Bond market revolts against a hawkish Fed as Big Tech earnings split in two

Research and idea generation for personal use. Not investment advice. See full disclaimer at the bottom.

Top of mind

The Fed held rates steady Wednesday in a 9-3 vote — but the dissent (three officials wanted a hike) and Chair Kevin Warsh's "family fight" framing spooked the bond market more than the equity market. The 10-year Treasury yield jumped over 8 basis points to 4.70%, and the 30-year hit its highest level since 2007, as traders concluded the Fed is behind the curve on inflation ahead of this morning's GDP and core PCE releases. Stocks sold off hard Wednesday (Dow -2.19%, its worst day since April 2025) and CNBC reports the Treasury selloff is continuing into Thursday. Overnight, Big Tech earnings underscored the market's growing bifurcation: Microsoft jumped roughly 8% after hours on a broad beat, while Meta sank about 10% on an EPS miss tied to ballooning AI capex — with Apple and Amazon still to report tonight.

Market snapshot

(S&P, Nasdaq, Dow, 10Y, and VIX are Wednesday, July 29 confirmed US closes. Futures reflect Thursday, July 30 premarket. Sources: CNBC, Yahoo Finance, Washington Post, Bloomberg.)

Asset	Level	Change	Notes
S&P 500	7,316.15	-1.52%	Wed close; more than 10% off its intraday record
Nasdaq Composite	24,442.94	-1.74%	Wed close; chip-stock selloff was the main drag
Dow Jones	51,594.14	-2.19% (-1,153.18 pts)	Wed close; worst single-day drop since April 2025
10Y Treasury	~4.70%	+8 bps	Wed close; sharpest bond reaction to the Fed decision
30Y Treasury	>5.2%	+10+ bps	Wed close; highest level since 2007
VIX	20.66	+13.45%	Wed close; up from ~18 Tuesday but still shy of panic levels

Read-through: Wednesday was a genuine risk-off day, not a rotation day — every major index fell together, led by a nearly $1 trillion wipeout in chip-stock market value on AI-capex anxiety (Intel -6%, AMD -8%, Micron and Seagate both down more than 8%, SanDisk -14%) after disappointing results from Korea's SK Hynix. The more important signal, though, is in rates: long-dated Treasury yields did the heavy lifting on the selloff, which tells you the market's real complaint isn't the "hold" itself but the idea that holding now means a bigger catch-up move later. Thursday's premarket futures (Dow +0.07%, S&P +0.21%, Nasdaq 100 +0.50%) point to a modest bounce, but a bounce built on Microsoft's earnings beat rather than any resolution of the rates story.

Headlines & analysis

1. Fed holds rates steady in a 9-3 vote; bond market reads it as hawkish

Source: CNBC, CNN So what: Three regional Fed presidents (Hammack, Kashkari, Logan) dissented in favor of a hike, and Warsh raised the committee's year-end inflation expectations while flagging a possible rate hike later in 2026. Equity markets fell on the news, but the bond market's reaction was sharper and more durable — the 10Y and 30Y both sold off hard, with the 30Y hitting a 2007-era high. That's the market pricing a Fed that may already be behind, not one that just bought itself time.

2. Microsoft blows past estimates; Meta misses on earnings despite a revenue beat

Source: CNBC, company earnings releases So what: Microsoft posted fiscal Q4 EPS of $4.74 (vs. $4.24 expected) on revenue of $90.0 billion (vs. $87.6 billion expected, +18% YoY), with Azure revenue topping $100 billion for the full fiscal year and 30 million paid Copilot users — real, monetized AI demand, and shares jumped about 8% after hours. Meta's revenue beat ($60.8B vs. $60.17B expected) but EPS missed badly ($6.18 vs. ~$7.17-7.22 expected, roughly 14% below consensus), and next-quarter revenue guidance ($62.5B) also came in soft versus the $63.15B estimate — shares fell about 10% after hours as investors balked at the spending required to sustain Meta's AI buildout.

3. Chip stocks shed roughly $1 trillion in value on AI-capex jitters

Source: CNBC So what: A weak report from Korea's SK Hynix reignited fears that AI infrastructure spending is running ahead of demand, dragging down Nvidia, AMD, Intel, Micron, Western Digital, and SanDisk over the past two sessions. This is the same anxiety that's been building since Monday and is now colliding directly with tonight's Amazon and Apple prints, both of which carry their own capex-guidance stakes.

4. GDP and core PCE both due at 8:30am ET today

Source: Investing.com, Axios, BEA So what: Economists expect Q2 GDP to show the economy grew at a 1.8% annualized pace (down from 2.1% in Q1) alongside core PCE — the Fed's preferred inflation gauge — rising 0.2% month-over-month, with the year-over-year rate ticking down slightly to 3.3% from May's three-year high of 3.4%. Given how the bond market just reacted to the Fed's own inflation framing, a hotter-than-expected PCE print this morning could extend Wednesday's yield spike rather than calm it.

5. Apple and Amazon report after tonight's close

Source: AppleInsider, TipRanks, company guidance So what: Apple is expected to post Q3 revenue near $108.9 billion (+16% YoY) and EPS of $1.89 — notably, this is expected to be Tim Cook's last earnings call as CEO before COO John Ternus takes over September 1. Amazon consensus sits near $196 billion in revenue and $1.82 EPS, with AWS growth and 2026 capex guidance the focus. Both reports land directly on top of this week's already-heated debate over whether AI spending is being matched by real monetization — the same question that just split Microsoft and Meta in opposite directions.

Ideas — long-term core

Quality businesses, durable competitive advantages, reasonable valuation. Hold horizon: years.

