What's going on, everybody?

I hope you all had a great week. This is Friday Five, Issue #11, and this was the week a war got worse and the market shrugged.

On Tuesday the United States completed a wave of strikes against Iran, and Iran struck back at American positions in Jordan, Kuwait, Bahrain, Iraq and the United Arab Emirates. Brent crude settled 8.87% higher on the week, at $95.63. West Texas Intermediate settled 10.68% higher. Not one barrel of that was demand. It was fear about supply, priced into the one thing supply actually constrains.

Here is what makes it interesting. Gold fell 5.04% in the same window. The volatility index finished the week at 14.32, below the 14.51 it started at, having spiked to 16.34 on the day of the strikes and then given it all back. And on Thursday, with Brent still near $96, energy was the only sector that fell while the S&P 500 rose 1.05%.

A market that genuinely feared a war would have bought gold. This one bought crude, sold the companies that burn it, and left everything else more or less alone. That is not complacency and it is not panic. It is a market making a specific judgment: that this is a supply problem for one industry rather than a crisis for all of them. Whether that judgment is right is the most important open question in this issue, and I do not think it is settled. Alright, let's get into it.

The Week in 60 Seconds

●  The United States struck Iran on September 1 and Iran hit American positions in five countries.

●  Brent settled 8.87% higher at $95.63, and none of that move was demand.

●  Gold fell 5.04% in the same window while the volatility index finished lower than it started.

●  Dell raised its full year revenue forecast by $25 billion and rose 15.81% the next session.

●  Seven of eleven sectors fell and small caps lost 1.53%, yet the S&P 500 still rose 0.22%.

This Week in One Minute

Measured from the August 27 close through Thursday, September 3, the S&P 500 rose 0.22% to 7,747.71 and the Dow rose 0.22% to 53,686.11, while the Nasdaq Composite added 0.16% and the small cap Russell 2000 fell 1.53% to 2,968.27. On the fund basis this newsletter uses for sector work, the S&P 500 ETF rose 0.27% and the equal weight version of the same index fell 0.63%. Energy led at 3.74% and Industrials finished last at negative 2.37%, with seven of the eleven sectors lower. Volatility eased from 14.51 to 14.32 after touching 16.34 on September 1. The ten year Treasury yield rose from 4.67% to 4.76%, having reached 4.80% mid week.

A note on the commodity numbers, because it matters this week more than most. Futures settle after the equity close, and when I write this on Thursday evening there is normally no settled bar for that Thursday yet. So the commodity and currency figures below are measured over an equal five session window that ends one settlement earlier, from August 26 to September 2. On that basis Brent crude rose 8.87% to $95.63 and West Texas Intermediate rose 10.68% to $91.01. Natural gas rose 4.01% to $2.96. Gold fell 5.04% to $4,366.30, silver fell 4.81% to $64.72 and copper fell 1.42%. The dollar index was close to unchanged at 99.56. Soybeans rose 3.79%, wheat 3.32% and corn 0.92%.

Three facts are worth carrying out of that.

One. Crude up 8.87% and gold down 5.04% in the same five sessions is an unusual pair. Both are supposed to rise when the world gets frightening. When only one of them does, the market is telling you it has priced a disruption to a physical supply chain, not a threat to the financial system. That distinction is the spine of this whole issue.

Two. Seven of eleven sectors fell and the index still rose. That is not a paradox, it is arithmetic. An oil shock is a revenue line for roughly 3% of the index by weight and a cost line for most of the rest, and the largest members of the index are companies whose biggest expense is engineers rather than fuel. Industrials, full of firms that buy diesel and freight, were worst at negative 2.37%.

Three. Three companies raised their own full year forecasts and all three were paid for it inside the week. Dell added $25 billion to its revenue outlook and rose 15.81% the next session. GitLab raised its profit outlook about 9% and rose 9.98%. Best Buy raised all four of its guidance lines, fell for three straight sessions, and then got bought back anyway. In August that did not happen, and I wrote about it twice. It is happening again.

The Week Ahead

On breadth. The index beat the average stock by 0.90 points this week, after 0.58 points last week. The four readings before that were plus 2.14, negative 0.69, negative 0.86 and plus 0.58, an average of plus 0.29, so this week sits 0.61 points above that average and the gap is widening rather than holding.

