The Market Tape — End-of-Day Wrap
SPY closed today at $767.58, down 0.75% from Tuesday's $773.38 and now about 1.3% below the $777.88 high the index set August 13 — a clean risk-off session, and one with an identifiable cause sitting underneath the index-level number rather than a diffuse macro story.
That cause was a single product announcement. Meta unveiled Muse, a new AI agent capable of booking travel and completing other transactions directly on a user's behalf, bypassing the search results and aggregator apps that currently sit in between. The most direct hit landed on the online travel intermediaries themselves: Expedia fell 7%, Airbnb dropped 6%, and Booking Holdings sank 5%, per 24/7 Wall St.'s Wednesday coverage, effectively pricing in the risk of an AI agent cutting them out of the transaction. Alphabet (GOOGL), which I don't hold, absorbed a similar disintermediation read on its own core search-and-ads business — shares fell 3.70% to $338.15, a much sharper drop than the broad index took. Charles Schwab got pulled into the same story from a different angle: a Citrini Research note flagged by Benzinga estimated the brokerage collects nearly $9 billion a year in “lazy cash” revenue from idle balances and elevated margin rates that persist largely because switching platforms is friction-heavy — exactly the kind of friction an AI agent doing the comparison shopping itself could erode.
Meta, the company that actually launched the disruptive product, closed up 1.10% to $744.72, and Microsoft added 0.52% — while enterprise software names that could end up supplying the infrastructure behind agents like Muse (Atlassian, Monday.com, Salesforce, ServiceNow) rallied even as the broader tech tape slid, a rotation this brief has flagged in a similar shape before. Intel fell 2.24% despite some optimism that agent-driven cloud demand could eventually help its server business, per Yahoo Finance — investors reportedly still want actual design wins before repricing that story. That split showed up across the three benchmarks I track: QQQ, the most exposed to both the stocks at the center of the story and the uncertainty around them, was the day's laggard at -0.87% to $740.96; DIA fell a close -0.71% to $514.31, with SPY landing in between at -0.75%. None of the three was spared, which is itself informative — this wasn't purely a mega-cap-tech story, it bled into the broader tape.
Crypto moved with the rest of the risk-off tape today rather than shrugging it off, a break from the pattern this brief has noted on several recent sessions. Bitcoin fell 2.65% to $84,405.98, ether dropped 3.55% to $2,675.46, and solana fell 3.99% to $114.37 — all real declines, roughly in line with or worse than the equity benchmarks.
On the position side, today's most concrete event was a decision of mine rather than a market move: I added another 25 shares of Home Depot (HD) in the Trading account (account ending ••••8840) this morning, filled at $299.58 at 12:12 p.m. ET, bringing the position to 75 shares at a $300.70 average cost — a continuation of yesterday's first buy into the name near a 52-week low. HD kept falling anyway today, down 2.78% to $296.87 from Tuesday's $305.35, and I couldn't confirm any company-specific news behind the drop, Muse-related or otherwise, so I'm logging it plainly as a position that's gone the wrong way twice in two days rather than reaching for an explanation that isn't there.
The rest of the book was a mix of unexplained moves and quiet ones. SoFi (SOFI), 200 shares in the Trading account at a $17.10 average cost, fell 3.26% to $16.60 with no ticker-specific news I could confirm — a bigger decline than the benchmarks took, in a name this brief has flagged for unexplained moves before. SpaceX (SPCX), 153 shares at a $124.25 average cost, fell 4.29% to $148.08, also without a confirmed catalyst; the Agentic account's SPCH (account ending ••••1117), the fund targeting twice SPCX's daily move, fell 8.55% to $9.73 for the same mechanical reason. Quantum Computing (QUBT) and KLA Corporation (KLAC) were both essentially flat — QUBT down 0.33% to $9.05 in the Trading account, KLAC down 0.35% to $187.66 in the Agentic account — a quiet day for both after recent volatility in each.
Two real dates are now closer. Initial jobless claims are due tomorrow morning, Thursday, September 24, at 8:30 a.m. ET, followed by August's durable goods orders on Friday. The bigger date sits about a week out: the Fed's preferred inflation gauge, August's PCE price index, lands Wednesday, September 30, alongside the Bureau of Economic Analysis's third estimate of second-quarter GDP — both confirmed directly against BEA's own release schedule rather than assumed.
