Oil Shock And Rates Test Thursday Stock Market Open

I checked the futures before the coffee finished brewing, and the screen already looked like an argument. Crude was up more than 4.5 percent on reports that Washington is weighing large-scale military operations involving Iran. Bond yields were climbing with it. Equity futures pointed to a sharp gap lower, and the South Korean Kospi had already dropped more than 2.5 percent overnight. None of that, on its own, tells you whether data-center spending is peaking. It does tell you that Thursday’s stock market open will be priced by people who would rather blame oil and rates than sit with a messier set of company numbers.

That habit is old. When a narrative gets crowded, any macro jolt becomes a convenient exit ramp. I’ve found that the cleaner the excuse, the more carefully you should read the actual prints. Chipmakers in Asia just delivered numbers that were, frankly, hard to spin as a peak. Consumer staples beat the quarter and then cut the year. A freight name lost a chunk of its target price. Banks got fresh buy notes. The tape is not one story. It is several stories arguing in the same hour.

What the Open Is Really Pricing

A lower open after a regional index slide is not a thesis. It is a mood. The mood this morning leans defensive because energy is spiking and the long end of the Treasury curve is being asked to absorb fresh supply. Traders will watch this afternoon’s 30-year auction after yesterday’s strong 10-year sale. Any market that lives and dies by bond actions is, in my view, less attractive than one that can ignore the auction desk for a few hours. That does not mean you fade every yield tick. It means you separate the excuse from the evidence.

Perhaps the most interesting aspect of this session is how quickly the data-center peak story got dusted off. Higher oil. Higher rates. Therefore spending on compute must be rolling over. That leap skips the overnight semiconductor results. Taiwan’s foundry giant and South Korea’s memory leader both put up numbers that look like demand, not a last hurrah. Memory pricing, advanced-node utilization, and customer commentary still read as tight. You can dislike the valuation of anything tied to artificial intelligence and still admit the order book has not rolled over in the data that just hit the tape.

A macro jolt is a reason to reprice risk. It is not, by itself, proof that a spending cycle has peaked.

Desk note, pre-open

Kospi’s drop of more than 2.5 percent will get cited all morning as confirmation. Fair enough as a sentiment read. Less fair as a fundamental read. Index moves in Seoul often exaggerate foreign flow and currency noise. When the underlying chip reports are strong and the index still sells off, you are watching positioning, not a sudden collapse in end demand. U.S. futures followed because they usually do when Asia tech wobbles and oil rips at the same time. Correlation is not causation, but it is how the open gets set.

Oil, Iran Headlines, and the Bond Desk

Oil up more than 4.5 percent is not a rounding error. It is a risk premium. Reports that the United States is considering large-scale military operations tied to Iran are enough to reprice the front of the crude curve before anyone knows whether those operations happen, how wide they would be, or how shipping lanes would respond. Markets do not wait for confirmation. They price the option.

That option has a second leg. Higher energy feeds inflation anxiety, and inflation anxiety feeds the long bond. Yields are higher into the 30-year auction. Yesterday’s 10-year sale went well, which is why today’s longer maturity matters. A sloppy tail, weak indirect demand, or a concession that fails to clear would give equity bears a second headline before the close. A clean auction would steal that headline back. I’ve sat through enough of these to know the equity reaction is often larger than the yield move itself. Traders hate ambiguity more than they hate a known higher yield.

• Crude’s jump is a geopolitical premium, not yet a confirmed supply outage.
• The 30-year auction is the afternoon’s real macro print.
• A market that needs a friendly bond sale is a market with less room for multiple expansion.
• Energy producers get the direct bid. Transport, chemicals, and discretionary names wear the cost.

Any market that depends on bond actions is less attractive than one that does not care about them. That line is worth keeping. It is not a moral claim. It is a duration claim. Companies with long-dated cash flows, high multiples, and no near-term pricing power feel every basis point. Companies that bill in the next two quarters and pass costs through feel less. Thursday sorts names along that line whether the headlines admit it or not.

The Data-Center Peak Excuse

Let me be plain. Using oil and rates to declare that data-center spending has peaked is convenient, and convenience is usually a tell. The spending cycle is enormous, concentrated, and already priced for a lot of perfection. Skepticism is healthy. Declaring a peak because crude rallied on a geopolitical headline is not the same thing as skepticism. It is a shortcut.