The gap is a size effect before it is a sector effect. The Russell 2000 fell 1.53% while the S&P 500 rose 0.22%, a 1.75 point spread that is nearly twice the equal weight gap. Underneath that, the damage was concentrated exactly where an oil move lands as a cost rather than a revenue: Industrials down 2.37%, Materials down 1.15%, both full of companies that buy fuel and freight instead of selling them.

This is a different signal from the four negative weeks that preceded it. Those were weeks when the average stock outperformed a falling index, which is broadening under stress. This is the index outperforming a falling average stock, which is narrowing under a cost shock. The mechanism is legible rather than mysterious, and that is what makes it worth watching rather than worrying about: it will unwind when the cost does.

The August employment report on September 4 is the first event in the window that reaches the median company rather than the index, and it lands about half an hour before this issue does, so I am writing without it. A weak print revives the rate cut trade, which is what small caps and the average stock need. A firm print with an oil driven inflation impulse behind it removes the cut without removing the cost, and that is the combination that keeps this market narrow. The August inflation print on September 11 is the second half of the test.

Market Risk Score

61 out of 100. Elevated. Up from 58 last week. The increase is entirely geopolitical: active missile exchanges across five countries and an 8.87% move in crude are a material change in the state of the world. What kept it from going higher is that the market's own price of fear went the other way. Volatility ended the week lower than it started, gold fell 5%, and credit sensitive parts of the market rallied on Thursday as yields eased back.

Rising real world risk alongside a falling price of fear is an unstable combination rather than a reassuring one. It means the market has taken a view, and if that view is wrong there is very little cushion in the price. That is the honest reading of 61.

Sector Rotation Update

Into: Energy 3.74%, Communication Services 1.77%, Financials 1.17%, Health Care 0.98%.
Out of: Industrials negative 2.37%, Technology negative 1.40%, Materials negative 1.15%, Real Estate negative 0.92%.

Energy has now led, trailed and led again in three consecutive weeks. That is not rotation, it is one sector being pushed around by dated headlines out of the Middle East, and I am not going to dress it up as a leadership change. I want to be specific about how little the Leading label is worth, because it is my own label: across the readings this newsletter has recorded, a sector I called Leading has gone on to beat the market in 1 of 8 following weeks, with an average edge of negative 2.13 points. A sector I called Slipping has been right 8 times out of 8, with an average edge of plus 3.63 points. I am good at spotting deterioration and bad at spotting continuation, and that asymmetry is why this week's forward call bets against the leader rather than with it.

There is one thing in the sector table that the weekly percentages hide entirely. Utilities finished down only 0.35%, which sounds like nothing. Inside it, California passed a wildfire bill on August 31 without the liability cap the industry expected, and Edison International fell 23% in a single session, its worst day in more than 25 years, with PG&E down 18%. No pick in this issue carries that exposure, but it is worth carrying the lesson: a legislative vote can reprice a regulated business faster than any earnings report, and it gives you no warning at all.

The Five

Every return below is measured from the August 27 close to the September 3 close, on split adjusted prices. All five beat both the S&P 500 and their own sector. Four of the five closed within 5% of a 52 week high, which is the honest hazard of a screen that looks for what went up, and I have said so in each write up rather than at the bottom.

Stock of the Week

Dell Technologies (DELL). One number decides this. Dell raised its full year revenue forecast by $25 billion in a single quarter, and the backlog behind it is $95 billion of AI server orders. It gets the top slot because it is the cleanest cause and effect in the issue: a dated filing on September 1, a 15.81% move the session after, and a technology sector that fell 1.40% over the same week so none of it can be attributed to the tide.

The honest caveat is the one I would want if I were reading this rather than writing it. AI servers carry thinner margins than the rest of what Dell sells. A mix that keeps shifting toward them can raise revenue enormously and leave profit roughly where it was, and revenue is the number that gets the headline. The thing to follow is not the size of the backlog. It is the margin it converts at.

What Could Change My Mind?

Two of these five picks are the same bet, and I would rather name it than let you find it. CF Industries and Deere both trace to the Iran supply shock, CF directly through nitrogen and Deere indirectly through grain and farm cash flow. Their measured daily correlation over three months is only 0.22, which is low, but a measured correlation describes the past and the driver describes this week. Forty percent of this issue turns on one geopolitical situation.

