Morgan Stanley: AI Capex Recharges Emerging Markets Earnings
Context and Chronology
Morgan Stanley’s research team identifies an unusual wave of forward earnings upgrades across a subset of emerging‑market issuers and traces the primary engine to accelerating corporate capital expenditure tied to AI infrastructure — chiefly chips, high‑performance memory, data centers and automation projects. The bank argues the mechanism is supply‑driven: concentrated, verifiable upstream order intake for foundries, memory suppliers and equipment vendors is increasing near‑term revenue visibility for firms embedded in global AI supply chains, which in turn explains the clustering of analyst upgrades.
Independent confirmations sharpen the picture. Public statements and guidance from leading foundries and memory houses, ASML’s noted backlog for lithography tools, and stronger quarters at large contract manufacturers (with market commentary citing material, step‑up revenue plans from some suppliers) show demand translating into upstream order books. A Reuters report that China cleared imports of Nvidia H200 accelerators for major domestic firms removed a procurement overhang in one important market, while TSMC’s results and wafer‑allocation signals indicate hyperscaler prioritisation for HBM and high‑performance DRAM.
Flow evidence complements the fundamentals: allocations into international and EM technology‑linked exposures have accelerated, the iShares MSCI Emerging Markets ETF (EEM) has rallied from prior lows and Asia‑Pacific benchmarks have outperformed year‑to‑date. Currency moves have amplified dollar‑listed returns and prompted portfolio managers to revisit hedging, further magnifying tactical tilts into EM names tied to AI capex.
But the episode sits inside a wider market debate that complicates interpretation. A high‑profile Citirini Research memo framed an agentic‑AI adoption stress scenario that drove abrupt trader rotations away from labour‑sensitive names and focused attention on concentration, adoption speed and systemic transmission. That memo prompted immediate pushback from large institutions and a White House economist, yet also triggered visible market reaction — quant funds and pension managers trimmed exposures, credit desks flagged wider spreads for smaller vendors, and discussions among supervisors intensified.
Crucially, public and private stress exercises diverge on severity and timing. Several banks’ internal stress runs and supervisory probes (European supervisors and the ECB signalled stepped‑up monitoring of labour and bank exposures) produced materially milder peak‑to‑trough outcomes than the most extreme scenarios in the memo, reflecting different assumptions about adoption speed, concentration of procurement, integration frictions and policy backstops. Markets, however, have at times priced demand‑feedback mechanisms faster than formal sell‑side forecast revisions have adjusted, creating a gap between price action and consensus analyst downgrades or upgrades.
Operational and financing frictions are immediate constraints. Upstream bottlenecks — substrate shortages, packaging/test throughput and long qualification lead times — together with permitting, construction and talent limits for onshore data‑center builds impose ceilings on how quickly capex can convert into recurring revenue. Private capital and insurers are already reworking financing terms for large infrastructure projects (tighter covenants, repriced coverage) and liquidity providers are monitoring vendor refinancing risks.
The net market effect is therefore powerful but concentrated: equipment makers, foundries, memory suppliers and packaging/test specialists linked directly to hyperscaler procurement show clearer revenue visibility and have seen multiples rerate, while many EM firms unconnected to AI supply chains risk relative underperformance. That concentration creates asymmetric outcomes — amplified upside where procurement, capacity and policy align, but rapid downside if export controls, supply shocks, permitting delays or sharper currency moves re‑emerge.
For portfolio managers and corporate strategists the playbook is twofold: selectively capture momentum among a compact set of winners (based on capex confirmations, order backlogs and margin pathways) while embedding active risk controls — stress‑testing demand reproducibility, monitoring supervisory signals, and considering financing/covenant fragility. Earnings calls and capex execution milestones this season will be high‑signal events for distinguishing durable spending from near‑term re‑rating driven by concentrated procurement.
Read Our Expert Analysis
Create an account or login for free to unlock our expert analysis and key takeaways for this development.
By continuing, you agree to receive marketing communications and our weekly newsletter. You can opt-out at any time.
Recommended for you
Earnings Season Puts Big Tech’s AI Spending Under the Microscope
The 2026 reporting cycle will force large technology companies to defend ramped-up AI infrastructure investments as investors demand clearer paths to profit; at the same time, direct demand confirmations from major foundries and a new U.S.–Taiwan trade arrangement are reshaping where and how that capacity will be built. Markets will weigh not only hyperscaler capex plans but whether upstream capacity growth — notably from firms like TSMC — meaningfully reduces delivery risk and shortens the timeline to monetization.
International stocks surge as markets recalibrate the U.S. AI story
Global and emerging-market equities have recently outpaced U.S. shares as investors re‑price the likely economic payoff from concentrated AI investment; that technical repricing has been amplified by policy and currency headlines out of Washington and Asia, prompting tactical reallocations and expanded hedging by institutions.
AI surge reshapes market winners and losers as enterprise software stocks tumble
A rapid narrative shift toward agent-style generative AI has triggered deep selling across many cloud and SaaS incumbents while concentrating capital on model builders, compute hosts and AI-security vendors. The change is rippling beyond equities into private‑equity and credit markets as hyperscalers accelerate capital plans and suppliers signal strong upstream demand that could both validate long‑term compute growth and tighten execution risks for smaller vendors.
NVIDIA Drives Market Nervousness as Jobs Data and Earnings Converge
NVIDIA’s latest report amplified investor anxiety, dragging technology names after shares swung on a mix of earnings signals and executive clarifications. Market participants are parsing CEO comments, downstream capacity moves and an upcoming U.S. jobs print on March 6 for clues on whether hyperscaler capex will convert into revenue versus simply inflating infrastructure costs.
Earnings, China Approvals and Tight Memory Supply Lift Global Chip Stocks
A combination of strong quarterly results at key equipment and memory suppliers and reports China has cleared purchases of Nvidia’s H200 helped lift chip stocks, reflecting both immediate demand and a reduced geopolitical overhang. Together with signs that foundries are confirming hyperscaler demand and will accelerate capex, the moves point to a multi-quarter lift in capital spending and selective revenue upside across the semiconductor chain.
U.S. Debt Markets Ride a Wave of AI Data‑Center Construction
A roughly $3 trillion AI data‑center build‑out is reshaping credit demand and expanding issuance across loans, bonds and securitized products, even as concentrated hyperscaler procurement, community permitting fights and repurposed crypto‑mining campuses introduce execution and political risks. Lenders, insurers and asset managers are widening underwriting lenses—adding covenant protections, stress tests and sector‑specific cash‑flow analysis—while regulators and rating agencies scrutinize leverage, tenant concentration and geographic clustering.
US AI Concerns Push Global Capital into Asia’s Chip Suppliers
Worries in US markets about AI-driven disruption are accelerating a tactical reallocation of capital into Asian semiconductor suppliers and related infrastructure, lifting regional benchmarks and re‑rating equipment, foundry and memory names. The shift is reinforced by industry results and policy signals — from ASML order backlogs to reports of Nvidia system access in China and stronger capex guidance at TSMC — but it concentrates risk in a handful of suppliers and geographies.
Applied Materials raises outlook as AI and memory demand fuels equipment spending
Applied Materials raised its fiscal Q2 revenue outlook well above Street estimates, citing stronger orders tied to AI accelerator and high‑performance memory production. Industry signals — large ASML bookings, TSMC’s capex confirmation and reports of eased export uncertainty for high‑end accelerators in China — corroborate the company’s read of accelerating demand, though long lead times and pull‑forward risk temper the outlook.