A busy week of big tech earnings will set the tone for stocks, and the whole story turns on one question: will the huge sums poured into AI actually pay off? That is the key thing to watch, along with what these companies say about their capex (money spent on buildings and equipment) and how AI is being used in their businesses.
Spending is up, and the market is nervous
Last week Google gave a positive signal by spending more, and it raised its outlook to possibly over $200 billion. Alphabet is spending because it has a backlog and real demand. Still, the market was not thrilled, because rising spending strains cash flow, and that is now being watched closely. The broad read is good: AI moves forward, the buildout continues, and that helps a range of stocks tied to it.
Sentiment right now is fairly negative because attention sits on how much these firms are spending rather than how much they are earning. There is fear that overspending could cause a collapse, echoing the late 1990s and early 2000s, when telecom firms laid too much fiber optic cable. At some point the buildout will be too much, but the engine is still running. The street should focus on the positive side, which is strong earnings.
Earnings momentum is still strong
The earnings trend remains positive and is still speeding up. This makes seven quarters of double-digit earnings growth. Second quarter earnings can be counted on for more than 20% growth, the second straight quarter at that level, which is three to four times the normal trend.
More telling: 137 companies in the S&P 500 have already reported, and half are raising their estimates. That share is higher than usual and better than three months ago, when they reported first quarter results. The rate at which those estimates are climbing is also higher. Rising outlooks like this are not what shows up at market tops. So the selling looks like fear, and any sell-offs should give long-term investors a chance to buy the dips.
What matters is the outlook, not the past quarter
For Meta, the test is proving AI is helping the business, whether through better advertising performance or more user engagement. The improvement must be larger than the cost of building the infrastructure. That is what the street wants to know, and it will show up in the earnings revisions and the guidance the company gives, not in what Meta reports about the three months that just ended in June. Forward guidance is the most important thing, because it answers whether the spending is worth it.
The main worry is that the hyperscalers (the giant cloud and AI firms) will not earn a good return on the capital they invested. That is a fair fear. But as long as revenues stay strong and earnings expectations keep improving, it is only a fear and not reality yet.
Beyond AI: power, the grid, and reshoring
Almost all the focus has landed on the AI buildout, which has meant neglecting two other real trends: reshoring (companies moving production back to the US) keeps going, and demand for power and electricity keeps rising. The US is expected to need 20% more electricity by the end of this decade. That points to infrastructure names tied to electrification, the grid buildout, utilities, and renewables that feed AI data centers. Many contractors have filled their backlogs with AI work, yet still have room to grow in other areas, which is the key point.
Crosscurrents to watch
Knee-jerk reactions are a concern. Google and Tesla saw selling, but then good news came from SAP, and a new tech IPO in Asia lifted optimism. Rising bond yields add pressure too. The two-year yield is climbing, and chip stocks are at elevated levels. The SMH chip ETF is up 55% year to date, with names like KLA, Lam, and Qualcomm in focus. Qualcomm reports this week.
Between all these earnings and the Fed meeting this week, there are many unknowns. For now it is a wait-and-see moment, and much more will be clear by the end of the week.
