Artificial intelligence (AI) has transitioned from speculative tech hype to the primary engine of modern macroeconomic growth. For wealth managers and financial advisors, client inquiries have pivoted from general curiosity to tactical asset allocation: where should capital be deployed within the AI ecosystem in 2026?
Deciphering the Five-Layer AI Investment Grid
At the Future Proof wealth management festival, industry experts mapped the complex AI value chain. Gina Sanchez, CEO of Chantico, outlined a structured five-layer framework to assist advisors in portfolio construction:
- Foundational Hardware: Chipmakers and semiconductor giants driving the computing revolution, led by Nvidia (NVDA) and Broadcom (AVGO).
- Hyperscale Infrastructure: The cloud platforms hosting the models, dominated by Microsoft (MSFT), Amazon (AMZN), and Alphabet (GOOG).
- Enterprise Software: The application and service layer, featuring Palantir, ServiceNow, and Adobe.
- Physical Infrastructure: Power generation, utility transmission, and data center cooling technologies required to prevent grid overload.
- Downstream Adopters: Private markets, venture capital startups, and traditional sectors (industrials, real estate) integrating AI for productivity gains.
The Hyper-Capex Wave: $1 Trillion and Beyond
A central pillar of the AI thesis is the massive capital expenditure (capex) of the tech giants. Joe Wilson, managing director at J.P. Morgan Asset Management and portfolio manager of the JTEK ETF, highlighted that the four major hyperscalers—Alphabet, Amazon, Microsoft, and Meta—are projected to invest $650 billion in AI capex this year. Wilson anticipates this spending will compound, reaching a run-rate of $1 trillion within three years. However, advisors must prepare clients for a non-linear path, as the hyperbuild phase faces temporary supply-chain and power constraints.
From a stock-selection perspective, structural advantages are key. Alphabet stands out as the sole player controlling its custom silicon, proprietary large language models (LLMs), massive cloud footprint, and unique consumer data. Conversely, defensive strategies might look to Apple, which has historically maintained a quieter AI profile, insulating it if sentiment briefly sours.
Navigating Froth and Valuations
Is AI in a bubble? David Wright, head of quantitative investments at Pictet Asset Management, and the panel assessed potential warning signs. Unlike the dotcom peak of 1999—where companies like Cisco commanded valuations implying growth beyond European GDP—current AI leaders support their multiples with real free cash flow. Nonetheless, advisors must watch for warning indicators:
- Estimates rising while stock prices stagnate, as seen when Nvidia’s forward earnings estimates rose 100% in six months without price movement, indicating market over-ownership.
- Speculative spillover into unproven quantum computing and micro-cap tech plays.
- Deflationary compute costs; Wright noted that training cycles have shrunk from months to days, suggesting total hardware spend could plateau sooner than expected as software efficiency increases.
Seeking Operating Leverage in Software and Healthcare
A key criteria for software companies in 2026 is their ability to demonstrate GAAP operating leverage. Wilson criticized software firms that underperformed hardware for a decade, urging them to show margin improvements rather than just media buzz. He contrasted Salesforce’s historic 19% GAAP margin with Texas Instruments’ 35–50%+ range. The consensus pick for the most undervalued downstream beneficiary is healthcare. AI applications in diagnostics, billing, insurance processing, and drug discovery are poised to transform the sector.
Frequently Asked Questions
How should advisors gain client exposure to AI in 2026?
Instead of chasing narrow, high-fee thematic ETFs, experts recommend a mix of diversified index funds, active technology strategies (like JTEK) that can dynamically classify emerging platforms, and private equity to capture early-stage application layers.
Is the massive $650 billion AI capex sustainable?
While the capex is projected to scale to $1 trillion, the build-out will likely experience cyclical pauses. Hyperscalers must eventually transition from model training to inference, where the returns depend on consumer and enterprise monetization.
Which sector is the most underowned AI play?
Healthcare is the unanimous consensus. Deep learning models applied to diagnostic scanning, administrative billing, and protein-folding drug discovery present major structural growth opportunities that are currently underallocated in traditional benchmarks.