What would a real peak look like? Order pushouts at the foundry. A break in memory contract prices. Customers talking about digestion rather than allocation. Power-purchase delays that actually cancel racks rather than slip them a quarter. None of that is what the overnight chip prints described. They described utilization and pricing that still favor the sellers. You can argue the stocks have run too far. That is a valuation argument. It is different from an end-demand argument, and mixing them is how people get chopped up on the open.

In my experience, the peak call arrives in waves. First the multiple compresses on rates. Then a single soft customer comment gets amplified. Then a regional index drop is treated as global proof. By the time the fundamental data catch up, the call is either early and painful or late and obvious. This morning looks like wave one wearing the costume of wave three.

PepsiCo: A Beat That Still Cut the Year

The beverage and snack giant beat on quarterly earnings and revenue. Organic revenue growth came in ahead of what the Street had penciled in. Management wants to build on a strong international business while acting with urgency in North America. That phrase, acting with urgency, is corporate for the home market is not where we want it. Volume, mix, and promo intensity in the U.S. still need work. A beat with that caveat is a mixed story, and mixed is the right word.

The part that will stick is the guidance. Full-year core earnings-per-share expectations were cut. You can beat the quarter and still tell holders the year is harder. That combination usually produces a shrug rather than a celebration, especially when rates are rising and the stock is owned in part for the dividend. Higher rates make dividend stocks less attractive on a relative basis. The coupon on the long bond is competition. When that coupon rises on the same morning the company trims its profit path, income buyers do not have to panic. They do have to re-rank.

I like the international cushion. I am less impressed by a North America plan that still reads as a repair job. Snacks and soft drinks are not broken franchises. They are franchises arguing with a tired consumer and with retailers who have learned to push private label. Urgency is the correct posture. Urgency is not a number. The cut to core earnings per share is the number.

Dividend stocks do not become bad businesses because the 30-year yield pops. They become less special. The spread you are paid to own a slow-growth staples name shrinks when Treasuries pay more for doing nothing. That is arithmetic, not a verdict on the brand. If you own it for the coupon and the defensive beta, today’s setup asks you to be honest about why it is in the portfolio. If you own it because you think North America volumes inflect next year, the guidance cut is the item to underwrite, not the oil headline.

Palantir Gets a Buy and a Moat Argument

A major investment bank upgraded the data-analytics company to buy from neutral and put a $230 price target on it. The note leans on a moat built through deep government ties, enterprise trust, and a habit of putting engineers on site with the customer. That last point is easy to skip and hard to copy. Software that requires the vendor’s people in the building is stickier than software that arrives as a login. It is also more expensive to deliver. The bull case says the expense is the moat. The bear case says the expense caps the margin story just as the multiple assumes software economics.

Shares have staged a large comeback since a late-July low. Upgrades into a comeback are not rare. They are also not free. The bank is arguing that government credibility has translated into commercial trust, and that on-premise engineering is a feature rather than a services drag. I can see the logic. I can also see why a stock that has already repaired a lot of damage does not need another narrative tailwind to stay expensive. A buy rating with a $230 target is a statement about durability. Durability in this name has always been the debate, not the demo.

Does the upgrade change the open? Only at the margin. On a morning when futures are heavy and oil is the headline, single-stock research fights for airtime. It matters more for holders deciding whether the July low was the clearing event. If you believe the government franchise and the on-site model compound, the note is cover. If you believe the multiple already prices that compounding, the note is late. Both can be true for different time horizons. That is the annoying part of sell-side work, and also the useful part.

An Australia AI Infrastructure Listing Looks Shaky

The proposed listing of Firmus Grid, an Australia-based AI infrastructure company backed by the leading accelerated-computing firm, is not finding a clean path to price. Reporting around the bookbuild says the process of settling an offer price closed on schedule, with no clear indication of structure. That is a polite way of saying the market did not hand the sellers the number they wanted, and nobody has agreed what the deal even looks like yet.

Three problems stand out, and none of them require a conspiracy. Placement looks poor. The operating track record is not long enough to carry a premium multiple without a leap of faith. The valuation being discussed has been high relative to what buyers will underwrite in a risk-off tape. Add a structural flaw that should have been fixed before the roadshow: more than half of the shareholders are free to sell on the first day. That is not a lockup. That is an invitation for the stock to meet supply before it meets a story.

