● AI Liquidity Supercycle
Liquidity cycle until 2026, the essence and decisive battleground of the AI investment rally unlike the dot-com era.
This article covers the structure of the liquidity cycle that will continue until 2026, the reality of this AI rally compared to the dot-com bubble, the logic of G2 (USA/China) co-prosperity, the “jobless growth” created by the spread of robotics and AI and the gap between the stock market and the real economy, and even the real ceiling such as power grids, infrastructure permit lead times, and stablecoin liquidity.
It compiles sector-specific roadmaps and risk checklists, a portfolio playbook, and a timeline for 2025 to 2026 all at once.
Reflecting the global economic structure, economic outlook, interest rates, liquidity, and core keywords of the stock market, it connects these to a practical strategy.
Headline briefing: Key points of the 2025 first-half review and the 2026 outlook.
Despite concerns over tariffs and controversies over the AI peak, investment volumes actually expanded in 2025.
Unlike the dot-com cycle, the AI cycle now features a structure where actual corporate investments and productivity innovations are connected to cash flows.
The US-China competition is evolving from “destruction” to “co-prosperity,” intensifying overlapping investment pressures in semiconductors, power, and robotics.
As robotics and AI become more prominent, jobless growth becomes a reality, deepening the gap between a robust stock market and a sluggish real economy.
Liquidity remains favorable until 2026, but the ultimate constraints are determined by physical limitations such as power grids, power rates, and cooling infrastructure.
How this cycle is different from the dot-com era.
In 1999, infrastructure was laid down with the expectation of future profits, but nowadays hyperscalers continuously invest with cash flows and profit models in place.
Enterprises are already paying for AI copilots, Palantir-like decision-making, and automation, with ROI becoming tangible.
With a big tech business base formed around “$2.5 trillion,” the addition of AI as a productivity lever leads to an exponential acceleration in commercialization.
The companies with the largest financial resources are both investors and demand drivers simultaneously, ensuring an endless virtuous cycle from CapEx to revenues to reinvestment.
Liquidity cycle until 2026: Decoupling between a slowing real economy and a strong capital market.
Economic slowdown leads to interest rate cuts and fiscal expansion, activating a “Bad is Good” dynamic.
While the real economy shows signs of weakening in consumption and services, the stock market is lifted by an increasing weight of large tech companies in market capitalization.
As robotics and AI gain more weight, the sensitivity to employment declines, potentially resulting in a prolonged phase where “the economy is sluggish but the market remains strong.”
Liquidity is mostly channeled into risk assets through government bond issuance, fiscal spending, and MMF fund reallocation.
The market is more sensitive to liquidity and interest rate trajectories than to global economic variables.
The logic of G2 co-prosperity: Competition expands total investment.
The United States and China are not mutually destructive but rather compete by enhancing their own competitiveness, each expanding total demand through bold investments.
A structure is emerging where the United States, China, and Korea simultaneously benefit from semiconductors, equipment, memory, and power infrastructure.
Despite a domestic slowdown, China sees sectors like AI, semiconductors, and robotics attracting talent and capital, resulting in conspicuous sector growth.
The United States is accelerating the reorganization of manufacturing and power investments through CHIPS and IRA, providing parallel benefits to supply chains in Korea and Taiwan.
Sector roadmaps for 2025~2026: What are the key areas?
Data center CapEx: GPU/AI accelerators, HBM high-bandwidth memory, CPO/optical communication, rack-scale power, and both immersion and liquid cooling are growing simultaneously.
Software/platform: Copilots, agents, vertical AI, RAG, security, ops stacks, and enterprise data platforms are being monetized.
Robotics/humanoids: Automation in logistics, manufacturing, and retail sites begins to translate into revenue, with the long-term TAM growing in proportion to the rate of job replacement.
Edge AI: On-device inference spreads across PCs, mobile devices, and industrial IoT, structurally benefiting NPUs and low-power AI chips.
Power/infrastructure: Substations, transmission and distribution, distributed power, gas/nuclear power, microgrids, and overhaul of power rate systems determine the physical ceilings for AI expansion.
Materials/parts: Demand for HBM stacking, advanced packaging, copper, high-purity chemical materials, and high-performance cables expands concurrently.
Stock market vs. real economy gap: Why is it widening?
The top 10 tech stocks in the index account for more than half of the market cap, concentrating profits and liquidity.
While consumer goods and domestic services slow down, the index attempts to hit new highs driven by AI, power, and manufacturing restructuring.
An optical illusion emerges as factories increasingly relocate overseas and robots are deployed on-site, resulting in a muted impact on domestic employment and wages.
In conclusion, the trajectories of tangible economic impact and stock market strength diverge.
Risk checklist: Factors that could halt the rally.
Delays in power grid capacity, power rates, and permit approvals may postpone data center expansions.
