Cool CPI, record Wall Street earnings, and IBM's worst trading day in its 115-year history. Published by @dailyanalysts, close of trading.
The single most important thing that happened today wasn't the market's +0.4% gain — it was IBM erasing $68 billion of market cap in one session (-25.2%, its worst day on record, worse than Black Monday 1987) because enterprise clients diverted IT budgets toward memory chips and servers ahead of expected price hikes. That is a real-time, real-money confirmation of the AI-hardware-crowds-out-software thesis that's been building all month. My stance: bullish on AI hardware/memory, cautiously bullish on the broader tape, moderate conviction. Cool June CPI (3.5% YoY vs. 3.8% expected) bought the Fed room to stay on hold, but Iran-driven oil risk and a hawkish Chair Warsh mean this relief is fragile, not a green light to chase everything.
| Index | Close | Change | % Change |
|---|---|---|---|
| S&P 500 | 7,543.59 | +28.25 | +0.4% |
| Dow Jones Industrial Average | 52,508.27 | +9.63 | +0.02% |
| Nasdaq Composite | 26,107.01 | +233.83 | +0.9% |
| Russell 2000 | 2,964.76 | +11.60 | +0.4% |
The Dow's flat close hides the real story: IBM alone sliced roughly 425 Dow points off the average on its own — without it, the Dow would have looked like a genuine tech-led rally day, similar to the S&P and Nasdaq.
| Sector | ETF | % Change | Driver |
|---|---|---|---|
| Best: Technology | XLK | +1.29% | Chip/memory rebound (SOX had dropped 15-16% from its June peak) + a cybersecurity spike |
| Energy | XLE | +0.37% | Oil holding above $79 on Iran/Hormuz risk |
| Financials | XLF | +0.20% | Record bank trading and IB revenue (mixed stock reactions though) |
| Worst: Healthcare | XLV | -1.93% | No single catalyst — rotation out of defensives into risk-on tech/chips |
| Consumer Staples | XLP | -1.38% | Same rotation-out-of-defensives dynamic |
| Real Estate | XLRE | -0.49% | Rate-sensitive; still digesting last week's yield spike |
The Cboe Volatility Index fell to 16.41 (-4.4%). In plain terms: professional traders are paying less to insure their portfolios against a big swing over the next 30 days, which means near-term fear eased. But 16.41 is still elevated versus the mid-teens "calm" range seen earlier this year, and the underlying SOX (semiconductor index) volatility just hit its highest since 2020 — the "calm" headline number is masking real turbulence underneath in individual AI-linked names.
The 10-year Treasury yield eased to roughly 4.59%, down modestly after touching 4.63% pre-CPI. Falling yields on cool inflation is normally an "all clear" signal for stock valuations, particularly rate-sensitive growth and small caps. But the move was modest, not a rout — because traders still see a ~60% chance of a Fed hike by September (down from higher odds a day ago, but not gone) given the Iran-war oil risk. Consequence that matters: a 10-year push back above 4.65% would start compressing AI-multiple stocks again; a durable move back under 4.40% would be the real all-clear for the small-cap/cyclical rotation trade.
S&P 500 7,470-7,415. That's the support zone the index has bounced from twice in the last week during the Iran-oil-shock scare. As long as the S&P holds above ~7,415 on a closing basis, this remains an orderly pullback within an uptrend (S&P is still +10.2% YTD). A decisive close below that zone would confirm the market is pricing in a real oil-driven inflation re-acceleration, not just a one-day scare.
Catalyst #1 — Cooler-than-expected June CPI. Headline CPI fell 0.4% month-over-month (biggest drop since April 2020), pulling the annual rate down to 3.5% from 4.2% in May, versus 3.8% expected. Core CPI was flat month-over-month (2.6% annual vs. 2.9% expected). (CNBC) The move was driven almost entirely by a 5.7% monthly plunge in energy prices tied to the (now-collapsing) Iran ceasefire — meaning this is a lagged, possibly one-month read that oil's renewed spike toward $80-85 could reverse within a print or two.
Catalyst #2 — Megabank earnings. Five of the largest US banks (JPMorgan, Bank of America, Citigroup, Wells Fargo, Goldman Sachs) all reported the same morning for the first time in years, and all five beat estimates, powered by an 86% surge in JPMorgan equities-trading revenue and a 55% jump in Goldman's investment-banking fees — both directly tied to AI-driven capital-markets activity (IPOs like SpaceX, AI infrastructure financing). (CNBC live blog)
Narrative challenged: "inflation is beaten, cuts are coming." Fed Chair Kevin Warsh explicitly rejected the "mission accomplished" framing the same morning, telling Congress the inflation fight isn't over, and futures still price a ~60% chance of a hike by September. (CNBC) The bond market only partially believed the good CPI news — yields fell just a few basis points, not the kind of move you'd see if traders thought the Fed was truly done.
