The equity market has been on a remarkable run, stringing together nine consecutive days of gains and nine straight weeks of advances while setting fresh records along the way. That momentum is not built on sentiment alone. The foundation is genuine: companies tied to the artificial intelligence and data center theme keep posting real earnings, positive cash flows, and tangible improvements to their bottom lines. As long as that remains true, AI and the infrastructure behind it should continue to drive the market through the remainder of the year. But beneath the headline strength lies a question that will define the next chapter of this cycle.
The Two Phases of the AI Trade
It helps to think of the AI boom in two distinct phases. The first phase has been about building infrastructure — the chips, the servers, the power generation, the cooling systems, and the data sets that make large-scale computing possible. This is the phase the market has been living in, and it keeps revealing new legs. After the initial chip frenzy, attention rotated into the server makers, producing sharp run-ups in the hardware names that supply that layer of the stack. Each time one segment matures, another emerges to carry the trade forward.
The second phase, however, is the one that matters most and has yet to fully arrive: monetization. So far, nearly all of the money being made is being made by the inputs feeding into giant data centers, not by the outputs those data centers produce. The pivotal moment will come when a company stands up and credibly says it is making real money on AI rather than simply selling shovels to the people digging the mine. Will it be a software company turning AI capability into revenue? A physical-goods manufacturer that has used AI to make a consumer or industrial product cheaper, better, or more efficient in a way that genuinely moves the economy? Honestly, it is hard to say which companies will capture that meaningful, durable revenue. That uncertainty is the central unresolved tension in the entire rally.
Software's Fall and Tentative Recovery
The software sector offers a useful case study in how quickly enthusiasm can turn. These companies enjoyed an early surge several months ago, only to fall off dramatically afterward — some of them crashing outright in what amounted to a kind of software apocalypse. The past week or so has brought a modest bounce back into a few of these beaten-down names, and that recovery deserves close attention.
The reason software is worth watching is that, of all the categories in the market, software companies are arguably the best positioned to actually monetize what AI produces. But they face an existential fork in the road. The defining question for each of them is whether they will be replaced by AI or whether they will learn to work alongside it. Names that once anchored portfolios but have since collapsed need to prove they can harness AI rather than be displaced by it. Even the historical core offerings of the largest, most established software franchises are not exempt from this question. The market is now demanding evidence of which side of that line each company will land on, and the answer is far from obvious.
Why Selected Hardware Names Remain Essential
While the software story is about adaptation and survival, certain hardware names look more like structural necessities. At some point the AI cohort will stratify — some companies will keep executing and thriving, while others will simply fail to deliver. Identifying which is which means examining individual companies rather than buying the theme wholesale.
A handful of chip and memory names stand out as essential. The leading semiconductor foundry is an unavoidable part of the supply chain; it manufactures the chips that the marquee designers produce, the silicon that powers the entire revolution. There is no realistic way around it. Memory, too, remains compelling — despite an incredible run, the dominant memory maker still looks cheap from a fundamental, historical perspective. These are the kinds of names that earn their place not on hype but on indispensability.
This Is Not the Dot-Com Bubble
It is tempting to draw parallels to the late-1990s technology mania, but the comparison does not hold. The decisive difference is earnings. Today's leading AI companies are demonstrably successful and are reporting concrete earnings improvements quarter after quarter — a stark contrast to the profitless speculation that defined the dot-com era. The valuations rest on real cash flows, not on promises. That said, the bubble framing does not vanish entirely; it simply migrates to the monetization question. Until someone proves out genuine profit from AI outputs, the market is extrapolating a future it has not yet seen delivered.
Expanding the Lens Beyond Technology
One of the more interesting opportunities lies in companies outside the technology sector that nonetheless benefit from AI and other powerful tailwinds. A low-cost domestic steel producer is a prime example. It stands to benefit from tariffs, and its use of electric arc furnaces means it requires far less fossil-fuel input while producing steel from recycled materials more cheaply than its competitors. Crucially, it also offers a turnkey solution for building the enormous data centers that AI demands — making it simultaneously an AI story and a story that stands on its own.