Why this matters: a single AI product launch moving Alphabet, Expedia, and Charles Schwab in the same session is a reminder that the AI-disruption story isn't confined to chipmakers and data-center names anymore — it's starting to touch the business models of companies whose whole value proposition is sitting between a customer and a transaction. And on the position side, today's a useful reminder that not every red print has a story behind it: SOFI and SPCX both fell harder than the market with nothing I could find to explain why, worth sitting with rather than backfilling a narrative that isn't actually there.
The Market Tape — End-of-Day Wrap
SPY closed today at $773.50 — exactly where it closed Monday, a full round trip after touching $774.69 at the open and $774.67 as late as 3:30 p.m. ET. The index spent most of the session green, gave almost all of it back in the final half hour, and finished the day precisely unchanged. That's a more interesting number than a garden-variety flat close: it means today's real story sits below the index level, not in it.
That story was a genuine divergence between the three benchmarks. QQQ led by a wide margin, up 0.78% to $747.22, while DIA lagged the whole session and closed down 0.34% to $518.03 — SPY landed in between, flat, because it's a blend of both. The gap traces to a real, identifiable split: oil extended its slide for a fifth straight session on reports of diplomatic efforts to de-escalate the U.S.-Iran standoff, and Treasury yields eased alongside it, a combination that tends to favor growth and tech over the more industrial, rate-sensitive names that anchor the Dow. Semiconductors did the heaviest lifting on the Nasdaq side: the Philadelphia Semiconductor Index added 1.8% and the State Street SPDR S&P Semiconductor ETF (XSD) rose 1.7%, per MT Newswires' Tuesday afternoon sector wrap, with President Trump separately rejecting calls for new AI-development regulation — the kind of headline that tends to keep the AI-infrastructure trade running rather than cooling it off.
Crypto stayed muted by comparison: bitcoin closed up 0.82% to $86,379.13, ether up 0.58% to $2,755.01, and solana up 1.31% to $118.31 — real gains, but nothing like Monday's short-squeeze-driven pop toward $85,000 that Benzinga flagged crypto-linked equities riding at the time. Today reads as a straightforward equity/AI-infrastructure story, not a broad risk-on move spilling out of crypto.
The two chip-adjacent names I hold both moved with that sector tape rather than their own news. Quantum Computing (QUBT), 500 shares in the Trading account (account ending ••••8840) at an $8.81 average cost, closed up 2.12% to $9.15 — I couldn't find any company-specific headline behind it today, so I'm logging it as riding the broader semiconductor strength rather than its own story. KLA Corporation (KLAC), the Agentic account's position (account ending ••••1117, 2.28 shares at a $189.93 average cost), closed up 2.18% to $187.97 for the same reason — still down a little over 1% from my own cost basis despite the gain.
SpaceX (SPCX), 153 shares in the Trading account at a $124.25 average cost, had an actual catalyst behind today's move: Bloomberg reported, via a company presentation shared at an event in London, that SpaceX's Grok Bot AI agent has topped 400,000 weekly users less than a month after launch, up 24% week over week as of Sept. 14. Shares closed up 1.76% to $154.53, now up about 24.4% on cost. The Agentic account's SPCH, the fund targeting twice SPCX's daily move, closed up 3.51% to $10.61 for the same mechanical reason, up roughly 30% on its $8.16 average cost.
SoFi (SOFI), 200 shares in the Trading account at a $17.10 average cost, had a real, company-specific reason to move today too: SoFi and Mastercard announced that stablecoin settlement is now live across SoFi Bank's entire $25 billion debit and credit card program, using SoFi's own SoFiUSD stablecoin on Mastercard's global payments network. Shares closed up 1.12% to $17.16, essentially flat against my cost basis — a real product announcement, but not the kind of news that necessarily moves a stock hard on the day it lands.
The other real event in the Trading account today was a new position, not a held one. I bought 50 shares of Home Depot (HD) this morning in two 25-share fills around 10:28 and 10:29 a.m. ET, averaging $301.26. That's a buy into weakness by design: HD closed Monday at $297.20, a fresh 52-week low, down 0.93% on a day the Dow itself gained. Today the stock rebounded 2.76% to $305.39 — but I couldn't confirm any specific news behind the bounce, so I'm logging the trade plainly, as a decision made into a beaten-down name near multi-month lows, not as a read on what today's rebound means. Checking prices again close to filing this, roughly 45 minutes past the bell, none of today's positions — HD included — show a material move in the early after-hours prints.
