For most of 2026, equity markets have behaved as if geopolitics and monetary policy were background noise to one overwhelming theme: artificial intelligence. The Philadelphia semiconductor index is up roughly 83% this year, on pace for its best year since 1999. The S&P 500, despite a soft month, still sits about 18% above where it was a year ago. Tech stocks surged over 80% in the first half of the year before wobbling.

That serenity is now being tested from three directions at once. This is an attempt to lay out, factually, what is happening and why these threads are connected.

Thread one: a war that markets can no longer ignore

The conflict between the United States and Iran has escalated sharply through July. As of this week, US Central Command has completed its ninth consecutive night of strikes on Iranian military targets, while Iran has launched retaliatory strikes against US positions in Bahrain, Jordan, Kuwait, Oman, Qatar and Syria. At least three American service members have been killed, and Houthi forces in Yemen have declared a maritime embargo on Saudi Arabia, deepening the Red Sea shipping crisis.

The economic transmission channel is oil. Commercial traffic through the Strait of Hormuz — the chokepoint for roughly a fifth of global oil supply — has largely ground to a halt, with vessel owners pausing transit attempts and the International Maritime Organization urging shippers not to risk the passage. Brent crude briefly broke $90 per barrel in Monday's Asian trading before settling near $88–89, up around 30% from its July lows and roughly 20% for the month. Earlier in the year, during the March escalation, Brent spiked above $109; the current rally is smaller but the direction of risk is the same.

Prices eased somewhat after Iran's Foreign Ministry confirmed it had received proposals from international mediators and signalled that negotiations with the US could continue if they served Iran's national interests. Markets are now trading headline-by-headline: escalation adds a war premium, any diplomatic signal removes part of it.

Thread two: inflation is back, and central banks have turned

The oil shock is not staying in the energy market. Eurozone inflation climbed to 3.2%, its highest reading since 2023, driven by energy and services — and the European Central Bank responded by raising its deposit rate by 25 basis points to 2.25%, its first hike since 2023 and a decisive pivot back toward tightening. The EU's economy chief has warned that if Brent were to hold around $100 with elevated gas prices, bloc-wide inflation could exceed 3% for the year and shave up to 0.4 percentage points off already-modest growth.

The United States faces a harder version of the same problem. Core inflation was running at 4.2% as of the May reading — more than double the Federal Reserve's 2% target — before the latest oil surge. Goldman Sachs has modelled that a return to $100 oil would add roughly 3–4 basis points to monthly core inflation in the coming months. And the Fed itself is under new management: Chairman Kevin Warsh, speaking at the ECB's Sintra forum, declined to signal the July decision but was blunt that "prices are too high". The institution that spent 2024–25 easing is now, at minimum, on hold — and possibly worse for equities.

This matters for stocks through a simple mechanism: higher rates raise the discount applied to future earnings, and no part of the market has more of its value parked in the distant future than AI infrastructure.

Thread three: the $700 billion question reports earnings this week

Which brings us to the third storyline. The four largest hyperscalers — Microsoft, Alphabet, Amazon and Meta — are now expected to spend over $700 billion on capital expenditure in 2026, most of it on AI data centers, up from earlier projections that were themselves considered aggressive. Estimates for Alphabet alone run to $175–185 billion this year.

The strain is showing in the cash flows. Alphabet's free cash flow is projected to fall about 67% this year to roughly $21 billion as capex consumes its operating cash. By contrast, Apple — which has conspicuously declined to join the infrastructure arms race — is expected to generate a record $140 billion in free cash flow, and its 16% gain this year makes it the best performer among the Magnificent Seven, partly because traders fleeing the AI selloff have treated it as a haven. Microsoft, meanwhile, is down about 20% in 2026, on pace for its worst year since 2022.

Earlier this month the doubt turned into a rout: semiconductor stocks shed over $1.3 trillion in market value as Wall Street questioned whether record AI capital spending can generate matching returns, with Intel, Micron and AMD under particular pressure. Notably, analysts characterized this not as a demand problem but as a valuation and returns problem — dot-com-era multiples meeting a hawkish Fed. Many still frame the selloff as a "mid-cycle reset" rather than the end of the cycle, and JPMorgan expects semiconductors to find a bid again given continued earnings growth.

This week delivers the first hard evidence. Alphabet and Tesla report after Wednesday's close, with Intel following Thursday, in what analysts are calling the most comprehensive single-week data point yet on whether the AI spending cycle is producing real returns. Intel's report carries its own drama: the stock has rallied more than 163% in 2026 on conviction that its manufacturing turnaround — anchored by the frontier 18A process node — is real rather than aspirational. Alphabet enters with its own tailwind, climbing after reports that Google is developing a Gemini-integrated server chip to ease AI compute constraints.

The signal to watch is capex guidance, and it cuts both ways. Raised guidance is bullish for the AI supply chain (Nvidia, chipmakers, data center REITs, utilities) but signals ever-greater strain on the spenders' own cash flows. Cut guidance relieves the spenders but implies the revenue isn't arriving fast enough to justify the buildout — a chilling message for everything downstream.

How the threads knot together

Here is the uncomfortable arithmetic. The AI trade is a bet on enormous profits arriving years from now, financed by unprecedented spending today. That bet is most fragile precisely when interest rates rise — and rates are rising because of inflation, and inflation is rising because of oil, and oil is rising because of a war whose trajectory nobody can forecast. A single diplomatic breakthrough in Tehran would ease all three pressures simultaneously; a strike on Iranian energy infrastructure, or a genuine closure of Hormuz, would tighten all three at once.

There is also a historical echo that professional investors are openly discussing. Strategists have noted that 2026's price action closely resembles the second half of 1999 — a market that kept climbing through elevated volatility, punctuated by sharp pullbacks, before the eventual reckoning. The comparison is suggestive, not predictive: unlike 1999, today's mega-caps are enormously profitable, and the S&P 500 trades at its lowest forward price/earnings-to-growth ratio since 1995 — meaning that if the earnings actually arrive, current prices are defensible. The entire debate reduces to whether they arrive.

What to watch next

Three dates and one variable. Alphabet and Tesla report Wednesday, July 22; Intel reports Thursday; the Federal Reserve meets in the final week of July for Warsh's first fully-owned rate decision amid a war-driven inflation impulse. The variable running through all of it is the Iran conflict — specifically whether mediator proposals turn into actual negotiations, and whether the Strait of Hormuz reopens to normal traffic.

For long-term investors, the practical takeaway is narrower than the drama suggests: this week won't settle whether AI transforms the economy — it will settle whether the market's timeline for that transformation was priced correctly. Those are very different questions, and conflating them is how bubbles inflate and how panics overshoot.

Note: Market figures are as of July 20–21, 2026 and move constantly; verify current prices before relying on any number here. Nothing in this article is investment advice.