A first-day free float above half the register is not price discovery. It is a supply event wearing a ticker.

I do not see how that structure makes sense for anyone who has to hold past the opening print. New issues tied to artificial-intelligence infrastructure already carry narrative risk. Layer on an unproven record, a stretched valuation, and a register that can sell immediately, and you have a deal that asks public buyers to be the exit. Some deals deserve that role. This one, on the facts available, looks like it was sequenced for the sellers. When the book closes with no structure indicated, the market has already answered.

There is a broader lesson for anyone chasing AI infrastructure paper. Backing from a famous chip designer is a distribution advantage, not a cash-flow guarantee. Power, sites, customers, and contract length are the assets. A logo on the cap table is marketing. In a week when oil and rates are already pressuring long-duration stories, marketing is a thin bid.

Constellation Brands Loses a Buy, Not the Argument

A global bank cut Constellation Brands to hold from buy. The worry is the quarter reported Tuesday night, specifically how Modelo and Corona performed, even with a World Cup backdrop that should have helped beer volumes. The analysts say they still support management’s strategy. They also say they do not see a clear catalyst yet. That pairing is familiar. Strategy intact, timing unclear, rating slips to the middle.

The market, to its credit, gave the company the benefit of the doubt after an imperfect quarter. I like that reaction more than I like pounding the table. Beer imports with strong brand equity can absorb a soft three months. They cannot absorb a story in which the two flagship labels stop being the growth engine and nobody names the replacement. World Cup benefits that fail to show up in the print are a yellow flag, not a broken franchise. Yellow flags are why holds exist.

What would change my mind? A clean read on depletions, not just shipments. Evidence that promotional spend is buying sustainable volume rather than renting a month. A catalyst that is not “the next holiday.” Until then, supportive of the strategy is a fair place to stand, and not pounding the table is the honest addendum. Premium beer is still a consumer trade. Consumer trades on a morning of higher oil and higher yields do not get the benefit of a multiple argument. They get judged on the next volume print.

1. Separate brand equity from the latest depletion trend.
2. Treat a missing World Cup lift as information, not as noise.
3. Ask what the catalyst is before you pay for the strategy.
4. Keep position size honest while the rating sits at hold.

Netflix Still Has a Buyer Ahead of Results

Ahead of third-quarter results, one large brokerage reiterated a buy on the streaming leader. The framework is simple. If you believe revenue can still grow at a double-digit pace and margins can still expand, the stock is too cheap at these levels. The price target was nudged down by $3, to $80. A cut that small is a rounding exercise. It tells you the rating is intact and the model was marked to market, not torn up.

Everyone is worried about engagement. That worry is not imaginary. Hours, churn, and the advertising tier’s ability to hold attention are the live questions. The brokerage points to green shoots: a title called East of Eden, five Ice Age films licensed from the studio parent of those pictures, and cloud gaming as an extra surface. Green shoots are not a quarter. They are a reason not to assume the engagement debate is already lost. I have found that streaming arguments go extreme in both directions. Either the platform is a utility with pricing power, or it is a hit-driven media stock in disguise. The truth is usually a blend, which is unsatisfying and closer to how the cash flows behave.

Is the stock too cheap if double-digit growth and margin expansion both hold? On that if, yes. The if is the position. A $3 trim on the target does not weaken the if. A weak engagement print would. Into the report, the bull case is a valuation case wearing an operating case. The bear case is that engagement has quietly become the multiple. Thursday’s macro noise will not settle it. The quarter will.

Levi Strauss and the Wrong Pair of Jeans

Not a great quarter from the denim company. Shares fell almost 4 percent. The main problem was direct-to-consumer, which grew 2 percent organically against a consensus near 6.7 percent. That gap is the whole story in one comparison. Wholesale can be managed. The store and the site are where the brand either converts or excuses itself.

The chief executive’s explanation is specific enough to be useful. The back-to-school campaign pushed the wrong styles. The customer wanted low-rise. The floor had something else. They pivoted. Improvements are already showing up in the current quarter. I will take a merchandising miss over a demand collapse. A merchandising miss can be fixed in a season. A demand collapse takes a cycle. Calling the miss correctly, and showing early evidence of the pivot, is the encouraging part of an otherwise soft print.

