Alphabet and Tesla reported earnings after the market closed on July 22, and both were punished in the same trading session for the same reason. Alphabet shares fell 7.13% to $317.69 on July 23 despite stronger-than-expected revenue and an 82% increase in Google Cloud sales. Tesla dropped 14.52% to $319.69 after its capital expenditure surged and free cash flow turned negative.
For nearly two years, markets frequently treated larger artificial intelligence budgets as evidence of future growth. The July 23 selloff suggests investors are beginning to apply a different test. Spending on data centres, chips, autonomous systems and computing infrastructure still supports the long-term AI thesis, but it is increasingly being judged against the cash those investments consume and the returns they are expected to generate.
The clearest comparison is in the annual spending plans. Alphabet now expects 2026 capital expenditure of between $195 billion and $205 billion, placing the midpoint near $200 billion. Tesla expects to spend more than $25 billion, close to three times its roughly $8.5 billion of capital expenditure in 2025. In both cases, investors focused less on the size of the opportunity and more on the immediate claim those programmes place on cash flow.
Alphabet and Tesla Fell in the Same Session
Alphabet produced results that would ordinarily support a higher stock price. Quarterly revenue reached $119.80 billion, exceeding the approximately $116.93 billion expected by analysts. Google Cloud revenue rose 82% year over year, providing further evidence that demand for AI computing, enterprise cloud services and related infrastructure remains strong.
The stock nevertheless fell more than 7%. The market reaction was driven by Alphabet’s decision to raise its full-year capital expenditure forecast to between $195 billion and $205 billion from a previous range of $180 billion to $190 billion. The revision added approximately $15 billion at the midpoint.
Alphabet spent $44.9 billion on capital expenditure during the quarter. That investment contributed to quarterly free cash flow of negative $5.9 billion, shifting attention away from the revenue beat and toward the financial cost of expanding AI capacity.
Tesla experienced an even larger decline during the same July 23 session. Its shares lost 14.52% and closed at $319.69 after the company reported its first negative free cash flow quarter in more than two years.
Tesla’s quarterly capital expenditure increased 142% year over year to approximately $5.8 billion. The company also guided investors to expect more than $25 billion of capital expenditure for the full year as it funds AI computing, autonomous driving, robotaxi development, robotics, battery production and manufacturing capacity.
The companies entered the quarter with different businesses and different financial profiles. The market nevertheless reached a similar conclusion on both: revenue growth and long-term AI ambitions did not fully offset concern about the speed at which investment was consuming cash.
The Capex Numbers Changed the Earnings Story
Alphabet’s earnings beat established that demand was not the immediate problem. Google Cloud was growing rapidly, total revenue exceeded expectations and the company remained positioned at the centre of the AI infrastructure buildout.
The issue was the amount of capital required to support that growth. At the midpoint, Alphabet’s revised forecast implies approximately $200 billion of capital expenditure in 2026. That is an infrastructure programme large enough to alter how investors evaluate the entire company, even when operating businesses continue expanding.
Tesla’s spending programme is smaller in absolute terms but larger relative to its recent history. A full-year budget above $25 billion would be close to three times its 2025 capital expenditure. Investors are therefore being asked to assign value not only to Tesla’s current vehicle business but also to projects whose financial returns depend on future adoption of autonomous transport, robotics and AI services.
Neither reaction proves that the investments are mistaken. Alphabet may need enormous computing capacity to protect its position in search, cloud computing and generative AI. Tesla may need to spend heavily before robotaxis and humanoid robots can become commercial businesses. The market is instead asking how long shareholders must fund those plans before the spending produces measurable returns.
Negative Free Cash Flow Was the Trigger
Capital expenditure does not reduce reported revenue, but it directly affects free cash flow. That distinction explains why Alphabet’s revenue beat failed to protect the stock.
Investors can tolerate rising expenditure when operating cash flow comfortably covers the investment and leaves significant cash available for buybacks, acquisitions or balance-sheet growth. The interpretation changes when capital expenditure pushes quarterly free cash flow below zero.
Alphabet’s negative $5.9 billion figure made the cost of its AI programme visible in a single number. Tesla’s first negative free cash flow quarter since early 2024 produced a similar warning. Both companies are investing in projects that may generate substantial future revenue, but both asked investors to accept weaker present cash generation in exchange for that possibility.
The selloff indicates that Wall Street may be placing a higher burden of proof on incremental AI spending. Announcing a larger data-centre programme is no longer automatically read as a bullish signal. Investors increasingly want evidence that earlier investment is improving margins, revenue growth and cash generation before rewarding the next increase.
The Read-Through for Nvidia and AI Infrastructure Stocks
The reaction matters for companies supplying the AI buildout. Nvidia, AMD, Broadcom, Arista Networks, Vertiv, Super Micro Computer and data-centre operators have benefited from expectations that hyperscalers will continue purchasing processors, networking systems, cooling equipment and power infrastructure at increasing rates.
Alphabet’s higher forecast is positive for suppliers because it confirms continued demand. The stock reaction is more complicated. Should investors continue punishing customers for the effect of AI spending on free cash flow, management teams may eventually face pressure to moderate investment or demonstrate stronger returns from existing capacity.
That creates a distinction between near-term supplier revenue and the valuation applied to that revenue. Semiconductor and infrastructure companies may continue reporting strong orders while their customers face greater scrutiny over the economics of those purchases.
The same question applies to quantum-computing and early-stage AI infrastructure companies whose valuations depend on spending expected years into the future. When markets become less willing to reward capital-intensive growth without near-term cash generation, companies with limited revenue and high infrastructure requirements can be repriced more aggressively than established suppliers.
FinanceFeeds previously examined the valuation debate surrounding Nvidia, including a $302 bull case and a $152 bear case. Alphabet and Tesla’s declines reinforce the central issue in that debate: AI demand may remain strong while investors become less willing to pay any price for exposure to it.
What Would Prove the AI Capex Thesis Wrong?
The remaining hyperscaler earnings reports will test whether July 23 represented a broader market shift or a reaction to company-specific cash-flow figures.
If the next group of large technology companies raises capital expenditure, reports weaker free cash flow and still sees its shares hold or rally, the claim that Wall Street has turned against AI spending would be too broad. Investors may simply have judged Alphabet and Tesla’s spending increases as excessive relative to their individual results.
If additional companies report strong revenue, increase AI budgets and suffer similar stock declines, the repricing thesis becomes harder to dismiss. That pattern would indicate that investors are applying a new condition to the AI trade: capital expenditure must produce visible operating leverage rather than only promise future market leadership.
Alphabet and Tesla did not fall because investors stopped believing in artificial intelligence. They fell because two of the largest listed companies showed, in the same earnings cycle and the same market session, how expensive the race has become. The next stage of the AI trade may therefore be determined less by who spends the most and more by who converts that spending into cash first.
