The Nasdaq 100 has just closed out its strongest quarter in six years, gaining roughly 20% over the three-month period. Understanding what produced that performance requires separating three distinct pillars that support the market and shape its movements. The most important of these, and the one that ultimately matters most in the long run, is fundamentals — specifically, earnings. Everything else complicates the picture, but over any meaningful horizon it is strong earnings that drive market performance.
Pillar One: AI Fundamentals and Earnings
The first and dominant pillar is the AI trade and the earnings and valuation premiums it generates. This is where the most important story lives, and it comes with significant sector bifurcation in performance.
Pillar Two: The US–Iran War
The second pillar is the US–Iran war, which has been receding into the background. Following the memorandum of understanding signed on June 18th, the market appears to be reacting far less to the tweets and messaging surrounding the conflict. Its fading influence is a welcome development for market stability.
Pillar Three: The Macroeconomic Picture
The third pillar is macroeconomic theory and the lingering aftermath of the war — its effects on oil prices, on inflation, and the uncertainty introduced by the prospect of a new Fed chair. The macro picture always complicates matters, but when everything is stripped back, market performance still comes down to fundamentals and earnings.
Are We Overpaying for Growth, or Still Early?
A natural question is whether investors are still in the early innings of the AI-driven bull market, or whether they are now paying too much for growth. The answer is that it is not really a matter of overpaying for growth. Instead, what is happening is a rotation and a sharpening of focus onto who is actually making money in AI.
This reflects a maturation of the AI trade. Very early on, anyone doing anything in AI was rewarded indiscriminately. That has shifted decisively toward a "show me the money, show me results" mentality. Strong performance has increasingly concentrated among the recipients of the enormous AI capital-expenditure spending program — the chipmakers, the semiconductor companies, and the memory companies. Names like Micron, Nvidia, and AMD have done extremely well. The semiconductor index is up about 63% within the Nasdaq 100, a remarkably strong result.
By contrast, the market has taken a more skeptical view of the hyperscalers — companies such as Microsoft and Meta that are the ones spending all the money on AI. Estimates point to roughly $700 billion of AI CapEx spending by 2026. The market is more cautious on the spending side of this equation, while consistently rewarding those on the receiving side — not only for the earnings they have already booked, where orders have been phenomenal, but also for their strong forward-looking forecasts.
The Ping-Pong Rotations Within Technology
On any given trading day, particularly the first day of a new quarter, there can be a huge rotation beneath the surface — out of the long-crowded trades and into the hated, unloved, underweighted parts of the market. Even within technology itself, rapid-fire rotations between hardware and semiconductors resemble an Olympic ping-pong match.
The market is still trying to get a handle on these moves. The earnings for these companies are genuinely there; the open question is simply what the correct valuation on them should be. A sell-off in some of the best-performing names on the first day of a new quarter is not surprising. Part of it may be window dressing — fund managers wanting their quarter-end and month-end statements to show that they owned all the biggest performers such as Micron, AMD, Nvidia, and Intel, then lightening up on a short-term basis once the quarter has closed. But the fundamentals remain very strong. Micron, for instance, recently reported earnings that blew away estimates and raised forward guidance. So while some of the movement is short-term positioning, the longer-term trade remains very much intact.
The Startling Gap Between Semiconductors and Software
One of the most striking statistics from this quarter is the enormous gap between semiconductor and software performance. Breaking down the Nasdaq 100's roughly 20% quarterly gain reveals the mechanics: bucketing all the semiconductor and hardware companies together, they added about 23 percentage points to performance. Bucketing the software companies together, they detracted about three percentage points. Netting those out leaves the index up about 20% for the quarter.
That performance differential is extraordinary — it is difficult to identify the last time a gap of that magnitude appeared. The software companies are the spenders right now, which is why the market is treating them with more skepticism. This is actually a healthy dynamic: rather than rewarding everyone involved in AI, the market is asking "who is winning right now?" The clear winners are the recipients of AI CapEx, the companies building out the infrastructure.
When Will Software Capture More AI Value?
There will come a time when software begins to capture more of the AI value currently flowing to the semiconductor names, though it is impossible to say exactly when. That turn will come as the return on investment becomes visible. Around 25% of S&P 500 companies are already starting to report meaningful contributions from AI. As those effects take hold and investors can properly evaluate the benefits of AI, software should begin to pick up again. For now, however, the market is simply rewarding those bringing in more tangible orders.
Dispersion and the "Neural Nine"
A useful reframing of the current market is the "neural nine" — the mag seven plus Broadcom and Micron — and the notable dispersion appearing among these names. For a long time the narrative was "mag seven versus everybody else," but that is not what the past quarter, or even the year, has actually shown. The mag seven have generally underperformed the rest of the Nasdaq 100. Microsoft and Nvidia have done very well, but several of the other mega-cap names have not kept pace.
The more accurate framing, then, is software versus semis rather than mag seven versus the other 93 constituents. Evidence for this comes from the Nasdaq equal-weighted index, which has kept up well with the cap-weighted Nasdaq 100 — up close to 20% for the quarter and about 20% year-to-date. The equal-weighted index performing so strongly signals a broadening out into other semiconductor names. This is a thematic trade rather than a mag-seven concentration trade, which marks a meaningful shift from the pattern seen last year.
The Valuation Paradox: Rising Prices, Falling Multiples
A particularly interesting point concerns valuation. Looking at the Nasdaq 100 over the last 18 months — from the end of December 2024 through the end of June — it is up somewhere around 44% on a cumulative basis. Yet over that same stretch, the forward price-to-earnings ratio has actually contracted by about 8%.
In other words, multiples have been shrinking even as prices have surged. This is possible because earnings have been growing so fast, and forecasted earnings continue to be ratcheted up, that the denominator is outpacing the strong gains in price. Far from signaling a stretched, overheated market, this pattern suggests that the rally is being underwritten by genuine and accelerating earnings growth rather than by pure multiple expansion — an observation echoed across a number of mid-year market outlooks.
