J.P. Morgan shifts toward more technology and Taiwan, reducing India allocation
AI summary card
J.P. Morgan shifts toward more technology and Taiwan, reducing India allocation
In an environment of simultaneous stagflation pressure and renewed acceleration in the AI cycle, the report recommends increasing Asia AI/technology exposure toward Taiwan while downgrading Indian equities to neutral.
- AI-related stocks accelerated upward in April, as improvements in Anthropic revenue, model capabilities, token consumption, pricing, and funding collectively eased commercialization concerns.
- Price increases across the AI hardware supply chain have spread from memory to packaging, substrates, networking, optical modules, PCB, cooling, power, and MLCC, driving upward revisions to margin expectations.
- As the macro environment shifts from Goldilocks to stagflation, the report recommends favoring structurally growing areas that are less economically sensitive, including AI, security resilience, energy, and materials.
- Taiwan and technology were upgraded to overweight, with the TWSE target raised to 43,000/48,000/36,000; the MXASJ target was raised to 1,200/1,450/900.
- India was downgraded to neutral due to still-elevated relative valuations, earnings risks, dilution from IPO/QIP and major shareholder sell-downs, insufficient AI and semiconductor exposure, and monsoon and food inflation risks.
Report interpretation
Overview
This Asian equity strategy report discusses the combined impact of the post-Middle East conflict macro focus shifting from risk premium to stagflation, and the renewed acceleration of the AI cycle driven by improving commercialization signals and rising hardware prices. J.P. Morgan believes that in an environment of a two-speed economy and sticky inflation, Asian equities should allocate more toward structurally growing areas that are less economically sensitive, especially AI, technology, security resilience, and some commodity-related stocks; at the same time, exposure to energy-sensitive, cyclical, and catalyst-poor sectors should be reduced.
Core views
The core view is 'more tech/Taiwan, less India.' The report upgrades technology and Taiwan to overweight, believing AI upside is expanding and supply-chain price increases are not over; it upgrades utilities to neutral to reflect power and energy security demand; it downgrades consumer discretionary and communication services to neutral, downgrades consumer staples to underweight, and downgrades Indian equities to neutral. The report remains positive on structural themes in China such as AI winners, capital expenditure beneficiaries, energy security, and robotics, but believes recovery in consumer confidence, governance, shareholder returns, reflation, and anti-involution repair remains slow.
Analysis framework
The report uses a combination of top-down macro scenario judgment, regional and sector allocation adjustments, AI industry-chain price and margin observations, index target revisions, and analysis of India's relative valuation and earnings risks. Its logic chain is: stagflation weakens cyclical and consumption themes, AI commercialization and hardware price increases improve growth visibility, which in turn increases exposure to AI supply chains such as technology, Taiwan, and Korea, while tactical allocation to India is reduced due to valuation, earnings, and dilution pressures.
Methodology notes
Shift from Goldilocks to stagflation
The report believes that although there is a negotiation path for the Middle East situation, higher commodity prices and tighter financial conditions may leave lasting effects, shifting the market from broad-based growth and policy support toward an environment with higher inflation pressure and more pronounced growth divergence.
Structurally growing areas that are less economically sensitive
In a stagflationary and K-shaped economy, the report favors growth themes such as AI, security resilience, and energy security, whose demand is less dependent on the economic cycle, while reducing weights in traditional cyclicals, non-AI consumption, and some value equities.
Supply tightness drives upward revisions to prices and margins
The report tracks price increases across memory, packaging, substrates, networking, optical modules, PCB, CCL, cooling, power, and MLCC, and uses upward revisions to margin expectations to explain the performance of Asian AI hardware stocks.
Valuation premium and earnings visibility
The report cites India's valuation premium relative to MSCI EM, peer-market valuations, earnings downgrades, equity dilution, and sector technology exposure as the main reasons for downgrading India to neutral.
Asset mapping & comparison
Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).
- Taiwan equitiesUpgraded to overweight, one of the core beneficiary markets of Asia's AI and technology broadening trade.