Still. Almost 4 percent down is the market saying the direct channel is the valuation, and 2 percent versus 6.7 percent is not a small miss in a channel that is supposed to be the growth proof. Apparel investors have heard “wrong styles, we fixed it” before. Sometimes it is true by the next quarter. Sometimes it is the start of a run of excuses. The current-quarter improvement is the item to track, not the apology. If low-rise is what the customer is buying, the registers should say so without a footnote.

Direct channel scorecard:
  Organic growth reported: 2%
  Consensus bar: 6.7%
  Share reaction: about -4%
  Management read: style miss, pivot underway
  What matters next: current-quarter conversion

Consumer discretionary on a day of rising fuel costs has a second problem that has nothing to do with rise or low-rise. A household that spends more at the pump spends less on the extra pair. That macro overlay does not erase a self-inflicted assortment error. It does mean the recovery has to show up in a slightly worse spending backdrop. That is a harder test, and it is the test they just signed up for.

FedEx Freight and the Volume Question

A regional brokerage cut its price target on FedEx Freight to $129 from $160 and kept a hold. The cut is large enough to matter. The rating staying put tells you this is a reset of expectations, not a new short thesis. This earnings season, the question across less-than-truckload is whether a pickup in shipment volumes can sustain. Fuel surcharges will add noise. They always do when crude jumps 4.5 percent in a session. Noise is not the same as trend, but it will clutter the year-over-year comparisons that lazy models rely on.

For this particular freight business, shipment counts are the number I want. Not the surcharge. Not the mix anecdote. Counts. It is why adding to the position before that print is hard to justify. A hold with a lower target is the market’s way of saying the asset is real and the inflection is unproven. Less-than-truckload is a volume game with operating leverage that cuts both ways. Sustain the pickup and the target cut looks cautious. Miss the counts and $129 was not cautious enough.

Industrial activity, inventory restocking, and the simple question of whether shippers are moving more partial loads are the inputs. Oil complicates the read because fuel recovery mechanisms lag and then catch up. Analysts who want a clean volume signal will have to strip the surcharge out. If they do not, the group will look better or worse than it is, depending on the week crude chose. Thursday is one of those weeks.

Banks: Two Buys, One Hold, a Preference for Discount

Another desk put buy ratings on Wells Fargo and Goldman Sachs, and a hold on Bank of New York. The Wells argument is the asset-cap story maturing. Years of a regulatory ceiling are giving way to a bank that can compete for balance-sheet growth again. Maturing is not the same as finished. It does mean the constraint that defined the stock is no longer the only paragraph in the note. Buyers want operating leverage from a freed balance sheet without a fresh credit surprise. That is a reasonable ask. It is not a free one.

Goldman gets the nod over its investment-bank rival on valuation. A discount to a close peer is not automatically a bargain. It is a starting point. If advisory, trading, and asset-management fees hold up while the multiple stays behind, the relative case works. If the discount exists because the market doubts the durability of those fees, the case is a value trap with a better lobby. I lean toward the discount being at least partly undeserved when capital markets are open and the firm is not the one wearing a fresh strategic question mark. That is an opinion. The quarter will grade it.

On the trust banks, the same note is positive on the group and prefers State Street to Bank of New York. Custody and asset-servicing businesses rise and fall with market levels, net interest, and the stickiness of mandates. Positive on the group with a preference inside the group is how you stay constructive without pretending the names are identical. A hold on one of them is not a condemnation. It is a ranking. Rankings are what you want from research on a morning when the index is heavy and you still have to own financials somewhere.

Financials on a day of rising yields are a split tape of their own. Net interest income can like a steeper curve. Deal calendars and market-sensitive fees can dislike the risk-off tone that comes with an oil spike. You do not get to own “banks” as a single idea this morning. You own a cap-removal story, a valuation-discount broker, or a custodian, and you should know which sentence you underwrote.

How These Threads Collide at the Open

Step back and the session has a shape. Geopolitical oil and a long-bond auction set the discount rate and the risk premium. Asian index weakness sets the tone even though the chip fundamentals overnight argued the other way. Then the single-stock tape refuses to cooperate with a single narrative. Staples beat and guide down. A software name gets upgraded into a comeback. An AI infrastructure listing cannot clear a price. Beer loses a buy. Streaming keeps one. Denim misses the direct channel on a style error. Freight loses target price and keeps a hold. Money-center and trust banks get sorted by story rather than by sector beta.