Defense and energy provide another avenue. A storied British engineering company is positioned to benefit from the necessary defense build-up taking shape across Europe, while also carrying a compelling energy angle through its work on small modular reactors — a meaningful advantage in an environment of rising energy prices and oil supply concerns. Energy, in turn, loops back to AI through the immense power requirements of data centers.
Geopolitical insulation matters here as well. An Italian oil company illustrates the point: it pipes the majority of its oil directly from the northern Caspian Sea in Kazakhstan and is pursuing new projects across Africa, keeping it outside the hottest geopolitical flashpoints and more insulated from the instability surrounding Venezuela and Iran. In a fractured world, the location of a company's supply matters as much as the commodity itself.
The Problem of Market Breadth
The concentration of the market is a genuine challenge. With the largest technology names dominating the index and the other 493 constituents trailing behind, allocating capital is genuinely difficult. The sharp earnings growth concentrated in the AI space makes those mega-cap stocks nearly impossible to ignore. Yet when a single name already represents close to 8% of the entire index, it becomes very hard to justify overweighting it further — the position is already enormous.
Navigating that tension requires selectivity across the rest of the market. Some sectors warrant caution: healthcare and real estate look unappealing in the current environment. Others remain attractive — utilities and energy still hold appeal — but even there the discipline is to hunt for the genuinely well-executing companies rather than buying entire sectors indiscriminately.
FOMO and the Risk of a Sharp Pullback
The market has climbed roughly 20% from its late-March low, advancing in what feels like a straight line. That has produced a great many impressive returns — including for plenty of investors who did little research and simply made a lucky guess on something they believed was connected to AI. The result is a tremendous amount of enthusiasm, and that is precisely what makes the moment fragile. When a market runs this hot, a single earnings miss or one unexpected headline anywhere in the world can erase the optimism in an instant. A pullback from these levels is entirely plausible.
The Macro Crosscurrents: Fed, Tariffs, Inflation, and War
Several macro forces are converging, and each carries weight. Monetary policy is shaping up to be a fascinating saga. There has been clear political pressure for lower interest rates, and a sympathetic appointee has been installed to that end — yet the market appears to be betting against rates falling. The tension between political intent and market expectation will be compelling to watch.
On trade, a proposal for a 10% tariff, possibly higher, on imports from some 60 countries echoes the disruptions of earlier in the cycle. Tariffs cut both ways — they can punish importers while protecting domestic producers like the low-cost steelmaker mentioned earlier.
Inflation remains the variable that touches everything. The painful lesson of 2022, when there was nowhere to hide because stocks and bonds both performed poorly, is not the current reality — but the risk has not disappeared. Energy is extraordinarily pervasive, and not only energy itself but the entire universe of chemicals and materials drawn from the ground. If conflict in the Middle East persists, additional inflationary pressure could follow. International bodies have already warned of a global slowdown, with growth potentially slipping toward 1.8% because of the Iran conflict, even as the inflation outlook has been revised higher — a worrying combination of slower growth and stickier prices.
The Scenarios That Could Spoil the Year
Looking toward year-end, two scenarios stand out as the most concerning. The first is geopolitical. The prevailing assumption — or at least the prevailing message — has been that the Iran conflict is a shorter-term issue. But a quick resolution is hard to envision. If the war is still active by the close of the year, that would be deeply troubling, and the market could very plausibly trade lower than it does today.
The second scenario is more paradoxical, and it strikes at the heart of the AI promise itself. As AI delivers real productivity gains, those gains may arrive hand in hand with layoffs. Major financial and technology firms have already signaled they intend to slow hiring and lean harder on AI instead. The danger is a corrosive mix: higher inflation on one side and lost jobs on the other. That combination — rising prices alongside shrinking employment — is the outcome most worth worrying about, and it is the shadow side of the very revolution currently powering the market to record highs.
Conclusion
The market's strength is real, anchored in genuine earnings rather than empty speculation, which is what separates this moment from past manias. But the rally now stands at an inflection point. The infrastructure build-out has been spectacularly profitable for the suppliers of inputs; the next and far harder test is whether anyone can convert AI into durable profit from its outputs. Until that proof arrives, investors should stay constructive but disciplined — favoring the indispensable hardware names, watching the software survivors closely, broadening beyond pure technology, and keeping a wary eye on the macro crosscurrents of war, inflation, and the very job losses that AI's success might bring.