Graticule Levinson: Iran Outcomes and AI Hardware Rally

Adam Levinson, chief investment officer at Graticule Asset Management Asia, laid out two interconnected market threads on Bloomberg’s Insight on Apr 24, 2026: the likely trajectories for Iran after renewed regional tensions and the accelerating AI hardware cycle he describes as "unequivocally the greatest" in modern computing. Levinson’s comments intersect geopolitical risk and technology capital allocation, key inputs for institutional portfolio strategies; the conversation flagged both near-term dislocations and durable secular demand for compute capacity. The Bloomberg segment (Bloomberg, Apr 24, 2026) comes as AI-driven demand ripples through semiconductor supply chains, lifting equities such as Nvidia and equipment suppliers like ASML. These dynamics test conventional risk premia, requiring investors to calibrate between short-term volatility tied to geopolitics and longer-duration earnings upgrades from AI infrastructure spending.

Context

Levinson’s remarks arrived against a backdrop of intensified focus on Iranian outcomes following a string of events in 2025–2026 that elevated tail-risk pricing in energy and defence sectors. Iranian crude export flows since 2022 have been uneven as sanctions, insurance costs and re-routing have constrained volumes; supply concerns pushed Brent volatility higher in episodic spikes through 2024–25 (source: Bloomberg commodity flows, 2024–25). For institutional investors the practical issue is not a single binary outcome in Tehran but the probability-weighted set of scenarios: diplomatic de-escalation with sanctions relief, continued proxy-based asymmetric escalation, or episodic attacks that temporarily disrupt shipping and risk premia.

On the technology front, Levinson’s claim that the market is in a historic hardware cycle parallels observable capital spending trends: the headline-grabbing market cap expansion of AI leaders is one visible metric — Nvidia exceeded a US$1 trillion market capitalisation in May 2023 (Bloomberg, May 2023) — but the cycle extends to memory, foundry investment, and the EUV lithography machines that underpin advanced node scaling. Equipment suppliers and foundries are reporting multi-year order books; SEMI and industry filings indicated semiconductor equipment billings surged after the 2020 trough, supporting capex that has materially re-rated supply-side capacity (SEMI, industry reports 2021–2025). For policymakers the relevant linkage is that elevated demand for compute capacity increases strategic sensitivities — governments are more likely to treat semiconductor supply chains as national security assets rather than pure commercial ecosystems.

Levinson’s dual focus neatly captures a central risk management task for allocators: isolate idiosyncratic event risk (Iran-linked disruptions) from secular demand trends (AI compute). The former can create transient valuation dislocations; the latter can compress horizons and justify higher current multiples if earnings upgrades are persistent. Institutional investors should therefore differentiate instruments that will reflect short-duration repricing (commodity-linked equities, shipping insurers) from instruments whose value accrues over multi-year capacity expansions (equipment vendors, foundry capex beneficiaries).

Data Deep Dive

Three data points help quantify the two-legged theme Levinson outlined. First, the Bloomberg interview itself was broadcast Apr 24, 2026 and frames the timing of the market reaction referenced in this note (Bloomberg, Apr 24, 2026). Second, Nvidia’s market capitalisation milestone — surpassing US$1 trillion in May 2023 — remains a useful reference for the scale of investor reallocation into AI equities in recent years (Bloomberg, May 2023). Third, semiconductor equipment billings, which are a proximate measure of industry capex, rebounded strongly after 2020; industry tallies showed equipment billings roughly doubled from the 2019 low into the 2021–2022 surge (SEMI, industry data 2019–2022).

Those numbers illustrate two mechanisms. The market-cap milestone is a demand-side signal: investors have repriced future earnings to reflect AI-driven TAM expansion. Equipment billings are the supply-side counterpoint: foundries and suppliers are contracting multi-year revenues to expand wafer starts and advanced-node capacity. Together these create an earnings trajectory where hardware revenue and margins can rise materially if adoption follows existing enterprise and cloud provider procurement plans. The data also show that capital intensity is shifting risk onto balance sheets: extended lead times for EUV tools and memory modules mean that revenue recognition is delayed but order books firm, embedding investment cyclicality for equipment vendors.

A comparative lens is useful: hardware demand growth is outperforming overall IT spend. IDC and similar forecasters have repeatedly upgraded AI infrastructure spending estimates; across cycles, AI-dedicated infrastructure has grown faster year-on-year than aggregate IT budgets, with some estimates indicating compound annual growth rates in the high single to low double digits through 2026 (IDC, selected forecasts 2024–2026). That compares with typical IT budgets that expand at mid-single-digit rates, illustrating why hardware companies tied to AI capture outsized revenue growth versus peers in legacy enterprise hardware markets.

Sector Implications

Semiconductor equipment and foundry-related equities are the most direct beneficiaries of a hardware cycle that Levinson emphasised. ASML, the Dutch supplier of EUV lithography systems, occupies a critical choke point for advanced nodes; backlog dynamics for high-end machines can create multi-year revenue visibility for the supplier. Meanwhile, fabless designers and GPU providers such as Nvidia capture demand downstream via sales to cloud providers and hyper-scale customers. The valuation mechanics differ: equipment suppliers trade on multi-year order visibility and capital intensity, whereas GPU vendors trade on margin expansion and platform monetisation.

Energy and defence sectors are sensitive to the Iran narrative Levinson discussed, but impact magnitude is scenario-dependent. A short-lived escalation that raises Brent by a few dollars per barrel for a week is primarily a headline risk with limited equity-market persistence; a protracted disruption that removes material barrels from global seaborne trade would re-rate energy stocks and shipping insurers materially. Investors should quantify exposures: for example, a sustained 1 million barrels per day (mb/d) supply shortfall historically translates into double-digit percentage moves in the oil price if inventories remain tight (IEA analyses, historical shocks 2011–2022).

