Wall Street ended high on Thursday, January 15, 2026, after another star performance by tech stocks. Nasdaq Composite regained 0.86% from Wednesday’s slump. And as it has been tradition for the past few years, the broad market gains when tech stocks rise; the S&P 500 gained 0.80%.
Since last year, the artificial intelligence (AI) infrastructure boom has been a key driver of the tech rally that carried the broad market. This won’t change in 2026, although more catalysts are coming into the picture, including robust Q4 2025 earnings. But, as we will discuss, the latter ties into the AI supercycle, which is just another way of saying that AI’s dominance as the key catalyst of the current and future tech rally isn’t about to end any time soon.
AI Supercycle in Semiconductors and Infrastructure
In December last year, Goldman Sachs projected that capital expenditure by US-based AI companies could reach $700 billion this year if aligned with past tech investment cycles; they gave the example of the late 1990s telecom boom. The bank noted that consensus estimates for AI hyperscaler capex in 2026 had risen to $527 billion, up from $465 billion earlier in 2025. To justify their projections, Goldman Sachs argued that Wall Street analysts have consistently underestimated AI-related spending; actual growth has exceeded forecasts for two years running.
Globally, Gartner expects AI spending to touch $2.52 trillion this year. This will be a 44% year-on-year increase as enterprises shift from pilots to large-scale AI deployments. And, AI infrastructure will dominate this capex with about $1.37 trillion. Gartner noted that GPU-rich data centers and AI-optimized servers will gobble up much of this spending.
And this infrastructure push is already being validated by robust earnings in key sectors. For instance, TSMC, a company that produces the chips powering the AI boom, released blowout Q4 2025 earnings yesterday, January 15. Revenue came in at $33.73 billion, surpassing estimates and up 25.5% YoY and 1.9% quarter over quarter. The company explained that 77% of the income came from advanced chips, under 7nm, tied to AI.
According to Alan Lancz, president of Alan B. Lancz & Associates Inc., TSMC’s blockbuster performance is a huge psychological boost for tech-focused investors. Lancz noted that before the earnings, “there was some worry as far as valuations – that they were getting a little too far ahead of themselves.” But with TSMC’s Q4 2025 earnings, that notion has “been kind of squashed.”
Earnings Resilience in Tech Stocks
TSMC told investors it expects to earn around $35.8 billion in revenue in the next quarter, and that the gross margin could range from 63-65%. And, the company expects its capex for 2026 to be no lower than $52 billion.
Interestingly, positive guidance is the trend right now among S&P 500 tech stocks. According to FactSet, 107 S&P 500 companies issued EPS guidance for Q4 2025, and of these, 50 (47%) were positive. This is well above the 5-year average of 42% and the 10-year average of 40%. FactSet noted that this quarter marks the second-highest percentage of S&P 500 companies issuing positive EPS guidance since Q3 2021.
One of the implications from this reality is that the tech companies have incredibly high confidence in Q4 performance. And, they are quite certain about the resilience of their margins because demand for their products and services, especially those linked to AI, is strong. This also validates Lancz’s perspective that worries about valuations are misplaced.
Risks and Near-Term Considerations
Be that as it may, some worry that tech stocks’ stretched valuations is an actual risk. The expensive valuations of US tech stocks is fueling the belief that for once, after several years, the leadership of the S&P 500 will broaden.
Angelo Kourkafas, senior global investment strategist at Edward Jones, told Reuters that 2026 is the year where the market might see “some true broadening of leadership.” He added: “The conditions are likely in place for that broadening to happen, especially when you sprinkle in and consider elevated valuations, there are some pockets of value that can be found looking beyond technology.”
For Larry McDonald, founder of The Bear Traps Report, 2026 is likely the year capital rotates out of AI and semiconductors into value sectors. His reasoning is that AI’s success depends on massive energy and materials investment. For example, over 850 data centers are projected in the US over five years, and this requires huge power infrastructure.
And for AI growth to be sustainable, McDonald argues that energy and commodity stocks must double or triple. But risks such as Not In My Backyard (NIMBY) opposition could block about 200 data centers and seriously impact AI expansion.
In sum, the AI trade still is the main driver of tech stock gains, at least in the near future. One doesn’t see how this will stop being the case even further ahead, given the amount of capex going into the sector.