So here is the risk I am actually worried about: a credible Iran de-escalation headline unwinds the crude and nitrogen bid in a single session and takes CF Industries and Deere down together. Brent settled 8.87% higher this week entirely on the strikes and the retaliation, and none of that is demand, which means none of it needs a demand event to reverse. It has already happened once this year on words alone. After the Treasury Secretary said on August 4 that a Hormuz deal could come within days, crude fell 6.6% in a week, and that is recorded in Issue #7. CF sits at 97.1% of its 52 week high on the thinnest volume confirmation of the five; Deere sits at 98.4% of its high at 38.6 times trough earnings. Neither has much cushion.

The reading that would confirm it: Brent crude settling below $85 a barrel on or before September 18. I put the odds of that at about 25%.

Since I am naming what I avoided, here is the reasoning. The obvious trade this week was refiners, and I own none of it. Refining margins are at extraordinary levels and the refiners have risen more than 80% this year, but the futures market itself prices September crack spreads near $69.92 and August 2027 crack spreads at $44.38, more than 35% lower. When the market that trades the thing tells you it does not expect the thing to last, buying the equity on the current number is a decision you have to make deliberately rather than by momentum. Adding a refiner would also have made three of five picks the same Iran bet, and two is already more than I am comfortable with.

The thing that would worry me most is the opposite of my whole read: an oil driven inflation print on September 11 landing on a Federal Reserve that has to publish new projections on September 16. That combination removes the rate cut without removing the cost, and it would hit the four picks in this issue that buy energy rather than sell it, at the same time as the market's price of fear is sitting near a one year low with nothing built in.

Scoreboard

Here is how last week's five picks did, both in the week they appeared and in the week since. The first column is the window they were selected from, August 20 to August 27. The second is measured from the August 28 close, the morning the issue landed, to the September 3 close, over which the S&P 500 ETF rose 0.50%.

Five out of five beat the index in the week after they appeared, and the average pick rose 4.22% against an index that rose 0.50%. That is the best forward week this newsletter has had, and I want to be careful about how much weight to put on it, because one week is one week. The two names I would point at are the two nobody would have picked out of that list. Dollar General was the smallest gain in Issue #10 at 4.40% and it was the largest gain since, at 6.81%. Cleveland-Cliffs was fourth and finished second. The three big software re-ratings that led the issue did fine and did not lead.

That second column is the one worth watching, because the first is not a track record: these names are screened for having risen that week, so beating the index in their own week is the selection rule restating itself. Across 9 issues and 45 picks with a full week of history behind them, the average pick has gained 0.02% in the week after publication against a benchmark that rose 0.66% over the same windows, and 46.7% of them beat the index. That is behind, and one excellent week does not change it. I publish it next to the winners because a scoreboard that only shows the winners is not a scoreboard.

Terms in This Issue

Second-order effect. The knock on consequence of an event, one step removed from the obvious one. A war in an oil producing region raises fuel prices, which is first order. It also raises fertilizer prices, because fertilizer is made from natural gas, which is second order and usually less crowded.

Equal weight index. A version of an index that counts every company the same, so the smallest counts as much as the largest. Comparing it with the normal version shows whether a move was broad or came from a handful of giants.

Backlog. Orders a company has taken but not yet delivered. It shows what future revenue is already contracted, which is why a backlog figure can move a stock more than the quarter just reported.

Comparable sales. Sales growth from stores and channels a retailer already had a year ago, stripping out new openings and closures. It is the closest thing to a like for like measure of whether more people are actually buying.

Settlement price. The official closing price of a futures contract, set by the exchange at a fixed time. It matters because futures trade almost around the clock, so the price on a screen in the evening is not yet a close and can move before it becomes one.

Falsifier. A condition stated in advance that would prove a forecast wrong. Writing one down before the fact is what stops a forecast from being reinterpreted after it, and it is why every forward view in this newsletter carries one.

A Possible Next Step

I'm testing interest in a future Edge Report with entry zones, invalidation levels, sizing tiers, and exit frameworks. Nothing is for sale and nothing is built yet. If that is something you would actually read, tell me and I'll build it.

The Friday Five is published for educational and informational purposes only. It is not investment advice, and nothing in it is a recommendation to buy or sell any security. I am not a registered investment adviser or broker dealer. Every figure is measured from public market data and public filings on the dates stated, and past performance does not indicate future results. Do your own research and consider speaking with a licensed professional before making any investment decision.

thefridayfive.org