Financials and FX also react differently across scenarios. In short-duration spike scenarios, safe-haven assets and CDS spreads drive regional bank risk premia; in broader geopolitical realignments, trade finance and sanctions compliance impose structural costs for institutional investors. These differences inform hedging choices: tactical hedges for transient risk versus structural portfolio tilts for multi-year regime shifts.

Risk Assessment

The two principal risks to Levinson’s juxtaposition are timing and the path-dependence of geopolitical events. On timing, hardware cycles are lumpy: multi-year capex commitments can undershoot demand if generative adoption stalls or if software efficiencies reduce commodity compute needs. Conversely, demand could overshoot if new generative AI applications rapidly commercialise beyond current expectations. This timing uncertainty widens dispersion in forward earnings estimates and increases alpha opportunity for active managers.

On geopolitical path-dependence, Iran-related outcomes are not binary; markets price probabilities across a continuum. A negotiated settlement with sanctions relief would compress energy risk premia but could increase liquidity in certain Middle Eastern equities. A prolonged period of asymmetric attacks or state-on-state escalation would raise insurance costs and increase structural energy prices, with ripple effects into inflation and central bank policy. For fixed-income investors, such a scenario could steepen real-yield curves through commodity-driven inflation surprises.

Operational risks include supply-chain concentration. Advanced lithography and certain specialty chemicals are concentrated among a handful of suppliers; any export controls, sanctions, or targeted cyber incidents could create step-function supply constraints. That concentration is both the economic moat for suppliers and the single point of failure for the broader AI hardware expansion.

Outlook

Levinson’s recommendation to "buy the dip" for hardware names presumes persistent secular demand that will lift multi-year earnings. If AI adoption proceeds along current enterprise and cloud procurement plans, equipment and foundry suppliers will enjoy multi-year visibility that supports higher multiples. However, investors should calibrate entry points based on order-book digestion — early-cycle multiples tend to be volatile as markets reconcile backlog fills with near-term revenue recognition.

A prudent institutional construct blends exposure across the value chain: selective positions in established GPU and AI platform leaders, complemented by thematic exposure to equipment suppliers with backed-up order books, and hedges against short-duration geopolitical spikes. Risk management should explicitly account for lead times: orders placed today for EUV and advanced packaging capacity can take 12–36 months to translate into incremental wafers and revenue.

Where timing matters most is in active rotation between cyclical hardware names and software/service providers. As is often the case in technology accelerations, profits accrue both to the stack's top (software, platforms) and the bottom (hardware, equipment). Allocators must therefore maintain flexibility to rotate when hardware orders convert into revenue and when software monetisation curvature becomes evident.

Fazen Markets Perspective

Our contrarian read diverges from the pure buy-the-dip narrative in one respect: not all hardware-exposed companies deserve equal conviction. The greatest hardware cycle claim is credible in aggregate, but dispersion will be substantial. Firms with deep balance-sheet capacity to fund multi-year capex cycles and those with integrated supply-chain control will compound returns; by contrast, smaller, highly levered manufacturers without secured long-term contracts are vulnerable in a downcycle. We therefore advocate a selection approach that emphasises balance-sheet strength and contractual visibility over headline thematic exposure.

We also see a non-obvious risk: policy fragmentation. Export controls, subsidy wars and national industrial policies could reshape comparative advantage faster than market participants expect. That would create winners among domestically protected firms and losers among firms reliant on globalised supply chains. Institutional investors should thus map political risk exposure at the company and country level and stress-test valuations under plausible trade-restriction scenarios.

A tactical note: volatility associated with Iran outcomes is likely to be transitory relative to the multi-year hardware demand curve. Tactical hedges — short-duration options or time-bound commodity forwards — can absorb event risk without distorting long-term exposure to secular winners in the hardware cycle. For longer-term allocations, rebalancing rules that capture dips while controlling concentration risk will likely outperform indiscriminate buying.

Bottom Line

Levinson’s synthesis — geopolitical event risk in the near term and a durable AI hardware cycle over the medium term — captures the two dominant forces institutional investors must price in 2026. Tactical hedges for Iran-related dislocations and selective exposure to balance-sheet-strong hardware and equipment suppliers are pragmatic ways to reconcile short-term volatility with longer-term secular upside. topic topic

Disclaimer: This article is for informational purposes only and does not constitute investment advice.

FAQ

Q: How quickly would an Iran-related disruption affect AI hardware supply chains? A: Direct impacts on silicon wafer supply are unlikely because advanced packaging and lithography capacity are geographically and commercially separate from Middle East oil logistics; however, an escalation that materially raises energy prices could increase operational costs for fabs and data centres, raising unit economics for marginal compute until prices adjust. Historical precedent shows commodity shocks affect sectors through margins and capex delays rather than immediate physical shortages in high-tech inputs (IEA/Bloomberg post-2011 analyses).

Q: Could policy changes derail the hardware cycle? A: Yes — export controls or large-scale subsidy programs could reallocate capital and reorder competitive advantage in months to years. The semiconductor industry’s multi-year lead times mean policy shifts can create permanent winners and losers; investors should stress-test models for scenarios where supply chains regionalise or key inputs face tariffs or restrictions.

Q: What is the historical comparison for this hardware cycle? A: The closest analogue is the datacentre build cycles of the late 2000s and 2010s when cloud migration created sustained hardware demand. The difference today is concentration: a smaller set of hyperscalers driving orders and a higher share of value captured by specialised hardware vendors versus general-purpose server OEMs. This changes the dispersion dynamics and the potential for concentrated winners.