Bottlenecks in HBM and advanced packaging, equipment lead times, and yield issues could constrain the pace of CAPEx.
If inflation picks up, a tangled downward interest rate trajectory may intensify valuation pressures.
If AI inference costs erode gross margins, the pace of profitability improvement may decelerate.
Content copyright, personal data, and safety regulations could slow down deployment.
Strengthened tariffs and export controls may deliver short-term shocks to specific supply chains.
Key points rarely covered elsewhere.
At the top, “power” is the determining factor: Regional differences in power permit capacity, rates, and cooling methods define the physical ceiling for AI market cap.
The practical pathway of liquidity: A line from fiscal deficits to government bonds to MMF fund reallocation to reduced risk premiums that leads stock prices.
The fluctuations in stablecoin market cap can serve as a high-frequency leading indicator of liquidity for risk assets.
Inference Opex risk: In user expansion phases, inference costs may temporarily erode margins, making price and architecture optimization a performance variable.
The disconnect in Korea: Despite strong exports and good performance from large corporations, offshore factory relocations and automation may dampen domestic employment impacts.
China’s talent concentration: Despite a domestic slowdown, sectors like AI, semiconductors, and robotics may continue to experience excess growth due to the concentration of talent.
Strategy playbook: Portfolio management for 2025~2026.
Core-satellite: The core comprises global mega-cap AI/power infrastructure, while the satellites consist of HBM, optical communication, cooling, robotics, and edge AI.
Regional barbell: Diversify across the United States (platforms, software, power facilities) + Korea/Taiwan (memory, equipment) + selectively in China (semiconductors/robotics supply chains).
Factor mix: Overweight growth/quality, approach sectors sensitive to interest rates only after confirming a shift in profit momentum.
Rebalancing rules: Adjust exposure based on data center power permit indicators, GPU lead times, HBM prices, and credit spread expansions.
Cash management: Utilize dollar-cost averaging and event-driven strategies in line with liquidity events (policy meetings, fiscal executions, major CAPEx guidance).
Hedging: During inflation reheat periods, reduce short-term duration exposure and defend with exposure to energy, utilities, and copper.
Timeline scenarios: Base/Bull/Bear.
Base: Gradual downward trend in interest rates, easing of power infrastructure permit bottlenecks, sustained data center CAPEx, and a one-sided rise dominated by large tech stocks.
Bull: Simultaneous advances in power/cooling innovations and inference efficiency improvements lead to an AI software profitability upgrade, with early acceleration of robotics commercialization.
Bear: A rebound in interest rates due to inflationary reheating, prolonged bottlenecks in HBM and power, slower deployment due to regulatory tightening, and a temporary reversal of multiple re-rating.
News format summary: Numbers that speak now.
Investment scale: Cumulative AI/data center CAPEx for 2024-2025 shows a double-digit increase year-over-year.
Market structure: An expanding weight of top tech stocks in the S&P, with a prolonged stagnation in 400 non-AI companies.
Labor inflection: The introduction of robotics and AI lowers employment elasticity, marking clear signs of “jobless growth.”
G2 drive: A “file-building” competition with simultaneous investment from the US and China expands total demand in semiconductors, power, and robotics.
Ceiling: Power grids, power rates, and permit delays impose physical constraints, while inference costs limit software margins.
Checklist 10 items: Check them immediately.
Are government interest rate trajectories and inflation expectations diminishing?
Are data center power permit capacities and cooling investment plans on the rise?
Are GPU/HBM lead times and prices stabilizing?
Are the conversion rates to paid services and ARPU of copilots/agents increasing?
Is the gross profit margin improving relative to inference costs?
Are changes in power rates and regulations favorable to AI expansion?
Are CHIPS/IRA implementations and permit lead times shortening?
Is domestic demand for AI/robotics in China and talent inflow continuing?
Are credit spreads and MMF fund flows moving toward risk assets?
Are content, personal data, and safety regulations being systematically and predictably organized?
< Summary >
Liquidity remains favorable until 2026, and unlike the dot-com era, this AI rally is underpinned by cash flows.
US-China competition expands total investment, driving concurrent demand in semiconductors, power, and robotics.
With the spread of robotics and AI, the gap between real economic impact and stock market strength may widen due to jobless growth.
The ceiling is defined by power grids and inference costs, with power, cooling, HBM, optical communication, robotics, and edge AI being the key beneficiaries.
The strategy focuses on a core of mega-caps with infrastructure/compute satellites arranged in a barbell pattern, rebalanced based on signals from interest rates, power, and CAPEx.