- Strengths
- TSMC and Taiwan's technology supply chain hold key positions in AI computing power, semiconductors, and hardware supply, and the TWSE target has been raised.
- Weaknesses
- The report notes that some Taiwan technology stocks have risen too quickly, and retail buying plus technical positioning may be crowded.
- Comparison
- Compared with India, Taiwan has more direct exposure to AI, data centers, and semiconductors.
- Risks
- A pullback in AI commercialization, reversal in hardware prices, downcycle in the supply chain, and short-term technical correction.
- Korea equitiesMaintained at overweight, benefiting from the memory and AI hardware cycle.
- Strengths
- Memory is one of the strongest-performing categories in Asia's AI stack, and Korean technology had already significantly outperformed earlier.
- Weaknesses
- If AI demand or memory prices peak, earnings and valuations may come under pressure.
- Comparison
- Together with Taiwan, it forms the main allocation direction for the upcycle in Asia AI hardware.
- Risks
- Semiconductor cycle volatility, customer capex adjustments, and insufficient sustainability of price increases.
- China equitiesMaintained at overweight, but with emphasis on structural opportunities rather than a broad recovery.
- Strengths
- AI winners, capex beneficiaries, energy security, and robotics themes remain attractive; A-share liquidity has improved.
- Weaknesses
- Recovery in consumer confidence, governance/shareholder returns, reflation, and anti-involution remains slow, and China's internet sector lacks visible catalysts.
- Comparison
- Compared with India, valuations are more attractive, but fundamental recovery remains uneven.
- Risks
- Export stocks face stagflation and tariff risks, and broad reflation may stall due to household balance-sheet repair.
- India equitiesDowngraded to neutral.
- Strengths
- The long-term structural growth story remains intact, and policy stability plus high growth had supported a scarcity premium.
- Weaknesses
- Valuation remains high relative to EM, earnings face downgrade risk, IPO/QIP and major shareholder sell-downs create dilution, and large caps have low exposure to AI, data centers, and semiconductors.
- Comparison
- The report believes EM markets such as Korea, Brazil, and China offer cheaper entry points with similar or better forward growth.
- Risks
- Energy supply disruptions, monsoon below long-term average, El Nino, food inflation, and foreign outflows.
- Asia Technology sectorUpgraded to overweight.
- Strengths
- AI demand, token consumption, model capabilities, improved funding, and hardware price increases jointly support margin and revenue expectations.
- Weaknesses
- Some share prices have risen extremely quickly, and valuations and positioning may be crowded.
- Comparison
- Better suited than traditional cyclicals and non-AI consumption themes to a quality-growth framework under stagflation.
- Risks
- AI services lack natural moats, open-source and low-cost Chinese models suppress commercialization, and capex is reassessed.
- UtilitiesUpgraded to neutral.
- Strengths
- AI data centers, power demand, and energy security investment improve demand visibility.
- Weaknesses
- Utilities may still be affected by high interest rates and capex pressure.
- Comparison
- Compared with traditional cyclical sectors, it benefits from themes of power and resilience investment.
- Risks
- Interest rates staying high, constrained regulatory returns, and energy cost volatility.
- Consumer Discretionary and Communication ServicesDowngraded to neutral.
- Strengths
- Some China consumer and internet names may still benefit from younger-generation health demand or tactical rebounds.
- Weaknesses
- A stagflationary environment, slow recovery in consumer confidence, and insufficient catalysts for China's internet sector weaken the appeal.
- Comparison
- Weaker than AI, security resilience, and commodity-related quality growth themes.
- Risks
- Weak demand, sticky rates and inflation, and policy and competitive pressure.
- Consumer StaplesDowngraded to underweight.
- Strengths
- Traditional defensive characteristics still exist.
- Weaknesses
- The report believes it is sensitive to commodity prices and will come under pressure in a stagflation shock.
- Comparison
- Compared with energy, materials, and AI technology, the report gives it lower allocation priority.
- Risks
- Rising raw material costs, food inflation, and margin compression.
