Artificial intelligence on capital markets

Artificial intelligence is not a new concept, but it is a new capital market phenomenon. With the massive use of generative models, increasing investments in computing power and widespread commercial use, AI has crossed a threshold: It no longer only influences individual companies, but entire market segments, valuation models and capital flows.

At the start of 2026, many investors are faced with a familiar but wrongly asked question: Is AI a bubble?

However, this question is inadequate for long-term wealth strategies. Capital markets do not react to technological upheavals in a binary way, but cyclically, selectively and often with a time delay.

An overview of the most important things:

• Market structure: AI is a structure-changing factor for capital markets in 2026, not just a hype.
• Technology cycle: Euphoria → overvaluation → correction → long-term substance repeats itself.
• reviews: Overvaluations in AI are cyclical; stable cash flows and market leadership are critical.
• Diversification: Countries, segments and company sizes significantly reduce cluster risks.
• Regional opportunities: US stocks drive innovation, EU companies provide stability, Swiss companies provide niche opportunities.
• Portfolio structure: Combine core (stable companies) and satellite positions (growth companies).
• Fundamental criteria: Revenue growth, margin stability and infrastructure dependency determine long-term success.

Technological cycles and market mechanics

Technological innovations rarely have a linear effect on capital markets. Historical data shows recurring patterns: euphoria, overvaluation, correction, and long-term integration. Artificial intelligence (AI) differs in its cross-sectional impact on cloud, industrial automation, finance and healthcare. This leads to opportunities for companies with technological leadership, but also to concentration risks in portfolios, as few companies control a large part of market-relevant AI applications.

An example: During the dot-com bubble of the late 1990s, Internet stocks rose dramatically, but many companies had no cash flows. Today, the leading AI companies already have sales and stable business models. This shifts the type of risk, but doesn't change the need to carefully analyze cyclical and fundamental data.

Assessment and concentration risks

The valuation of AI companies is often based on long-term earnings assumptions that are discounted to the present. High expectations for margins, economies of scale and market shares lead to high volatility as soon as results do not materialize as forecast.

The capital intensity of AI is high: data centers, energy-intensive hardware, and specialized specialists generate real costs that can influence investment cycles, inflation, and interest rates at a macro level. A critical factor in 2026 is the “energy bottleneck”: Without a massive scaling of power grids and the expansion of baseload capacities (in particular through modern nuclear power and smart grids), the growth of data centers is reaching physical limits. Investors must therefore assess the energy dependence of their tech positions and increasingly consider utilities as indirect beneficiaries of AI. Strategic portfolios must take these factors into account in order to properly weight risks, with an objective and independent asset manager provides decisive guidance in evaluation and allocation.

‍→ Arrange your free, non-binding initial consultation on site.

Historical comparison cycles

Technological innovations often follow recurring patterns that can be observed over centuries. From railroads to electricity and the Internet to mobile technologies, markets are showing similar phases of euphoria, overvaluation, correction, and long-term stabilization. An understanding of these historical cycles helps investors to categorize short-term volatility and correctly assess the long-term economic substance of new technologies such as artificial intelligence.

Waves of technological innovation at a glance

• railway
  
  • Peak phase: 1840s
  • Correction: 1847—1849
  • Long-term impact: industrialization, long-term productivity
• electricity
  
  • High phase: 1890—1905
  • Correction: early 20th century
  • Long-term impact: basis of modern industry
• internet
  
  • Peak phase: 1998—2000
  • Correction: 2000—2002
  • Long-term impact: digital economy, platform models
• Mobile
  
  • Peak phase: 2007-2011
  • Correction: 2012
  • Long-term impact: mobile platforms, new economies
• KI
  
  • Peak phase: 2023—
  • Correction: open
  • Long-term impact: structural change in software, industry, services

‍

AI stocks — global selection and structure

Selecting suitable AI stocks requires a deep understanding of the global market structure and the different segments within the sector. Artificial intelligence is not limited to a single market or region: US companies are driving innovation and scaling, European companies offer stability and niche expertise, while Swiss companies are showing operational strength, particularly in the B2B and special segments. A strategic approach takes into account both the technology segment (hardware, platforms, software, industrial automation) as well as country risks, assessment levels and cycle sensitivity. The following overview provides an initial framework for classifying relevant titles.

Companies in the AI ecosystem — overview by segment and risk profile

• NVIDIA
  
  • country: USA
    
    
  • Segment: Chips/ Hardware
    
    
  • Characteristic/Risk profile: Cyclical, high margins, central to computing power
    
    
• microsoft
  
  • country: USA
    
    
  • Segment: Platform/ Integration
    
    
  • Characteristic/Risk profile: Diversified, stable cash flows, cloud + AI
    
    
• alphabet
  
  • country: USA
    
    
  • Segment: Data & Models
    
    
  • Characteristic/Risk profile: Research-intensive, long-term monetization
    
    
• amazon
  
  • country: USA
    
    
  • Segment: Cloud infrastructure
    
    
  • Characteristic/Risk Profile: Capital-intensive, AWS dependency
    
    
• meta
  
  • country: USA
    
    
  • Segment: Platform/Generative AI
    
    
  • Characteristic/Risk profile: Research-intensive, growth potential, regulatory risk
    
    
• Apple
  
  • country: USA
    
    
  • Segment: Hardware + AI integration
    
    
  • Characteristic/Risk profile: Stable, profitable, long-term cash flows, platform effects
    