[Related articles…]
Power Grid Crisis: Bottlenecks of AI Data Centers and Investment Strategies
Stablecoin Liquidity and the Stock Market Correlation until 2026
*Source: [ 경제 읽어주는 남자(김광석TV) ]
– ‘유동성 사이클’ 2026년까지 이어진다. “닷컴버블때와 다르다” 본격화 될 AI 투자 랠리 어디까지 가는가? | 경읽남과 토론합시다 | 김기훈 대표 1편
● AI Biotech Gold Rush
2026 AI Bio Investment Roadmap: A One-Stop Guide Covering Large, Small, and Equipment Strategies, FDA Alternative Testing Methods, Big Tech Platform Wars, and Currency & Interest Rate Variables.
In 2026, the key themes of AI bio are laid out, including the weight to be assigned to large-cap and small-cap stocks, promising sectors, and an actual portfolio allocation guide.
With the FDA’s trend of accepting alternative animal testing driving the fusion of AI and bio, coupled with ADC and bispecific antibody pipelines, the beneficiaries of domestic equipment, parts, and materials (소부장) from automation and analysis equipment, and macro variables such as exchange rates, interest rates, the dollar, inflation, and Nasdaq, everything required to make informed investment decisions is included.
It specifically addresses the three major bottlenecks—data, certification, and equipment—that are rarely covered elsewhere, along with the types of companies that can resolve these issues.
[Headline News Summary]
There is a growing recognition that efficiency improvements in bio are difficult without AI, leading to increased technology transfers and collaborations between major pharmaceutical companies and AI firms.
In anticipation of 2026, ADC and bispecific antibodies in cancer treatments are cited as the most popular and promising segments.
Among domestic large-cap stocks, Celltrion has been noted for its potential revaluation due to factory acquisition effects, shareholder-friendly policies, and an ADC pipeline.
While small-cap stocks have high profit potential, the individual differences are significant, so a mixed portfolio of large and small caps is recommended.
Automated research and manufacturing and the rise in demand for cell analysis equipment indicate that domestic companies producing automation and analysis devices are among the potential beneficiaries in bio equipment, parts, and materials.
Abroad, representative companies like Thermo Fisher Scientific in bio equipment and materials are being reclassified as defensively positioned due to their low sensitivity to economic cycles.
As the competition for drug discovery platforms intensifies with NVIDIA’s BioNeMo and Clara and Alphabet’s Isomorphic Labs, the FDA’s acceptance of alternative non-animal testing methods is seen as a catalyst that accelerates AI adoption.
[Large-Cap vs. Small-Cap: How Much and Where to Invest]
Large caps tend to be more favorable in terms of joint rallies across sectors and come with liquidity and better access to information, making volatility management easier.
Small caps, while capable of delivering tenfold returns based on clinical, contract, and regulatory events, carry the risk of “falling alone,” so diversification is essential.
A recommended portfolio (example) might include 30% domestic large caps, 20–30% domestic small caps, 30% ETFs (a mix of Korean active and U.S. index funds), and 5–10% micro-caps/thematic stocks.
Considering exchange rate and dollar exposure, incorporating dollar-denominated assets or hedged ETFs can be advantageous for guarding against macro risks.
[AI x Bio: Key Points in the Platform War]
Data quality and quantity, computing, and regulatory compliance are the three factors that determine the success of AI-driven drug discovery.
NVIDIA is expanding its ecosystem with BioNeMo (protein and molecule generation/simulation) and Clara (medical imaging and life science computing), and its proof-of-concept trials with major pharmaceutical companies are on the rise.
Alphabet’s Isomorphic Labs is speeding up the commercialization of “AI-driven drug design” by increasing large-scale technology transfer agreements with major pharmaceutical companies.
Once a platform becomes the standard, lock-in effects will occur through APIs, toolchains, and data pipelines, drastically raising entry barriers.
It was also noted in the original conversation that the narrative set by big tech leaders will likely play a critical role in driving sector momentum in the future.
[Regulatory Changes and Cost Structures: Why Now for AI]
The FDA’s stance on allowing non-animal alternative testing is interpreted as a signal that it is actively reviewing evidence based on organoids, cell-based models, simulations, and AI models.
This change is expected to drastically reduce time and costs in the preclinical phase and help filter out experimental failures in the early stages of drug discovery.
At the same time, AI workflows that meet data governance, verification, and traceability (GxP) guidelines may command a premium valuation.
[Macro Variables: Interest Rates, Exchange Rates, the Dollar, Inflation, Nasdaq]
A stable decline in interest rates lowers the discount rate for long-term cash flows, helping growth stocks and bio valuations to improve.
A slowdown in inflation can lead to renewed appetite for risk assets and alleviate clinical and manufacturing cost pressures.
While a stronger dollar and exchange rate can benefit companies with significant overseas revenues when measured in local currency, they may also result in higher borrowing costs in dollars and increased import costs, necessitating individual company assessments.
The risk-on/risk-off cycles of Nasdaq and bio indices quickly translate to domestic bio market sentiments, so timed phased buying in alignment with this interconnectedness is effective.