That is a lot of information for a gap lower to erase. Gaps feel like verdicts. They are usually just the first trade. The useful work is deciding which of today’s moves you would still believe if oil gave back half the spike and the auction cleared cleanly. If your data-center peak call depends on the spike, it is not a call. If your staples trim depends on the guidance cut, it survives the auction either way. Process sounds dull until the open is loud.

A few practical filters, the kind I actually use when the headline count gets silly:

• Separate geopolitical premium from confirmed supply loss before you resize energy.
• Do not let a regional index drop overrule company prints from the same region.
• Treat a guidance cut as more important than a quarterly beat.
• Treat a first-day unlock above half the register as a reason to skip a new issue.
• In freight, wait for shipment counts once fuel surcharges start shouting.
• In banks, own the specific sentence, not the sector label.

None of those filters require a forecast of what the military option becomes. They require a forecast of what you are actually betting on. Most bad mornings in the stock market are bad because those two forecasts get swapped.

Rates, Dividends, and the Relative Game

Higher rates do a quiet violence to anything owned for income. The dividend yield that looked generous against a sleepy long bond looks ordinary when the bond wakes up. PepsiCo is the clean example today because the fundamental news is mixed and the rate news is not. You can like the international franchise and still admit the stock has more competition from Treasuries than it did last week. That competition does not force a sale. It forces a comparison. Comparisons are where portfolio mistakes get corrected without drama.

The same rate move leans on long-duration growth, which is why the data-center complex feels the headline even when the chip numbers are good. Duration is not a moral category. A cash-flow stream that arrives later is worth less when the discount rate rises, full stop. If the cash-flow stream is also being revised up, the two forces fight. Overnight, the revision force showed up in semiconductors and the discount-rate force showed up in oil and yields. Futures chose the discount rate. That choice can reverse by lunch if the auction behaves and crude stalls. It can also extend if both keep moving.

I would rather own a business that does not need the auction to go well. That preference is not always available at a fair price. On mornings like this it is at least a sorting tool. Who needs lower yields to justify the multiple? Who can fund the dividend and the buyback if yields stay here? Who just cut the year and is asking you to wait? The answers are in the releases, not in the futures quote.

What I Would Not Do With This Tape

I would not declare a peak in data-center spending off an oil headline and a Kospi slide. The evidence from the foundry and the memory leader points the other way, and evidence beats mood. I would not chase an Australia infrastructure listing that cannot indicate a structure and that frees more than half the register on day one. I would not add freight exposure solely because a target was cut, unless the shipment counts are the thing I came for and I am willing to wait for them. I would not treat a $3 trim on a streaming target as a change of view.

I would also not ignore the auction. Pretending bonds do not matter because you prefer equities is how relative performance leaks away in a week. Watch the bid-to-cover, the tail, and whether indirect bidders show up. Then decide whether today’s yield move is a headline or a new range. Ranges change position sizes. Headlines change conversations. Know which one you are in before you trade the close.

Morning sort: oil premium + auction risk + company-level revisions. Trade the third. Respect the first two.

There is a temptation, every time crude spikes on a conflict headline, to rebuild the whole portfolio around energy and defense and to fade everything else. Sometimes that temptation is early and right. More often it is a one-day factor storm. The reports say large-scale operations are being considered. Considered is not ordered, and ordered is not executed. Pricing the option is rational. Pricing it as a base case before the afternoon bond sale is how cash accounts get spent twice.

A Closer Read on the Consumer Names

Three consumer stories landed in the same window, and they do not rhyme as neatly as a sector note would like. The snack and beverage company beat the quarter, grew organic sales ahead of plan, and cut the full-year core profit path while promising urgency at home. The beer company lost an upgrade-level rating because flagship import labels did not deliver even with a tournament tailwind, though the strategy still has supporters. The denim company missed badly in the channel it has asked investors to care about, explained the miss as a style error, and pointed to a pivot already visible in the current quarter.

If you squint, all three are stories about a consumer who is still spending but is pickier about where. International holds up better than North America in beverages. Brand equity in beer is not the same as a depletion beat. Jeans sell when the rise matches the moment, and they do not when it does not. None of that requires a recession call. It requires merchandising, price architecture, and a bit of humility about how much a sporting event can do for a quarterly volume line.