Two real dates are now a little closer. Weekly jobless claims, durable goods orders, and the final Q2 GDP read are due Thursday morning, and August's PCE index — the Fed's preferred inflation gauge — lands Friday, Sept. 26, alongside the final University of Michigan consumer sentiment read. Neither is a surprise addition to the calendar, but both are now inside the week.
Why this matters: a flat SPY close can hide a real day underneath it, and today is a clean example — falling oil and easing yields pulled money toward chips and AI-infrastructure names and away from the Dow's more industrial tilt, even as the headline index number looked like nothing happened. On the position side, today was a mix of real catalysts (SoFi's stablecoin rollout, SpaceX's Grok Bot user numbers) and moves that don't have one yet (QUBT, KLAC, and this morning's Home Depot bounce) — worth being honest about which is which rather than backfilling a story for a stock that simply moved with its sector.
The Market Tape — End-of-Day Wrap
SPY closed today at $762.65, up 1.14% from Wednesday's $754.05 — a full reversal of Wednesday's Fed-decision-day dip, and enough to put the index back above Tuesday's pre-Fed close of $757.86. It's still about 2% below the $777.88 high SPY set August 13, but today read like relief rather than more selling.
The relief had a specific, identifiable cause. The Federal Reserve, under new Chair Kevin Warsh, raised its benchmark rate 25 basis points to a 3.75%–4.00% range Wednesday in a unanimous vote — its first hike since July 2023 — while signaling at least one more increase is likely later this year. That's the kind of headline that should keep pressuring stocks, and it did on Wednesday; instead, today's session moved almost entirely the other way. Oil gave back some of its recent geopolitical premium, with WTI crude falling 1.3% to $101.15 a barrel and Brent down 1.8% to $103.91, and the 10-year Treasury yield eased to roughly 4.93%–4.95%, down from the 5.04% multi-year high it touched earlier this week, per Benzinga's Thursday wrap. The CBOE Volatility Index fell 11.4% to 15.69. Two data points reinforced the calmer mood: weekly jobless claims fell to 196,000, well below the 206,500 consensus and close to a 60-year low, and the Philadelphia Fed manufacturing index, while slowing to 37.8 from 47.4, still beat the 32.1 estimate comfortably.
That combination showed up unevenly across the three benchmarks I track. QQQ led with a 1.72% gain to $716.81, DIA lagged at just 0.59% to $518.26, and SPY landed in between at 1.14%. The gap traces to two different stories inside the index. Chipmakers ripped higher on a real, specific catalyst: Reuters reported this week that SK Hynix is in early talks with Intel about manufacturing memory chips at Intel's long-idle Ohio complex, adding to a memory-pricing story Intel CEO Lip-Bu Tan flagged earlier in the week (prices up five-to-sevenfold on tight AI-driven supply). Intel closed up 7.59% to $108.72 on the day, with Barclays upgrading the stock to Overweight the same day the Fed hiked and a Melius Research note floating a two-year, $200 price target. AMD (+6.38%), Micron and Oracle rode the same memory/AI-infrastructure wave, per Benzinga. DIA's underperformance had its own real cause on the other side of the ledger: as oil's geopolitical premium unwound, so did the defense trade that had been propping up the Dow — RTX, Lockheed Martin, General Dynamics and Boeing all fell today as that risk premium deflated, per Benzinga's Thursday market wrap.
Crypto barely moved by comparison — bitcoin up 0.10% to $76,455.08, ether up 0.60% to $2,448.37, and solana up 1.07% to $100.95 — the same pattern this brief has flagged before, where a big one-day equity move doesn't translate into a matching move in risk appetite broadly.
One of my own trades sits right in the middle of this story. I added 100 shares of SoFi (SOFI) in the Trading account (account ending ••••8840) yesterday, filled at $16.715 at 3:04 p.m. ET on the Fed-decision day itself, bringing the position to 200 shares at a $17.10 average cost. The logic behind SoFi's earlier weakness has been that higher long-term rates raise funding costs for a balance-sheet lender; today's yield decline should have been a tailwind by that same logic, but SOFI closed essentially flat, down 0.65% to $16.73 — a reminder that the yield-sensitivity story is directional, not a guarantee of a same-day reaction in either direction.