MSFT — Microsoft

• Thesis: Wednesday's print is the clearest evidence yet in this earnings season that a hyperscaler's AI capex is translating into real, monetized growth — Azure crossed $100 billion in annual revenue with 43% quarterly growth, and Copilot has scaled to 30 million paid seats. That's a meaningfully different data point than the market's broader AI-capex anxiety would suggest.
• Valuation note: Shares were already priced for strong execution before the print; the ~8% after-hours pop pushes the multiple further ahead of the sector median, so this isn't a name to chase into strength.
• Why now (or why patient): Patient accumulation on any post-earnings-pop pullback looks more sensible than chasing the gap — especially with a still-volatile rates backdrop that could pressure high-multiple tech broadly in the near term regardless of company-specific execution.
• Risks / bear case: Capital intensity remains enormous, and any deceleration in Azure growth or Copilot monetization would quickly reopen the same "spending without payoff" narrative currently punishing Meta. Rising long-term rates are also a headwind for any long-duration growth multiple.

Ideas — opportunistic

Catalyst-driven, time-bound, sized smaller. Hold horizon: days to months. Define exit before entry.

AMZN / AAPL — earnings reaction tonight

• Catalyst: Both report after today's close, directly on top of a GDP/PCE print this morning and Wednesday's Microsoft/Meta split — the market has just shown it will reward real AI monetization (Microsoft) and punish capex-without-payoff optics (Meta), and Amazon's AWS growth plus capex guidance will be read through exactly that lens.
• Time horizon: Overnight into Friday's open.
• What would invalidate: Soft AWS growth or a capex guide that raises spending without a clearer monetization story would likely extend Wednesday's AI-anxiety theme; Apple's setup is complicated further by the Cook-to-Ternus CEO transition being confirmed on the same call, adding a leadership-continuity variable atop the usual iPhone/services/China questions.
• Risk note: Binary, headline-driven event risk for both names on the same night — size for a large potential overnight gap in either direction, not for conviction on the outcome.

META — post-earnings dislocation

• Catalyst: A ~10% after-hours drop on an EPS miss (despite a revenue beat) is a large single-session repricing for a name with a still-strong core advertising business; whether that gap holds, extends, or partially retraces over the next few sessions is the tradeable question.
• Time horizon: Days to a few weeks — watch how the stock trades once the initial earnings-reaction volatility settles.
• What would invalidate: Further analyst downgrades citing the capex trajectory, or confirmation that ad-revenue growth (rather than just spending) is decelerating, would argue the gap is a repricing rather than an overreaction. A stabilization above the post-earnings low without new negative commentary would be the more constructive signal.
• Risk note: This is a bet on market overreaction to a spending story that the market has just shown, via Meta itself, it's willing to punish hard — size small and define an exit level before entering.

Portfolio-level guidance

Allocation and risk observations. Not specific buy/sell calls — those depend on a full picture this report doesn't see.

• Concentration check: Four mega-cap tech earnings reports in three trading days (Microsoft and Meta already out, Apple and Amazon tonight) is a concentrated risk event for any portfolio overweight large-cap tech or the broader AI-infrastructure trade — Wednesday and tonight together are effectively one continuous single-week risk event, not routine earnings.

• Rates positioning: The 30-year Treasury yield's move to its highest level since 2007 is the more important story than Wednesday's equity selloff — it reflects a market view that the Fed may be under-hiking relative to where inflation is heading. This is not the moment to extend duration on the expectation of rate cuts; if anything, the bond market's message this week has consistently leaned the other way.

• Cash & dry powder: Between this morning's GDP/PCE data, tonight's Apple and Amazon reports, and a bond market that's still actively repricing the Fed's reaction function, staying reactive to confirmed outcomes — rather than pre-positioning around any single expected result — remains the better risk-adjusted posture through the rest of this week.

• Risk regime read: VIX closed Wednesday at 20.66, up over 13% but still well below levels seen in genuine panic episodes — the equity market treated Wednesday as a real but orderly repricing, not a capitulation. The bond market's reaction was proportionally larger, which is worth weighing more heavily than the VIX print alone when sizing risk this week.

Watch list — tomorrow / this week

Earnings: Apple and Amazon report Thursday, July 30 after the close — Apple's call (5pm ET) will also mark Tim Cook's last as CEO before John Ternus takes over September 1.

Economic data: Q2 GDP (advance estimate) and June core PCE both release Thursday, July 30 at 8:30am ET; consensus looks for GDP growth of about 1.8% annualized and core PCE up 0.2% month-over-month (3.3% year-over-year). The next major data point is the July jobs report (nonfarm payrolls), due Friday, August 7 at 8:30am ET.

Fed / central bank: No scheduled Fed speakers immediately following Wednesday's decision, but watch for follow-up commentary on the September meeting given three dissents favored an immediate hike and Warsh flagged rising year-end inflation expectations.

Other: The US-Iran situation remains fragile — a ceasefire announced in recent days was followed by Wednesday's intercepted Iranian missile attack on US forces and a joint US-Saudi response in Iraq. This conflict has reversed direction multiple times in recent months; any further escalation or de-escalation headline could move oil and rates sentiment quickly in either direction.

Disclaimer

This report is prepared for personal research and informational purposes only. It does not constitute investment advice, an offer, or a solicitation to buy or sell any security. Information is drawn from public sources believed to be reliable but is not guaranteed accurate or complete. Markets change rapidly; data may be stale by the time of reading. Any "ideas" mentioned are research candidates, not recommendations, and do not consider any specific person's financial situation, objectives, or risk tolerance. Consult a licensed financial advisor before making investment decisions. Past performance does not predict future results.