The rate backdrop makes the pickier consumer more important, not less. When the risk-free yield rises, you are paid less for waiting on a turnaround. Urgency in North America has to show up in the next organic number, not in the adjective. A beer catalyst has to be named. A denim pivot has to clear the 6.7 percent bar it just missed, or at least close the gap enough that the direct channel stops being the objection. I am willing to give an operator one season to fix a style miss. I am less willing to give a staples giant a pass on a full-year cut just because the quarter cleared.

Semiconductors Versus the Narrative

Go back to the chip prints, because they are the fact the peak narrative has to climb over. Incredibly good numbers from the leading foundry and from the leading memory producer do not guarantee that every U.S. data-center landlord or every accelerator designer is cheap. They do undermine the claim that spending has already rolled. Foundry commentary on advanced nodes and memory commentary on pricing are about as close as public markets get to a real-time read on whether the buildout is digesting or still allocating.

Why, then, the Kospi drop of more than 2.5 percent? Positioning, currency, and the same oil shock that hit every risk market before New York opened. Foreign investors can sell a strong fundamental tape if they are reducing Asia beta into a geopolitical headline. That selling exports itself into U.S. futures because the correlation book is real. By the time cash equities open, the fundamental print and the flow print are telling different stories, and commentators pick the one that fits the lower open. I would pick the print. Flows reverse faster than order books.

There is still a legitimate worry inside the complex, and it is not oil. It is concentration. A handful of buyers, a handful of sites, a power constraint that does not care about earnings beats. If those buyers pause together, the overnight strength becomes a lagging indicator. Nothing in this morning’s headlines is that pause. Calling it a pause because yields ticked up is how the excuse earns its keep.

New Issues and the Quality of the Bid

Primary markets are a side door into the same mood. When a bookbuild for an AI infrastructure listing closes on schedule and still cannot indicate structure, the bid is selective. Selective is healthy. The unhealthy version is a deal that reaches the market anyway, with a loose lockup, and then trades as a disappointment that stains the next three filings. More than half the shareholders free to sell on day one is the detail I cannot get past. Alignment between new buyers and existing holders is the minimum. This structure does not meet the minimum.

Poor placement, a short track record, and a high valuation would each be manageable with the other two in better shape. Together they explain a process that ends without a price. Nvidia’s backing, to use the sponsor’s name plainly, gets meetings. Meetings are not orders at any price. In a tape already nervous about duration and about whether data-center spending is late-cycle, infrastructure paper needs contracts, power, and a lockup. It got a process deadline instead.

Skip it. There will be other ways to own power, sites, and compute without funding an exit on the opening cross. The fact that this one looks shaky is not a verdict on the whole theme. It is a verdict on this deal’s sequencing. Those are different sentences, and only one of them belongs in a portfolio.

Holding the Line Into the Auction

Between the open and the 30-year sale, the useful posture is boring. Let the oil spike settle into a range rather than a spike-and-reverse or a spike-and-extend. Read the company items that do not depend on the headline. PepsiCo’s cut. Palantir’s upgrade and the $230 target. The Firmus Grid book that could not clear. Constellation’s downgrade to hold. Netflix’s reiterated buy with a target at $80. Levi’s direct-channel miss and the style pivot. FedEx Freight’s target at $129. The bank ranking that prefers a discount and a maturing cap story over a custodian that sits in the middle.

If the auction is strong, some of the morning’s multiple compression can unwind, and the chip fundamental story gets a cleaner hearing. If the auction is weak, the excuse merchants will have a second wind, and dividend stocks will look even less special into the close. Either outcome is tradable. Neither outcome rewrites the company releases that already happened. That distinction is the whole edge available on a day like this, and it is smaller than the futures board makes it look.

I keep coming back to a simple preference. Markets that need the bond desk to cooperate are tougher to own at rich multiples than markets that can ignore it. Today the stock market is being asked to cooperate with oil, with a long-bond sale, and with a stack of earnings that refuse to point the same direction. You do not have to solve all three. You have to know which one your position actually needs. Most people, if they are honest, own the excuse and think they own the evidence. Thursday is a good morning to check.

Own the release you can defend after the headline fades. Rent the macro if you must. Do not confuse the two at the open.

The lower open will feel decisive for about twenty minutes. Then the auction time will show up on the calendar, crude will either hold the premium or leak it, and the single-stock arguments will still be sitting there, unimpressed. That is the part worth staying for. Not the first print. The argument underneath it.