as his, not checking them against our table. But the last five words are the whole disagreement. One of them is reading a survey of what people expect. The other is reading what things actually cost, with Brent closing at 98.85 in between them.
Both can be right for a while, which is the uncomfortable part. Expectations can drift down while energy drags a headline number up, and that gap is exactly where a central bank gets into trouble. It is also, for what it’s worth, the thing the thirty-year seems to be refusing to take a view on.
We’re not adjudicating this one. We don’t have the primary data to, and the referee arrives this week anyway: PPI Thursday, CPI Friday, then the Fed on the 15th and 16th. If Thursday and Friday come in soft, Sonders’ survey looks right and the long end has some explaining to do. If they run hot with oil where it is, McCullough’s broadening looks right and 5.249% was the honest number all along.
Around The World
Fourteen markets and the same two ends as at lunchtime. China’s FXI was the weakest at −2.5%, with India next at −1.5%. Brazil was the strongest at +2.0% and got stronger into the close. Korea second at +0.5%.
Europe was a shrug that turned slightly sour: France, South Africa and Mexico all gave up morning gains, and the developed blend finished down half a percent. Worth noting that China had the worst tape of the fourteen while the Chinese thirty-year sits at 2.161%, comfortably the cheapest money on the sovereign table. The country with the lowest cost of capital in the group also had the worst equity day. One session is not a thesis and we won’t turn it into one.
The yen is the line to watch, and it firmed into the bell: 153.9830, with the dollar down two tenths of a percent against it. The last figure we put in print for it was 155.6350, in Friday’s Market Intelligence, and that one was a mid-session read rather than a close, so the two aren’t quite like for like.
Bitcoin closed 78,511, down 0.9%, and IBIT finished −1.85% on the day and −10.59% on the year. It is one of only five names on the board above that are red year to date, out of the 21 carrying a year-to-date figure.
Trusted Voices
Only what was actually posted today, quoted as it ran. Where the page cut a post off, we quote to the cut and stop.
Liz Ann Sonders, Chief Investment Strategist at the Schwab Center for Financial Research. Both of her posts are above. The New York Fed expectations print is complete; the earnings-yield-versus-bond-yield post was truncated and we quoted it to the cut.
Keith McCullough, CEO of Hedgeye Risk Management. His commodity post is above and it’s the one that matters. Three more, all complete. On the rates trade: “Long $CLOX vs. Short $TLT = Macro Alpha” — for what it’s worth the short TLT leg neither paid nor cost anything today, TLT closing dead flat at −0.01%. We have no price for CLOX, so that’s half a trade we can see. On gold, before it fell 1.1%: “Nice job reducing your Gold positions on green last week”. And on the copper leg of the same commodity thesis: “Nice rip for our Copper Miners $ICOP ETF today”. Earlier in the day he also posted his signal count, “Signal Strength Stocks (update) = 61 Longs, 76 Shorts #timestamped” — 137 names carrying a signal, 15 more short than long, which on a day all four indices closed red is a tilt the tape agreed with. And one line worth keeping on the wall: “You want your process to be automatic”.
amit, who posts as @amitisinvesting. We don’t have a firm or a title for him, so we won’t invent one. His semis call and how it aged are the second section of this letter. He also posted twice on a Palantir and Nebius announcement, both cut off by the page: “$PLTR $NBIS PALANTIR NAMES NEBIUS AS ITS PREFERRED SOVEREIGN AI INFRASTRUCTURE PARTNER. - Palantir and Nebius announced a strategic AI infrastructure partnership, with Palantir naming Nebius its preferred sovereign AI infrastructure partner and planning to integrate Nebius”, and then, on going and doing the homework himself: “Called up Palantir Architect @chadwahl this morning to get his thoughts on this Palantir/Nebius Partnership and he confirmed what most people assumed: distributed compute across open source models is a sector of AI demand that is only growing. The future is being built in front”. Neither name is on the board above, so unlike the semis call we have no way to check this one. It’s his reporting and we’re passing it on as his.
Bottom Line
We did trade this morning. That’s the combo alert’s job, not this letter’s.
Four things from today, in the order we’d rank them.
Oil is the story and it is not finished. Brent closed at 98.85, $1.15 from a hundred, on drone strikes against tankers, explosions on Kharg Island, Houthi attacks on Saudi energy infrastructure and Hormuz running at about half its pre-war flows. The IRGC then told commercial crews near Kuwaiti and Bahraini ports to get off their ships. Whatever tomorrow does, the risk in that sentence is not priced by a 15.63 VIX.
The rotation faded and we said so. Intel, AMD and Broadcom held; Marvell, Sandisk and Micron gave it back or went red; NVIDIA was never in it. A call that’s right at lunchtime and half right at the bell is normal, and the only thing that would have been wrong is printing the lunchtime numbers under a story about the whole day.
Nothing broke. Outside the convertibles, nothing in credit moved a tenth of a percent, on a day with a shooting war in it. That is either reassurance or complacency and we genuinely can’t tell you which.
And the one we’d turn over tonight: the thirty-year closed exactly where it sat at lunch, 5.249%, while two people we trust read the same afternoon in opposite directions. Sonders has expectations falling on both tenors. McCullough has inflation broadening, with the oil price to back him. PPI Thursday, CPI Friday, the Fed the week after. One of those two readings is about to get a lot harder to hold.
Disclaimer
This is not investment advice. Nothing in this letter is a recommendation to buy or sell any security. Grow Your Pile and Squared T Capital publish what we do in our own accounts for education. Options involve substantial risk and are not suitable for every investor. A short put can require you to buy stock at the strike price and can lose more than the premium collected. Past results do not predict future results. Do your own work and consider speaking with a licensed advisor about your circumstances.
Past performance is not indicative of future results, and results shown are those of our own accounts and are not representative of any subscriber’s results. Read the Characteristics and Risks of Standardized Options before trading.
Tony Rihan and Tony Battista Grow Your Pile
Every options trade in all three portfolios is published, win or lose, with entry, exit and running P&L at growyourpile.com.