    
• Constellation Energy
  
  • country: USA
    
    
  • Segment: Energy/Nuclear
    
    
  • Characteristic/Risk profile: AI enabler; benefits from 24/7 power requirements of data centers
    
    
• ASML
  
  • country: NL
    
    
  • Segment: Semiconductor Technology
    
    
  • Characteristic/Risk profile: bottleneck provider, stable demand
    
    
• Infineon
  
  • country: DE
    
    
  • Segment: Chips/AI applications
    
    
  • Characteristic/Risk profile: Europe, diversified, moderate growth
    
    
• SAP
  
  • country: DE
    
    
  • Segment: Enterprise AI
    
    
  • Characteristic/Risk profile: Software, conservative
    
    
• FIG
  
  • country: CH
    
    
  • Segment: Industrial Automation/AI
    
    
  • Characteristic/Risk profile: B2B, stable income
    
    
• logitech
  
  • country: CH
    
    
  • Segment: AI periphery
    
    
  • Characteristic/Risk profile: Niche products, cyclical
    
    
• Sophia Genetics
  
  • country: CH
    
    
  • Segment: AI in healthcare
    
    
  • Characteristic/Risk profile: growth-oriented, higher risk
    
    
• Schneider Electric
  
  • country: FR
    
    
  • Segment: Energy Management
    
    
  • Characteristic/Risk profile: Infrastructure beneficiary; focus on efficiency and network stability
    
    

Portfolio insight:

Diversification across countries, segments and company profiles reduces cluster risks. Swiss companies offer stability and niche opportunities, US titles provide innovation drivers, European companies offer medium-term stability.

Portfolio Implications for 2026

Integrating AI into asset portfolios requires a disciplined, long-term approach. The technology has a cross-cutting effect, not in isolation, which is why investors must consider cross-sectoral effects. AI is not only influencing technology companies, but also industrial automation, financial services, healthcare and infrastructure. The allocation must therefore be broadly diversified and segmented according to risk profile in order to avoid unwanted cluster risks.

Another aspect is the cyclical nature of valuations. Historically, valuation bubbles often have short periods of extreme multiples, followed by corrections of 20-50% in specific sectors. A portfolio that disproportionately weights AI stocks can therefore react in a highly volatile manner, even if the fundamental technology remains intact. Strategic positioning should therefore be based on weighted allocations and scenario analyses that reflect both upturn and downturn phases.

A layering approach is also recommended:

• Core positions in established, stable companies (e.g. Microsoft, ABB)
• Satellite positions in infrastructure enablers (such as specialized energy providers or power grid operators that provide the physical foundation for scaling AI)
• Cash or alternative risk buffer layer to cushion short-term volatility

Regular rebalancing ensures that strategic weighting is maintained and that corrections do not lead to unwanted overexposure. Finally, investors should consider fundamental criteria such as revenue growth, margin stability, and infrastructure dependency, not just market trends.

Historical lessons for investors

The analysis of historical technology cycles provides clear information for structuring asset portfolios in times of disruptive innovation. Artificial intelligence is not an isolated trend, but part of a pattern that is repeated over centuries: euphoria, correction and long-term value creation. It is crucial for HNW investors to understand these cycles in order to systematically seize opportunities and limit risks.

Historical technology waves — market reactions and portfolio implications

• Railroad (19th century)
  
  
  • Cycle phase: euphoria → correction
    
    
  • Market reaction: Short, sharp price increases followed by crashes
    
    
  • Lesson for investors: Short-term exaggerations don't destroy long-term substance
    
    
  • Portfolio implication: Long-term positioning in core companies, risk buffer for cyclical fluctuations
    
    
• Electricity (1890—1905)
  
  
  • Cycle phase: introduction → overvaluation → stabilization
    
    
  • Market response: High valuations during pioneering phases, moderate corrections
    
    
  • Lesson for investors: Fundamental significance is evident in the long term
    
    
  • Portfolio implication: Structured allocation to fast-growing, stable companies
    
    
• Internet (1998—2002)
  
  
  • Cycle phase: dot-com boom → crash → recovery
    
    
  • Market reaction: Massive corrections, many companies disappear
    
    
  • Lesson for investors: Discipline beats timing
    
    
  • Portfolio implication: diversification across countries and segments, core/satellite strategy
    
    
• Mobile technologies (2007-2012)
  
  
  • Cycle phase: hype → consolidation → platform maturity
    
    
  • Market Reaction: Cyclical Overvaluations, Some Winners, Many Losers
    
    
  • Lesson for investors: Only technologically robust companies survive
    
    
  • Portfolio Implication: Focused Investments in Platform Leaders, Broad to Reduce Risk
    
    
• Artificial intelligence (2023—)
  
  
  • Cycle phase: peak phase → ongoing adjustment
    
    
  • Market response: sharp increases in valuations, ongoing corrections
    
    
  • Lesson for investors: Overvaluations are cyclical; structure remains long-term
    
    
  • Portfolio implication: limiting thematic allocation, risk management, layering of core/satellite positions
    
    

The overview shows that technological cycles always start off volatile but create substantial assets over the long term. Investors should not overestimate short-term market movements but implement structured scenarios and risk buffers to take advantage of long-term opportunities.

For detailed figures, short-term market dynamics and specific portfolio implications, we recommend the Monthly Report November 2025. This report complements the strategic analysis in this article, shows the development of the peak phase of AI on US, EU and Swiss markets, and provides practical advice on the allocation of core and satellite positions for 2026.