The rest of the Trading account split between an unexplained move and a real one. Quantum Computing (QUBT) jumped 6.13% to $8.49 — a much bigger move than the broad small-cap tape — but I couldn't find any company-specific news behind it today, so I'm logging it as an unexplained move rather than forcing a narrative onto it; the position is still down about 3.6% from my $8.81 average cost. SpaceX (SPCX), 150 shares at a $123.71 average cost, rose 2.51% to $154.66 on real news: Bloomberg reported today that the company has held internal talks about buying data from defunct startups to train its AI models, and separately, SpaceX has a confirmed Sept. 22 Starship flight on the calendar — its first attempt to reach orbit while carrying an operational payload of 26 Starlink satellites, per Benzinga. The position is now up about 25% on cost. The Agentic account's SPCH, the levered SpaceX-tracking fund (account ending ••••1117), rose 4.92% to $10.66, up roughly 30.6% on its $8.16 average cost for the same reason. KLA Corporation (KLAC), the Agentic account's other position, gained just 0.69% to $168.51 — a quiet day despite a bullish call-option sweep flagged in Benzinga's options-activity scan, and still down about 11.3% from my $189.93 average cost. It's a reminder that semicap-equipment names don't automatically ride the same wave as the memory makers whose capacity they help build.
Why this matters: a Fed rate hike is exactly the kind of headline that's supposed to cool a market down, and today was followed by one of the better sessions of the month anyway — a reminder that a single data point rarely tells the whole story on its own; oil retreating and a soft-landing-consistent jobs number did more work today than the hike itself. And on the position side, today was a good day to notice the moves that don't fit the story I'd have told about them ahead of time: SoFi shrugging off a yield decline that should have helped it, and Quantum Computing jumping 6% with no news behind it at all. Worth treating both as open questions rather than filling in an explanation that isn't actually there yet.
The Market Tape — End-of-Day Wrap
SPY closed today at $757.86, down 0.40% from Monday's $760.65 — a second straight loss, and now roughly 2.6% below the $777.88 closing high the index set August 13. But today's decline had almost nothing to do with the AI-safety story that drove Monday's session; it was a different set of mechanics entirely.
The real driver was energy. WTI crude jumped 3.7% to $105.12 a barrel and Brent rose 2.4% to $108.20, after Saudi Aramco reportedly canceled or delayed cargoes to European refiners, per Benzinga's midday wrap. U.S. diesel prices climbed to a record average of $6.26 a gallon, above the previous high set in 2022, with heating oil futures up 5.5% and gasoline up 3.2%. President Trump wrote on Truth Social that "the world's diesel price rise is mostly caused by the Russia/Ukraine War, not Iran." It showed up unevenly across the three benchmarks: DIA, the most industrial-heavy of the three, fell 0.60% to $521.36, a shade worse than QQQ's 0.56% drop to $705.18 and SPY's 0.40% decline, with the Energy sector the only one of the eleven S&P sector SPDRs posting a real gain on the day, per Benzinga.
Layered on top of the oil move, the 10-year Treasury yield extended its climb for a fifth straight session, touching 5.04% intraday before settling near 5.00%. Treasury Secretary Scott Bessent addressed it directly with lawmakers, saying "the need to address [the] deficit is one factor affecting 10-year treasury bond yield," while stressing Treasury doesn't "set equilibrium price for yield levels." That backdrop sits directly ahead of tomorrow's Fed decision: per Benzinga's citation of the CME Group's FedWatch tool, traders are now pricing roughly a 92% probability of a 25-basis-point hike at Wednesday's 2 p.m. ET announcement — which would be the Fed's first hike since July 2023, and is now less than a day away.
That yield move is a real, mechanical headwind for one of my own positions. SoFi (SOFI), 100 shares in the Trading account (account ending ••••8840) at a $17.48 average cost, closed down 3.12% to $17.10 — for the same balance-sheet-lender reason this brief has flagged before: higher long-term rates raise SoFi's funding costs and widen securitization spreads on a loan book it carries directly, not because of anything company-specific today. The position is now down about 2.2% from cost.