[Sector Checklist: Which Areas are Promising]
Cancer treatments (ADC, bispecific antibodies).
- Check the clinical stage of lead assets, safety profiles, differentiation compared to similar mechanisms, and potential for global license-outs.
AI Drug Discovery/Protein Design. - Data sources and quality, structured payment milestones in external pharma proof-of-concept deals, and regulatory compliance documentation are key.
Organoids, Imaging, Digital Pathology. - Automation in experiments and reproducibility in analysis are the leverages for revenue generation.
- Verify whether there are repeat orders or service-based contracts (maintenance, software subscriptions) targeting major pharmaceutical companies.
Bio Equipment, Parts, and Materials. - Equipment for cell washing/culture automation, label-free live cell imaging, and process analytics (PAT) are expected to see structural demand alongside the adoption of automation by major pharmaceutical companies.
CDMO & Reshoring. - As production hubs expand in the U.S. and Europe to mitigate tariffs and supply chain risks, the visibility of orders improves.
- Companies with a strong track record in regulatory approvals and quality systems command a premium.
[Bottom-Up Hints: Investment Points Illustrated by Examples]
Domestic Large-Cap.
- In the original conversation, Celltrion was noted for its potential revaluation due to anticipated revenue jumps from factory acquisitions, shareholder-friendly policies, and ADC/bispecific antibody pipelines.
- The shareholder returns and regulatory/policy responsiveness of large-cap companies can support their floors even in volatile markets.
Domestic Equipment & Parts. - Companies producing automation equipment and cell analysis/imaging devices initially penetrate research and pilot processes, with contract announcements and order disclosures with major pharmaceutical firms triggering movement.
International Leaders. - Thermo Fisher Scientific, possessing a broad range of life science equipment and reagents, is viewed as defensively positioned due to its irreplaceable position in research and manufacturing.
Note. - Specific contract or pipeline issues should be verified through public disclosures and press releases, while also checking clinical stages, milestones, cash positions, and cash burn rates.
[Portfolio Construction: Core-Satellite Strategy]
Core (30% Domestic & International Large-Caps).
- Their robust cash flows, production capabilities, and regulatory resilience provide downside protection.
Satellite (20–30% Domestic Small-Caps). - Target event-driven alpha, diversify across 5–7 stocks, and pre-define stop-loss and re-entry rules.
ETFs (30%). - Mitigate individual stock risk with a Korea active bio ETF, and leverage global momentum with a U.S. index-based bio ETF.
Options (5–10% Micro-Caps/Thematic Stocks). - Bet a small amount on themes such as AI, organoids, and ADC to manage volatility from seed investments.
Risk Management. - Cultivate habits of diversifying currency exposure, phased buying and selling, maintaining an events calendar, and monitoring key indicators like interest rates, inflation, and the dollar.
[Five Key Points Others Rarely Mention]
Lock-In of Data and Certification.
- In AI drug discovery, “verifiable data” is crucial over just large volumes, and companies with pipelines that support GxP, traceability, and audit readiness will secure sustained orders.
ARR Conversion of Equipment. - As analysis equipment and software transition to subscription models, they generate recurring revenues unrelated to economic cycles and have significant potential for multiple re-ratings.
Explosive Value of NAMs (Non-Animal Alternative Testing) Acceptance for Organoids and Digital Twins. - Companies with extensive references in institutional and regulatory responses will penetrate the market quickly.
Reshoring and Tariff Risks. - Relocating production sites necessitates re-standardization of facilities and reintroduction of verification equipment, thereby expanding demand for CDMO, automation, and validation devices concurrently.
Mismatch of Macro Variables. - When interest rates and the dollar decline, small and mid-caps tend to exhibit higher beta.
- Conversely, during periods of renewed inflation or surging exchange rates, a tactical shift toward large-cap, equipment, and strong cash-flow companies is necessary.
[Practical Checklist & Timeline]
Events Calendar.
- Pre-list global healthcare conference seasons, major pharmaceutical pipeline days, and key clinical announcements and regulatory decision dates.
Fact-Checking. - Document and verify details such as the scale of technology transfer agreements, milestone conditions, termination clauses, clinical stages, patient counts, and interim analysis designs.
Macro Monitoring. - Keep an eye on trends in interest rates, inflation, exchange rates, the dollar, and risk-on signals in the Nasdaq bio index.
Rebalancing. - Increase ETF weights in sectors with waning momentum, and implement “safety mechanisms” by reducing individual positions before major events.
[Related Posts…]
- AI Bio Era: FDA Alternative Testing Methods Opening a Speed Race in Drug Development
- ADC Boom and Organoid Automation: 2026 Korean Bio Equipment Checklist
*Source: [ Jun’s economy lab ]
– 내년 AI 바이오 이 기업이 오릅니다(ft.강하나 2부)
