AI is burning cash—turn in your homework tonight: Alphabet reveals $190 billion in capital expenditures, and Tesla faces extreme volatility

Alphabet and Tesla will release their 2026 Q2 earnings reports after the close of US stock trading, simultaneously with Taiwan time in the early hours of July 23. Intel will follow the next day. The options market has already positioned itself, pricing in a post–Tesla earnings one-day swing of 5.91%. That translates to a market value of about $83 billion, which is the largest expected earnings-related volatility for this company over the past year. What the market truly wants to know is not how much these two companies will earn, but what this year’s $700 billion in AI capital expenditures will actually deliver in return.
(Background: Google has rolled out three new models, including Gemini 3.6 Flash! 3.5 Flash-Lite has rocketed to extreme speed, and Gemini 4 pre-training has already started.)
(Background detail: Citigroup says, “the US stock seven giants are dead,” and shifts its bet to growth companies spanning six major industries.)

Table of Contents

Toggle

• Alphabet’s problem: where is the money going?
• Tesla sets a record for delivery innovations
• The three numbers to watch tonight

Key Takeaways

• The options market is pricing in a 5.91% swing after Tesla’s earnings, with roughly $83 billion in market value hanging in the balance
• Alphabet’s capital expenditures this year could top $190 billion, but Q1 free cash flow fell 47% year over year to just $10.1 billion
• Tesla Q2 deliveries hit 480,126 units, setting a record high; the market is only watching whether the ex-carbon-credit auto gross margin can hold at 18%

The options market’s pricing has already said everything. According to Benzinga, implied volatility before Tesla’s earnings reached 5.91%. Converting that using a $1.4 trillion market cap equals about $83 billion in market value being placed on tonight’s earnings report—by absolute dollar amount, the second-largest among all companies reporting earnings this week.

CNBC also noted that this is the largest earnings-driven volatility expectation traders have set for Tesla in a year. The last time they priced in a 6% move was in October 2025.

Alphabet, by comparison, is more restrained: implied volatility is about 5.4% for a one-day move, but its market cap is $4.3 trillion—any percentage point is still worth several hundred billions. After Intel’s after-hours release on July 23, implied volatility climbs as high as 12%. With three earnings reports squeezed into the same week, analysts described them as the most complete one-week test of the AI spending cycle so far.

Neither company’s positioning going into the market looks great. Alphabet closed last Friday at $346.77, down about 15% from this year’s high of $408.61. Tesla is down 18% year-to-date, the worst performer among the seven giants, and still about 25% away from its 52-week high of $498.83. Even before earnings start, the stock has already shown you what’s coming.

Alphabet’s problem: where is the money going?

Market consensus for Alphabet’s second-quarter revenue is between $113.6 billion and $116.8 billion (different data providers use different definitions), with earnings per share of $2.87, versus $96.4 billion and $2.31 in the same period last year. The results themselves don’t look bad.

What looks bad is the cash flow picture. In the previous quarter, Alphabet raised its full-year capital expenditure guidance to $180 billion to $190 billion, and told investors that 2027 will jump up again by a large step. In Q1, capital expenditures rose 107% year over year to $35.7 billion. At the same time, free cash flow fell about 47% year over year to just $10.1 billion. Spending more than twice as much as last year but ending up with less than half as much on hand—this “gap” is the question everyone is asking tonight.

The story behind the money is cloud. Google Cloud revenue rose 63% year over year in Q1, and backlog nearly doubled quarter over quarter to more than $460 billion. The market expects second-quarter cloud revenue to grow another roughly 65% to $22.5 billion, while search business revenue would be about $63.29 billion. Backlog is a beautiful number, but backlog isn’t the same as booked revenue. What to watch tonight is whether the pace at which these $460 billion in backlog turns into revenue can keep up with the pace of depreciation.

Another problem inserted into Gemini: at Google’s May I/O developer conference, it said Gemini 3.5 Pro went live in June, but by mid-July it still had not appeared. Bloomberg reported that Google updated its training data to improve Gemini’s code capabilities, but the outcome was not as expected; and what enterprise customers care about most is exactly this part.

Yesterday, Blockworks reported that Google simultaneously released three models, including Gemini 3.6 Flash, and that Gemini 4 has also entered pre-training. Three releases in the Flash series back-to-back, but the flagship Pro didn’t show up—by itself, that’s an answer.