The other real story of the day happened in Washington, not on a trading desk: the Senate's procedural cloture vote on the CLARITY Act, the crypto market-structure bill, failed to reach the 60 votes needed to advance, per CNBC, after last-minute Democratic objections to the bill's ethics and stablecoin-yield provisions. Coinbase (COIN), which isn't a position I hold, closed down 10.25% to $171.83 — roughly four times bitcoin's own decline — while Circle Internet Group fell nearly 9% intraday on the same news, per Benzinga. Bitcoin itself closed down 2.56% to $75,674.57, ether fell 3.96% to $2,399.43, and solana dropped 4.57% to $96.90: a real pullback, but a much smaller one than what hit the exchanges and legislation-linked equities most directly exposed to today's vote.
The weekend's AI-safety debate kept threading through my other two accounts today, this time from a new angle. At the All-In Summit, Elon Musk warned that existing chip fabs are running at full capacity and that a disruption in Taiwan could compound the shortage — the reason, he said, Tesla and SpaceX are building their own "Terafab" facility. SpaceX (SPCX), 150 shares in the Trading account at a $123.71 average cost, closed down 3.25% to $143.33, still up about 15.9% on cost; the Agentic account's SPCH, a fund targeting twice SPCX's daily move, fell 6.61% to $9.19, still up roughly 12.6% on its $8.16 average cost. KLA Corporation (KLAC), the Agentic account's other position (2.28 shares at a $189.93 average cost), was named in the same Terafab coverage as one of the equipment suppliers that could benefit if fab-building shifts outside Taiwan — it closed down just 0.39% to $168.43, a quiet day after Monday's 6% drop on the Third Point disclosure, still down about 11.3% from cost. Quantum Computing (QUBT), 500 shares in the Trading account at an $8.81 average cost, fell 1.90% to $7.82 with no company-specific news I could confirm today — it reads as broad small-cap weakness (the Russell 2000 fell about 0.9%, per Benzinga) rather than its own story, now down about 11.2% from cost.
Why this matters: three separate mechanisms moved money today — an oil-supply shock, a multi-year high in long-term rates, and a specific legislative setback for crypto market-structure rules — and none of them needed an AI headline to do it. Worth sitting with after a week where the AI-safety debate has dominated the narrative: real markets keep running more than one story at once, and a position like SoFi can have a genuinely bad day for reasons that have nothing to do with whatever's trending that week.
The Market Tape — End-of-Day Wrap
SPY closed today at $760.65, down 0.48% from Friday's $764.35 — a second straight loss that picked up almost exactly where this morning's brief left off, on the same weekend AI-safety story that hit futures before the open.
The session turned into a real capital rotation, not just a broad risk-off day. AI-infrastructure hardware kept bleeding: Nvidia fell 3.57% to $210.50, AMD dropped 4.45% to $493.17, and Intel slid 5.59% to $97.19. Chip-equipment names took it further — Applied Materials and Lam Research each fell roughly 6% and ASML sank about 5%, per 24/7 Wall St.'s Monday coverage of what it called the "AI pacing call reaching fab spending." Hewlett Packard Enterprise dropped 9% (on a downgrade after a big recent run, a company-specific reason layered on top of the sector weakness), with Dell off 5% and Super Micro down 8% in the same AI-server complex. On the other side of the rotation, megacap platform names actually gained: Alphabet rose 2.97% to $348.56, Meta added 2.56% to $664.61, and Microsoft climbed 1.97% to $505.40. Enterprise software did even better — ServiceNow, Adobe and Salesforce all gained, with coverage from 24/7 Wall St. and Yahoo Finance framing it as capital "sidestepping" the AI-infrastructure selloff. Cybersecurity names jumped hardest of all: CrowdStrike and Zscaler each rose roughly 12% and Palo Alto Networks about 11%, after Microsoft published a new "Humanist AI Code of Conduct" framework that investors read, per Benzinga's reporting, as a reason to own the security layer sitting on top of increasingly autonomous AI systems.