Key data
- TWSE target43,000/48,000/36,000Corresponding to the base, bull, and bear scenarios respectively; the report upgraded Taiwan allocation to overweight.
- MXASJ target1,200/1,450/900Corresponding to the base, bull, and bear scenarios respectively; higher Taiwan and Korea targets partly offset the India downgrade.
- MXCN and CSI-300 targets100/5,200The report maintains its base targets for MXCN and CSI-300 through the end of 2026.
- Anthropic ARR changeAbout $0.9 billion to about $3.0 billionThe report says Anthropic ARR rose from about $0.9 billion at the end of 2025 to about $3.0 billion in April 2026, easing AI commercialization concerns.
- Token consumption growthAbout 10x YoY to about 20xThe report says global token consumption growth accelerated from about 10x YoY at the end of 2025 to about 20x currently.
- AI hardware company margin expectationsAverage upward revision of more than 10%The report says that since September 2025, gross margin expectations for a broad AI hardware sample have risen by more than 10% on average.
- India relative MSCI EM valuation premium65%It has narrowed from the peak of 109%, but the report believes it remains high relative to markets such as Korea, Brazil, and China.
- India earnings forecast revisionCY26E/27E MSCI India EPS growth cut by 2%/1% to 11%/13%Energy supply disruptions and sector earnings risks are among the reasons for the downgrade.
- India fund flows and dilutionDomestic inflows of about $120 billion; FII outflows of about $37 billion; IPO/QIP and major shareholder sell-downs of about $64 billionThe report believes heavy issuance and sell-downs dilute existing shareholders and limit upside.
- India monsoon forecast2026 monsoon at 92% of the long-term averageThe report mentions that developing El Nino conditions could threaten rural incomes and push up food inflation.
Impact & implications
The investment implication is that Asian portfolios should increase exposure to AI hardware, semiconductors, and the Taiwan and Korea technology supply chains, while maintaining overweight positions in energy, materials, and financials; at the same time, weights in India, traditional consumption, communication services, and consumer staples should be reduced. If AI investment is seen as strategically necessary like defense or energy, global semiconductor spending may still have room to expand; however, some Taiwan and technology stocks have risen too quickly in the short term and may consolidate.
Risks
- AI commercialization risk has not disappeared; AI services lack clear network effects and stable moats, and technology leadership may continue to shift.
- Rising share of open-source models and low-cost Chinese models may suppress AI service pricing and profitability.
- AI hardware price increases and margin upgrades may be unsustainable; if supply-demand tightness eases, related stocks face valuation and earnings pullbacks.
- Short-term price action in technology sectors such as Taiwan is extreme, with crowded positioning and technicals, and near-term consolidation may occur.
- The Middle East situation is not fully resolved; if conflict persists, it may create nonlinear risks to commodity supply.
- Stagflation, high energy prices, and tighter financial conditions may pressure cyclicals, consumption, and high-valuation assets.
- India faces risks from still-high valuations, earnings downgrades, equity dilution, foreign outflows, the monsoon, and food inflation.
- China market fundamental recovery is slow, and improvements in consumer confidence, reflation, governance, and shareholder returns still need time.
What to watch
- Revenue, token consumption, model capabilities, price adjustments, and funding progress at Anthropic and other AI labs.
- Price trends in AI hardware segments such as memory, packaging, substrates, PCB, networking, optical modules, cooling, power, and MLCC.
- Whether gross margin and earnings expectations for Asian AI hardware companies continue to be revised upward.
- Whether Taiwan retail buying, RSI overbought breadth, and technology sector positioning continue to become more crowded.
- Whether the Middle East situation, energy prices, developed-market bond yields, and financial conditions remain tight.
- India corporate earnings revisions, IPO/QIP issuance pace, major shareholder sell-downs, FII flows, and monsoon data.
- Whether China's consumer confidence, anti-involution, shareholder returns, reflation, and internet catalysts improve.
- U.S. and global technology capex, cybersecurity-related spending, and AI data center power demand.