Finance · Daily Deep-Dive

One Earnings Season, Two Kinds of Fear — Alphabet and Tesla's Capex Collided With a Red Sea Tanker Attack

2026-07-24 · US Equities · AI Capex · Middle East Oil

On Thursday, July 23, the Dow fell 506.93 points (-0.97%), the S&P 500 dropped 1.21% to 7,408.30, and the Nasdaq sank 2.15% to 25,137.69 — briefly slipping below 25,000 for the first time since May. What broke the market wasn't one story. It was two, largely unrelated, landing on the same day.

What Happened

Alphabet's Q2 numbers weren't the problem. Revenue grew 24% year over year, cloud revenue surged 82%. What spooked investors was the guidance: Alphabet raised its 2026 capex forecast to $195-205 billion, after a record $44.9 billion in quarterly capex that left the company with -$5.9 billion in free cash flow for the quarter. Shares fell as much as 7% after hours before settling roughly 5% lower once trading opened.

Tesla's reaction was sharper. Q2 revenue came in at $28.24 billion (vs. $26.32 billion expected, +26% year over year), but adjusted EPS was $0.33 against a $0.50 estimate, adjusted EBITDA missed at $3.2 billion versus $4 billion expected, and operating margin collapsed to 1.4% from 4.1% a year earlier. CFO Vaibhav Taneja confirmed full-year capex would top $25 billion — quarterly capex had already jumped 142% year over year to $5.79 billion — and the company lined up $30 billion in new debt capacity to fund the robotaxi, Optimus, and AI data-center buildout. Musk called 2026 a "massive capex year." Tesla stock closed down 14.5% Thursday to $319.69, its lowest close since August 2025.

At the same time: Iran-aligned Houthi rebels struck two Saudi oil tankers in the Red Sea, roughly 10 ships turned back, and daily traffic through the Bab el-Mandeb Strait fell 34% that day. Brent crude crossed $100 a barrel for the first time in about two months. The 10-year Treasury yield rose to 4.71%, its highest level since January 2025, and CME FedWatch pricing swung to over 80% odds of a September rate hike.

Why: The Mechanism

Mechanism 1 · Capex stopped being a story and became a line item. For over a year, investors were willing to treat AI capex as an article of faith — the more a company spent, the more credible its AI story looked. This time, the spending showed up where it actually matters: Tesla's operating margin got cut by nearly two-thirds, and Alphabet posted negative free cash flow for the quarter. The market wasn't punishing spending itself. It was punishing the first real evidence of what that spending costs.

Mechanism 2 · An oil shock with nothing to do with AI landed on the same day. The Houthi tanker attack has no causal link to how fast Alphabet's cloud business grows. But Brent above $100 and a 10-year yield at 4.71% raise the discount rate applied to every equity — and hit hardest the stocks whose valuations depend most on distant future cash flows, which is exactly the profile of high-capex tech. Two unrelated shocks pulled the same valuations down from two different directions at once.

Mechanism 3 · Strong labor data worked against the market, not for it. Weekly jobless claims fell to 187,000 (vs. 210,000 expected), among the lowest readings in decades. Normally that's good news. Layered onto an oil-driven inflation scare and rising rate-hike odds, though, resilient employment became evidence the Fed has no reason to ease — good news read as a reason for tighter policy, not looser.

Numbers & Second-Order Effects

MetricNumber
Dow / S&P / Nasdaq (7/23)-506.93pts (-0.97%) / -1.21% to 7,408.30 / -2.15% to 25,137.69
Tesla stock-14.5% to $319.69, lowest since Aug 2025
Alphabet stock-7% after hours, settled ~-5%
Alphabet 2026 capex guidance$195-205B (record $44.9B in quarter)
Alphabet quarterly free cash flow-$5.9B
Tesla Q2 capex$5.79B, +142% year over year
Tesla FY2026 capex guidance>$25B + $30B new debt capacity
Tesla operating margin1.4%, down from 4.1% a year ago
Tesla free cash flow-$1.09B (vs. -$3.64B expected)
Combined 2026 capex, big four~$725B, +77% vs. ~$410B in 2025
Brent crudecrossed $100/barrel, first time in ~2 months
Red Sea attack impact~10 ships turned back, Bab el-Mandeb traffic -34%
10-year Treasury yield4.71%, highest since Jan 2025
September hike odds (CME FedWatch)>80%
Weekly jobless claims187,000 (vs. 210,000 expected), decades-low

Second-Order Effects: Who Wins, Who Loses

Bulls' vs. Bears' Strongest Case

Bull · The Demand Is Real

Alphabet's underlying numbers are clean — 24% revenue growth and 82% cloud growth are real demand, not a story. KKM Financial's Jeff Kilburg called the capex fear "short-term misguided," arguing the spending eventually converts to profit. Tesla's actual cash burn (-$1.09 billion) came in far better than the -$3.64 billion Wall Street feared — the cash reality was better than the headline suggested.

Bear · The Scale Is Out of Hand

Combined 2026 capex across the big four cloud spenders — Amazon (~$200B), Microsoft (~$190B), Alphabet ($195-205B), and Meta ($115-135B) — is tracking toward roughly $725 billion, up 77% from about $410 billion in 2025. Spending at that scale needs years of near-certain payoff to justify current valuations, and Tesla's margin compression to 1.4% shows the "spend now, profit later" thesis already eroding real profitability, not just theoretical risk. Layer on an oil shock and rising hike odds, and the cost of capital is rising at the same moment — a double squeeze, not background noise.

Critical Thinking: What the Narrative Misses

It's tempting to flatten this into a single "AI bubble cracking" headline, but that label conflates two very different things. The first is a legitimate question about the quality of AI capex itself — Tesla's margin numbers and Alphabet's negative free cash flow are verifiable, real signals worth taking seriously. The second is a geopolitical supply shock that has nothing to do with AI — Houthi tanker attacks don't change based on how many billions any tech company spends on data centers. Bundling both into one "risk-off" selloff makes for a tidier headline, but it obscures a more useful question: if your worry is AI return on investment, the move is reassessing capex-heavy holdings; if your worry is oil-driven rate hikes, the move is checking duration and rate-sensitive exposure. Those are two different trades, and doing both at once, driven by one blended headline, tends to do neither well. Gold and Bitcoin both fell the same day too — a detail that points toward a rising-real-yield repricing, not a classic flight-to-safety panic.

One Takeaway

Back on July 4, we wrote about weak jobs data reading as good news for markets, because it meant faster rate cuts. This is close to a mirror image: strong jobs data, in a backdrop of oil-driven inflation fear, read as bad news, because it removed an excuse to ease. The same economic indicator can carry opposite meaning depending on what else is happening that day. Next time a headline reads "jobs beat expectations, stocks fall anyway," don't assume the market is being irrational — check whether a separate, independent shock is quietly redefining what "good" and "bad" mean that day.

Sources

This is informational analysis, not investment advice.