H1 2026 in two words: Iran and AI

Half-year market reviews are usually exercises in naming a single dominant theme. The first six months of 2026 needed two, and Nikkei Asia's half-year snapshot, published 30 June 2026 UTC, names them plainly: Iran and AI. Each pulled the global tape in a different direction — one through the price of crude and the premium on Gulf risk, the other through the capex decisions of a handful of US-listed hyperscalers and their Asian supply chain. Together, they explain almost every outlier on the half-year leaderboard, on both the upside and the downside.

The point is not that geopolitics and artificial intelligence are unusual drivers of market behaviour. It is that, in the first half of 2026, neither behaved the way consensus expected at the start of the year. The Iran file did not de-escalate on the timetable Western desks had pencilled in. The AI file did not roll over on the timetable bears had pencilled in. That double non-event — two narratives failing to resolve on schedule — is the story of the half.

What the tape actually showed

Nikkei Asia's snapshot frames 2026's first half as a market defined by two gravitational pulls. The first is the war footing around the Strait of Hormuz and the broader Gulf, where Iran's regional posture and the diplomatic track around its nuclear file kept crude bid and kept defence, shipping, and Gulf-exposed insurers volatile. The second is the AI capex cycle, centred on a small number of US-listed chip designers and cloud platforms whose order books run through Taiwanese and Korean fabricators, with secondary exposure in Japanese lithography and Malaysian packaging. The two forces do not cancel each other out; they compound. Higher oil feeds inflation, which feeds rate expectations, which feeds the discount applied to long-duration AI capex.

The clearest expression of this is in the equity market's regional dispersion. Energy and defence names with Gulf exposure outperformed. Pure-play AI semiconductor names outperformed, but with widening single-stock dispersion — the leaders extended, the laggards in the same sub-sector lagged further. Traditional cyclicals in markets exposed to Chinese demand underperformed, in part because Chinese consumer recovery remained uneven and in part because the same Gulf risk premium that lifted crude weighed on import bills across South and Southeast Asia.

The Iran leg — why the timeline slipped

The Western consensus entering 2026 was that the diplomatic track on Iran's nuclear file would produce some form of framework arrangement within the first half. That did not happen. Through Q1 and into Q2, the public reporting — both from Western wires and from regional outlets such as Middle East Eye — kept the file in a state of managed tension rather than resolution. Iranian players, whether diplomats or sporting representatives travelling through compressed schedules with visa uncertainty, became a recurring visual shorthand for the friction. That shorthand is not just colour; it is a price signal. Every week the file stays open, the option premium on a Hormuz disruption stays in the market.

The structural point is that the Gulf risk premium is now a permanent line item in trading-floor risk models rather than a tail event. That changes how capital is allocated. Sovereign wealth flows out of pure Gulf equity beta into logistics, refining capacity outside the Strait, and LNG infrastructure with flexible routing. Insurance markets reprice. Even non-energy sectors — Gulf tourism, regional banks with concentrated sovereign exposure — trade differently when the baseline assumption is persistent tension rather than episodic crisis.

The AI leg — why the cycle did not break

The other half of the Nikkei frame is the refusal of the AI capex cycle to roll over. Entering 2026, the bear case was straightforward: the hyperscalers had over-ordered on AI infrastructure, the cloud revenue curve could not absorb the depreciation, and 2026 would be the year the order book corrected. The first half did not deliver that correction. Capex guidance from the leading US cloud platforms was maintained or raised; the Asian supply chain — foundry, HBM memory, advanced packaging, lithography — ran hot; and the equity tape continued to separate the names with credible AI revenue exposure from those whose AI positioning was narrative rather than operational.

What changed inside the cycle, however, was the shape of dispersion. The first half did not lift all AI-exposed names equally. It lifted a narrow set of names with credible forward order book, while names whose AI exposure was retrospective — companies that had announced AI strategies without corresponding capex or revenue conversion — were marked down as the market learned to discriminate. That is the more durable signal of the half: not that AI kept working, but that the market started to price AI scepticism into the laggards while leaving the leaders largely undisturbed.

Where the two files meet

The reason Nikkei names both is that they share a single underlying condition: a world in which the major policy questions — energy security, technology control, the price of money — are being settled in real time rather than deferred. When both files are open, capital does not behave the way it behaves when one or both are closed. The discount rate on long-duration capex rises when oil risk rises, because both feed inflation expectations. The discount applied to Gulf-exposed earnings falls when oil risk rises, because the same earnings become more defensible. These are not separate stories; they are two outputs of the same underlying volatility regime.

The structural frame, in plain terms, is this: 2026's first half was the half in which the global market stopped expecting either the Iran file or the AI capex cycle to resolve quickly, and started pricing both as persistent rather than transitional. That is a regime change in expectations, even if the underlying flows have not changed as dramatically as the multiple expansion on either side suggests.

What the second half is now priced for

If both files remain open, the second half of 2026 looks like a continuation of the first: continued energy and defence outperformance, continued narrow leadership in AI-exposed semiconductors, continued pressure on cyclicals exposed to Chinese demand and on emerging-market importers of energy. If one file resolves — a credible diplomatic track on Iran that lasts more than a news cycle, or an AI capex guidance cut from a hyperscaler large enough to reset the consensus — the dispersion inverts. The Iran resolution would compress energy, expand rate-sensitive duration, and pull capital back into the AI laggards that had been marked down on discount-rate grounds. The AI guidance cut would do the opposite: compress the AI leaders, expand the laggards on relative basis, and redirect flows back into defensives and energy.

The honest reading is that the market enters the second half priced for persistence, not resolution. That is a more fragile setup than it looks, because it means the asymmetry is in the resolution scenarios, not the persistence scenario. Nikkei's choice to name only Iran and AI is, in that sense, a description of where the asymmetry sits.

What remains uncertain

The half-year snapshot is clean about what drove the tape and opaque about what did not move it. Two questions are still open. The first is whether the diplomatic track on Iran can produce a framework durable enough to compress the Gulf risk premium rather than simply rotating it; public reporting through the first half did not give a clear read on that. The second is whether AI capex can sustain its current pace without a cloud-revenue inflection that has, to date, lagged the order book; the half's equity action suggests the market believes it can, but belief is not the same as confirmation. Until one of those resolves, the two-word frame holds.

This publication framed the half through the two drivers the wire named, rather than treating either as background — because in the first six months of 2026, neither was.

Wire provenance

This editorial synthesis draws on the following public wire/social posts:

• https://t.me/NikkeiAsia
• https://t.me/nikkeiasia
• https://t.me/NikkeiAsia
• https://t.me/nikkeiasia