One data point today made the weekend debate feel less like market mood and more like an actual operating decision: per Yahoo Finance, Nvidia and Palantir — along with Booz Allen Hamilton — are reportedly restricting their own use of advanced models from Anthropic and OpenAI over concern those models could retain or learn from sensitive corporate data, and are building an alternative using Palantir's software paired with Nvidia's own open Nemotron models instead. Anthropic's response, per the same report, is a new "Enterprise Frontier Safeguards" framework that stores customer activity data in infrastructure the customer controls. That's a real company decision tied to this weekend's essay, not just a repricing of sentiment.
KLA Corporation (KLAC), the small position I hold in my Agentic account (account ending ••••1117, 2.28 shares at a $189.93 average cost), closed down 6.01% to $169.79 — a sharper drop than Nvidia's, and one with a company-specific layer on top of the sector story. A regulatory filing reported today by Benzinga showed billionaire investor Dan Loeb's Third Point exited its entire KLA and Lam Research positions during the second quarter of 2026, while more than doubling its ASML stake — worth being precise that this is a lagging disclosure reflecting a decision made months ago, not a trade placed today, but it's real information that landed on the stock the same day it was already absorbing the broader capex-pacing worry. The position closed down about 10.6% from my own cost basis, which is actually a bit better than this morning's premarket read near $167 — some of the damage was pared back during the session even though the stock still finished the day down sharply.
The other real event in my own book today happened in the Trading account: I sold the entire 100-share Lululemon (LULU) position, filled at a $100.705 average price at 11:30 a.m. ET against a $99.28 average cost — a gain of roughly $142 before a $0.23 fee, a little under 1.5% on the shares. The sale landed the same morning Rothschild & Co Redburn cut its price target on the stock to $90 while keeping a Sell rating, already flagged in this morning's brief — worth noting plainly that LULU itself still closed up 1.34% on the day, at $100.30, a similar pattern to what this brief has flagged before on the stock absorbing bearish calls without much visible damage. Logging the trade as the decision made today, not as a read on where the stock goes from here. The rest of the Trading account moved without much company-specific news I could confirm: SoFi (SOFI) closed up 1.62% to $17.60, essentially flat against my $17.48 average cost; Quantum Computing (QUBT) fell 0.88% to $7.92, now down about 10% from my $8.81 average cost; SpaceX (SPCX) slipped 2.17% to $147.93, still up roughly 19.6% on my $123.71 average cost despite the pullback. The Agentic account's SPCH position, a fund targeting twice SPCX's daily move, fell 4.29% to $9.82 for the same mechanical reason — still up about 20.3% on its own $8.16 average cost.
Crypto shrugged off the whole AI-infrastructure story again, the same way it did in premarket trading this morning: bitcoin closed up 1.63% to $78,833, ether up 1.44% to $2,548, and solana up 2.4% to $103.37 — one more data point that today's rotation was specifically an equity, AI-capex story rather than a broad risk-off move. The divergence showed up across the major benchmarks too: QQQ, the most AI/semis-heavy of the three, fell 0.77% to $709.37, while DIA, the least concentrated in chips, fell just 0.30% to $524.23 — with even that smaller move getting some help from Caterpillar's roughly 4% drop on what Yahoo Finance described as the data-center power trade "unwinding" alongside the broader AI-capex pullback.
One thing is now a trading day closer: the Federal Reserve's two-day meeting begins tomorrow, Tuesday, September 15, with the rate decision due Wednesday at 2 p.m. ET — the same meeting this morning's brief flagged, with markets pricing an 88.5% probability of a 25-basis-point hike as of this morning, per Benzinga's Monday wrap citing the CME Group's FedWatch tool. Nothing in today's session changed that calculus directly; it's simply one day nearer now.
Why this matters: Bank of America's own read today, per Benzinga, was that Amodei's AI-safety warning is "noise" in what the firm still calls a "secular bull market" — worth holding next to today's very real, very large sector-level moves. A capital rotation out of AI-infrastructure hardware and into software and cybersecurity isn't the same thing as a broad market selloff, and today's crypto and Dow-adjacent action both back that reading up. Where it gets less academic is a position like KLAC: a stock can get hit by a real sector story and a real, if lagging, ownership-change disclosure on the very same day, and it's worth being able to tell those two forces apart rather than reading one closing number as a single cause.
The Market Tape — End-of-Day Wrap
SPY closed at $764.35, up 0.86% from Thursday's $757.83 close and about 1.7% below the $777.88 high the index set August 13 — a session almost entirely explained by one data point released before the opening bell.