Analysts didn’t collectively turn bearish. Guggenheim’s Michael Morris maintained a Buy rating and a $450 price target, treating the Gemini delay as an “entry point for all-end AI leadership” rather than a fundamental issue; BMO Capital raised its target to $455 and KeyBanc raised it to $445. But UBS cut its target to $400 and Wells Fargo lowered it to $416. Almost all the disagreement between bulls and bears boils down to the same sentence: is this $190 billion buying a moat, or just depreciation?

Tesla sets a record for delivery innovations

Tesla’s second-quarter deliveries were 480,126 vehicles, its best quarter in company history. The number had already been published in early July, but the issue is that this didn’t translate into the stock price.

The consensus numbers are highly divergent. On July 17, Tesla published its self-compiled analyst consensus on its investor relations page: revenue of $27.58 billion; non-GAAP earnings per share of $0.55; US GAAP earnings per share of $0.36; and net income of $1.5 billion. But versions circulating in the market range from $25.24 billion to $26.4 billion in revenue, and earnings per share range from $0.44 to $0.55. In the same company, for the same quarter, sellers differ by more than a couple dozen billion dollars in revenue—this kind of divergence is itself a source of volatility.

The real battleground is gross margin. Analysts expect the ex-carbon-credit automotive gross margin to be slightly above 18%. Whether this line can be held matters far more than selling a few more tens of thousands of vehicles. Cash is also tight. The market estimates Tesla’s Q2 free cash flow will be negative $3.25 billion, with capital expenditures nearing $6.7 billion—money invested into AI infrastructure buildout, capacity expansion, and Robotaxi deployments.

Record deliveries, yet Tesla is the one that has fallen the worst among the seven giants.

Morgan Stanley’s Adam Jonas describes Tesla as a stock that is “fully priced in front.” He maintains a Neutral rating and a $417 price target. Put into plain language: the current stock price has already factored in the success of Robotaxi and Optimus—without actual deployment, there’s no new reason for upside.

But the actual deployment progress is behind. Robotaxi currently operates across seven metro areas: Austin, the Bay Area, Dallas, Houston, Miami, Orlando, and Tampa. Of the originally planned list of seven cities for the first half, Phoenix and Las Vegas are not yet in place. Optimus, the humanoid robot, was originally planned to start small-scale production at the end of July or in August, but Musk issued a preemptive warning: each unit has about 10,000 unique parts, and early production ramp-up would be extremely slow; and by mid-July, that production line had still not started running.

On the other end of the spectrum, Wedbush’s Dan Ives expects Robotaxi to expand to 30-plus cities, and calls for Tesla’s market cap to reach $2 trillion to $3 trillion. But this goal is clearly conditional: Optimus commercialization must land by the end of 2026.

The three numbers to watch tonight

What tonight really tests is whether the roughly $700 billion to $725 billion in AI capital expenditures that global large technology companies spend in 2026 has started to become money. Alphabet’s capital expenditure scale has already expanded more than fivefold since 2023.

• Alphabet capital expenditure guidance: whether the $180 billion to $190 billion range will be raised again, and the pace at which free cash flow continues to bleed
• Google Cloud growth: the market expects about 65% year-over-year growth, while also looking at the revenue conversion cadence of $460 billion in backlog
• Tesla automotive gross margin: whether it can hold 18% ex-carbon credits, and the specific timelines for Robotaxi and Optimus

Worth noting is that market patience is being repriced. Citigroup only last week called the “US stock seven giants” dead, arguing to shift to growth companies spanning six industries—because when all the giants are spending the same money on the same track, does excess returns still exist?

Common Questions

When will the 2026 Q2 earnings reports for Alphabet and Tesla be released?

Both companies release after the close of US stock trading on July 22 Eastern Time—that is, in the early hours of July 23 Taiwan time. Intel will follow with its after-hours release on July 23. The options market is pricing Alphabet’s volatility at 5.4%, Tesla’s at 5.91%, and Intel’s at as high as 12%.

Why are these two earnings reports viewed as a stress test for AI capital expenditures?

Alphabet’s capital expenditures are expected to reach up to $190 billion, but in Q1 free cash flow fell 47% year over year to $10.1 billion. The market wants to confirm whether this money turns into actual revenue for Google Cloud and Gemini—not just depreciation expense.