Narrative strengthened: the AI-hardware-crowds-out-software thesis. IBM's own CEO put numbers on it: clients shifted their June capex toward servers, storage and memory "to secure supply-constrained infrastructure ahead of expected price increases," and it cost IBM $660 million in revenue it didn't see coming. (CNBC) This is the same memory-supercycle thesis behind our standing MU (Micron) buy call — and today it got confirmation from an unlikely source: the company being hurt by it.
What most investors are overlooking: buried inside the IBM story is a genuinely bullish read for one software subsector. Krishna told CNBC that Anthropic's "Mythos" AI-cybersecurity model is causing some enterprise deals to pause while clients reassess how much they need to spend on cyber defense — and cybersecurity stocks (CrowdStrike +12.1%, Palo Alto Networks +6.8%, Cloudflare +4.5%) ripped higher on that comment. The market read this correctly for security names but is still bundling ALL enterprise software (ServiceNow, Salesforce, Adobe, Workday, all down 2-6% today) into one "AI is eating software" bucket — when the actual evidence points to a bifurcation, not a collapse.
Real-world implication: memory and server component price hikes triggered by this capex rush will eventually show up as input-cost inflation for PC, server, and consumer-electronics makers in 2027 — meaning today's "good" CPI print is partly a temporary energy-driven illusion sitting on top of a hardware inflation wave that hasn't hit consumer prices yet.
Lucid Motors (LCID) -16.2% to $4.62 on an (denied) report it's weighing bankruptcy or going private. Whether or not the report is accurate, the swiftness of the drop signals just how little benefit-of-the-doubt capital-burning EV names get right now, in a market that is aggressively funneling capital toward AI hardware and away from anything resembling the last cycle's cash-burning growth story. That's a trend worth watching across the rest of the EV/legacy-growth complex.
| Idea | Type | Entry | Target | Invalidation | Timeframe | Conviction |
|---|---|---|---|---|---|---|
| Buy the ServiceNow (NOW) contagion overreaction | Contrarian | $102-107 | $112-115 | Daily close <$100 | 1-2 weeks | SPECULATIVE |
| Add to Micron (MU) on any pullback | Trend-following | $930-965 | T1 $1,200 / T2 $1,500 | Weekly close <$840 | 1-3mo / 6-12mo | HIGH CONVICTION |
| Healthcare (XLV) defensive add | Defensive | $155-160 | $172 | Weekly close <$150 | 1-3 months | WATCH |
1. Tomorrow's actionable idea (short-term traders): buy ServiceNow's dip. It fell right into the exact "overreaction bounce" zone flagged earlier today, on pure sympathy selling with no company-specific evidence of AI disruption to its actual business. Rationale: markets frequently punish innocent bystanders before sorting out which companies actually have the problem — that gap is the trade.
2. Contrarian move (long-term investors): keep buying the memory/AI-hardware complex (MU, and by extension the broader memory trade) even after today's rip. The crowd's instinct after a name is already up double digits year-to-date is to assume it's "priced in." Today IBM handed the memory bulls unsolicited, real-dollar proof that the supply/demand imbalance is worse than modeled — that's not a reason to fade the trade, it's a reason to stay in it.
3. Defensive position (long-term investors): scale into Healthcare (XLV) on today's -1.9% drop. There was no healthcare-specific bad news today — it was rotation out of defensives into risk-on tech. That kind of "no news" sector drawdown, in a sector trading near 52-week lows relative to the market, is exactly the setup for a defensive hedge that costs you little in an up market and protects you if the Iran/oil risk resurfaces.
Bull (30%): June CPI cooling proves durable, Iran/Hormuz de-escalates, Brent falls back under $75, the Fed skips both July and September hikes. S&P 500 grinds to new highs above 7,700; small caps (Russell 2000) lead as rate-cut hopes resurface; memory/chip names extend gains on confirmed AI capex durability.
Base (50%): Choppy, two-speed market continues — AI hardware/memory and Wall Street trading desks keep winning, legacy enterprise software and defensives keep lagging, oil stays elevated but doesn't spike further. S&P chops in the 7,400-7,650 range through Q3 earnings season.
Bear (20%): Iran conflict escalates further, Hormuz transit is meaningfully disrupted, Brent spikes through $90-100. The Fed is forced toward a September hike (or earlier), the 10-year yield breaks back above 4.65%, and the "cool CPI" narrative is fully reversed by August/September prints — AI-multiple compression resumes and small caps/cyclicals give back H1's rotation gains.
Our standing Micron (MU) HIGH CONVICTION buy call ($900-955 entry) is working — up another 4.9% today to $983.12, and today's IBM warning is direct, unsolicited confirmation of the underlying thesis (enterprise capex diverting to memory ahead of price hikes). Our ServiceNow bounce call, opened earlier today after the IBM-contagion selloff, saw NOW settle at $104.85 — right in the planned entry zone — with the trade still live into tomorrow.