That data point was August's CPI report, which held at 3.4% year-over-year, unchanged from July and in line with what the market had already priced in. The reaction was immediate: by 10:38 a.m. ET, the Dow was up more than 1%, the Nasdaq up 1.20%, and the S&P up 1.08% on the in-line print, with oil trading lower and gold catching a bid — the shape of a fairly standard "no surprise, no reason to reprice" relief rally. By the close, SPY's gain had narrowed to 0.86%, a less exuberant number than the mid-morning read but still a solidly green session, not a fade back to flat.
One position in my own book told a more interesting story underneath that macro tape. Lululemon (LULU), held in my Trading account, closed at $98.94, up 2.12% on the day — despite BMO Capital Markets initiating coverage this morning with an "underperform" rating, arguing the athletic-wear name faces a genuinely difficult turnaround (stale product, share losses to rivals, a bear case that puts the stock near $50) even with new CEO Heidi O'Neill only a few days into the job. That's about as bearish a note as a stock gets in a single morning, and the tape mostly shrugged it off. Worth being precise about what that does and doesn't mean: on a day the broad market is rallying on a friendly macro print, a name can absorb a negative analyst call and still close green just by riding the tide — that's a different, weaker kind of "up" than a stock rallying on its own news, and today's green close isn't itself evidence that BMO's thesis is wrong.
The rest of my Trading-account positions — SOFI, QUBT, SPCX — moved with the broad market today; I didn't find anything ticker-specific attached to any of the three, which fits a session where one macro print was doing the work rather than individual stock news. (This wrap draws on the Trading account only, not the Agentic account, so it isn't a full-book summary.)
One thing is now a few trading days closer: the Fed's rate decision lands Wednesday, September 16, following the FOMC's two-day meeting that starts Tuesday. Today's in-line inflation read is one of the last major data points the committee sees before that decision — it removes one argument for a hawkish surprise without, on its own, confirming anything about what comes next.
Why this matters: a rally built on an in-line inflation print is a "no bad news" rally, not a "good news" one — worth keeping that distinction in mind before reading too much bullish signal into today's green close. And a stock that rallies through a bearish analyst note on a day the whole market is up isn't the same as that stock proving the note wrong — two things worth keeping separate heading into next week's Fed decision, not blurred into "the market liked LULU today."
SPY, the S&P 500 ETF, closed Thursday at $758.04, down 0.57% — down from Wednesday's $762.40 and now about 2.6% below the $777.88 high it set on August 13. It was the fourth straight losing session for the major averages, and the story underneath today's tape was less about any single company and more about what's happening in the energy market.
WTI crude settled at $102.48 a barrel, up 6.7% on the day, with Brent finishing at $107.63, up 5.9% — both benchmarks back above $100 for the first time in months. The move tracks reporting of intensified attacks on tankers near the Strait of Hormuz following a reported U.S. strike on Iranian oil vessels earlier in the week. I'm not in a position to verify the underlying military reporting beyond what's been published by wire services, and it's genuinely still developing — but the market reaction to it is the relevant part here: a sudden jump in energy costs is itself an inflation input, and it landed on a day that already had inflation data on the calendar.
That data was this morning's Producer Price Index for August, released at 8:30am ET. The headline monthly number (+0.4%) actually matched what was expected. What moved markets more was the trend underneath it: the 12-month unadjusted rate came in at 5.4%, up from July's 4.8% pace, and the core reading (excluding food, energy, and trade services) still rose 0.3% for the month and 4.7% over the past year. A monthly print that's in line can still read as a warning sign if the year-over-year trend is accelerating — and that's what happened today, on the same session oil spiked.
One name in my own book moved on its own story rather than the broader tape: Lululemon (LULU) closed at $96.55, down 3.18%, and slipped back below my $99.28 average cost on the 100 shares I hold in my Trading account — giving back Tuesday's post-CEO-announcement bounce to $103.19, and then some. The average cost itself moved this week too, worth logging plainly rather than just showing the new aggregate: I bought 45 more shares Wednesday morning at 9:40am ET, filled at $98.49, on top of the 55 shares held since last week's guidance-cut selloff — that trade, not any restatement, is what moved the average from $99.93 to $99.28. Two real, dated items explain today's move better than the broad tape does: BMO Capital Markets initiated coverage on LULU Wednesday with an "underperform" rating and a $70 price target, and Levi & Korsinsky put out a notice the same day that it's investigating LULU's disclosures related to IEEPA tariff refunds. I can't independently verify either beyond the fact that both were published — an analyst initiation and a plaintiff-firm investigation notice are two different kinds of signal, and the second one in particular is worth treating as exactly what it is (a law firm soliciting a case) rather than confirmation of wrongdoing. Either way, it's a more plausible explanation for today's move in that specific name than anything in the broader tape.
The rest of the Trading account mostly just moved with the market — SoFi (SOFI) down 0.69% to $17.21, Quantum Computing (QUBT) down 2.74% to $7.82. SPCX had a wide session, swinging from $144.89 to $154.70 before closing at $147.81, up 0.17% on the day; I don't have a confirmed catalyst for that range beyond ordinary volatility in a name that already trades with a lot of it. (This wrap pulls positions from the Trading account only, same as it's scoped to do — the Agentic account isn't part of this read.)
One more thing now a day closer: August's CPI is due out tomorrow morning, Friday, September 11, at 8:30am ET. Today's PPI trend and the oil move both raise the stakes on that number a little — a hot CPI print on top of today's action would be a different market than a cool one.
Why this matters: today wasn't really a stock-picking day — it was an energy-and-inflation day, and the LULU move is a reminder that even on a macro-driven tape, individual positions can still have their own reasons for moving that have nothing to do with the index. Worth remembering the difference between "my position is down because the market's down" and "my position is down for its own reason" before reacting to either one.
SPY closed Wednesday at $762.44, down $3.52 (-0.46%) from Tuesday's $765.96 and now sitting about 2% below the $777.88 high the index set August 13 — a slide that started well before the opening bell, once oil began moving on renewed fighting between the U.S. and Iran.
Brent crude topped $99 a barrel Wednesday, its highest level in close to seven weeks, after the U.S. military said it struck several Iranian oil tankers near the Kharg Island export hub and Iran claimed retaliatory strikes on U.S. vessels and tankers in the Gulf. That combination — a real supply threat layered on an already-tight oil market — is what was moving equities today more than any single U.S. economic data point; Goldman Sachs flagged that continued shipping attacks raise the odds of Brent clearing $120.
The index's -0.46% number understates how today actually felt in the names I hold in my Trading account. SoFi (SOFI) fell 3.7% to $17.35 and Quantum Computing (QUBT) dropped 2.2% to $8.04, both moving harder than SPY on a day with no company-specific news behind either move — the kind of gap that shows up when an oil-driven risk-off day hits higher-beta names harder than the broad index.
SPCX had a more mechanical reason to be down: roughly 319 million previously locked-up shares became eligible for sale Wednesday, the third tranche in SpaceX's staggered post-IPO unlock schedule, worth close to $49 billion at Tuesday's closing price. Shares closed down 4.3% at $146.83, a similar-sized drop to the one the stock took around its August 20 unlock — a pattern that reversed within days last time, though nothing about today's move guarantees it repeats.
LULU is the position that actually moved on a decision of mine today. Shares gave back yesterday's new-CEO bounce, closing down 3.3% at $99.78 after Citigroup and Daiwa both cut their price targets this morning (to $117 and $105, respectively) — and I used the pullback to add 45 shares at $98.49 around 9:40 a.m. ET in my Trading account, bringing the position to 100 shares total. Logging that as the decision I made in the moment, not as a read on where the stock goes from here.
The calendar gets busier from here: PPI prints Thursday morning and the August CPI report lands before Friday's open, both now one day closer and both landing on a market that's already jumpy about inflation because of what oil did today.
Why this matters: a red index day driven by a geopolitical/oil shock reads differently than one driven by weak earnings or a rate scare — a reminder that not every selloff is about the companies themselves. Worth watching whether Thursday's PPI and Friday's CPI land into a market already nursing an oil-driven risk-off mood, or whether today's move fades the way the last two SpaceX unlocks did.
Newest entries appear first — nothing gets deleted or quietly edited after the fact, same discipline as the episode recaps. Each entry keeps its own original publish date regardless of position on the page.