Browse macro investment themes, each backed by one or more AI-validated theses with full basket analysis, scoring, and paper trade results.
EU Drinking Water Directive and ECHA universal PFAS restriction create legally mandated, non-discretionary spending on water treatment, site remediation, and filtration tech. A picks-and-shovels basket of utilities, engineers, and testing firms captures this multi-decade, €95B+ compliance cycle.
Companies that have inherited or built entrenched positions in the semiconductor supply chain possess durable competitive advantages. These positions—spanning lithography, specialty materials, advanced packaging, and foundry—create high switching costs and pricing power as AI, onshoring, and geopolitical fragmentation drive sustained demand.
AI agents represent the next paradigm shift in computing, transitioning from copilots to autonomous agents that can reason, plan, and execute multi-step workflows. This thesis centers on companies building infrastructure, platforms, and applications for the agentic AI era.
The explosive growth of AI and cloud computing is driving unprecedented electricity demand from data centers. Nuclear energy—both conventional and next-gen SMRs—is emerging as the only scalable, clean baseload power source capable of meeting this demand, creating a multi-decade investment opportunity across the nuclear value chain.
The investment thesis posits that uranium prices are near a government-backed support level, offering an asymmetric risk/reward profile. A rebound in uranium prices will disproportionately benefit domestic producers, developers, and physical trusts.
Western rare earth supply chains face a reckoning as EV motor, wind turbine, and defense magnet demand surges while China dominates 85%+ of processing capacity. Geopolitical tensions and supply security concerns are forcing massive investment in non-Chinese supply chains, creating opportunities across miners, processors, and magnet manufacturers. The thesis hinges on structural supply deficits in NdFeB permanent magnets forming this decade.
NVIDIA's shift to 800V DC (HVDC) architecture for AI data centers addresses exploding rack power densities (100kW→1MW+), creating a multi-year capex tailwind for power infrastructure, wide-bandgap semiconductors, and cooling specialists. Early movers with NVIDIA-validated reference designs are positioned to capture disproportionate share as 2026–2027 ramps begin.
Swarm drone technology represents a paradigm shift in both military and commercial applications, enabled by AI advancements, declining hardware costs, and lessons from recent conflicts. The thesis centers on autonomous coordinated drone networks disrupting traditional defense systems, logistics, and infrastructure monitoring at fraction of conventional costs.
Contrary to the 'AI is killing SaaS' narrative, AI is deepening moats for incumbent software platforms by embedding copilots, agents, and intelligent automation into existing workflows. SaaS companies with proprietary data, distribution, and switching costs are leveraging AI to expand ARPU, reduce churn, and create new product categories—driving a reacceleration in cloud spending.
PFAS 'forever chemical' contamination has triggered stringent EPA and EU regulations, creating a multi-hundred-billion-dollar remediation market spanning water treatment, environmental consulting, waste destruction, and specialized technologies. Established infrastructure and environmental services firms with proven PFAS capabilities are positioned to capture decades of mandated cleanup spending.
CPO supply bottleneck thesis
AI data center buildout is creating unprecedented electricity demand growth after two decades of flat US power consumption. BCA Research and major banks now project 15-25% electricity demand growth through 2030 driven by hyperscaler AI workloads, favoring independent power producers and regulated utilities with generation capacity in data center corridors.
As GPU rack densities exceed 100kW, air cooling fails physically and hyperscalers must adopt direct-to-chip and immersion liquid cooling. Vertiv and Modine are the leading public beneficiaries, with expanding order books and NVIDIA-validated reference designs for 2026-2027 buildouts.
AI data-center growth is now constrained by firm, dispatchable electricity rather than chips or capital. Interconnection queues, turbine backlogs, and multi-year lead times transfer pricing power to owners of existing nuclear/gas generation and grid equipment makers, creating a durable multi-year scarcity rent through 2030+.
Structural bull thesis on cybersecurity driven by AI-expanded attack surfaces, non-discretionary spending (~13% growth to ~$249B in 2026), and platform consolidation favoring scale leaders. Leading pure-play and platform vendors with >120% net retention and strong FCF are positioned to outgrow broader software.
With China dominating ~90% of rare earth magnet production and expanding export controls, Western supply chains face a structural squeeze. MP Materials and Albemarle (via specialty chemicals/energy storage exposure) represent US-aligned plays on reshoring critical minerals, though both trade under heavy technical pressure currently.
The DOE article highlights accelerating HALEU (high-assay low-enriched uranium) availability to unlock advanced nuclear reactors. Only a handful of Western companies can enrich uranium to 19.75% HALEU, creating a structural bottleneck with government-backed demand. This thesis targets the enrichment suppliers, fuel fabricators, and advanced reactor developers positioned to monetize this chokepoint.
Uranium enrichment—not mining—is the true bottleneck of the nuclear renaissance. Western capacity is dominated by Russian supply, HALEU is in structural deficit through 2030, and the DOE is funding domestic buildout, giving merchant enrichers with new centrifuge capacity exceptional pricing power.
Solar PV is now the dominant source of new global capacity (~664 GW added in 2025), driven by lowest-cost generation, AI/data center demand, and storage pairing. Thesis favors upstream/midstream players with pricing power and policy-protected positions over commodity module assemblers.
SpaceX's $8B spectrum purchase and 15,000-satellite D2D approval forces massive terrestrial densification. TowerCos, fiber backhaul providers, and T-Mobile as ecosystem partner stand to benefit disproportionately versus threatened legacy MNOs.
Peptides offer highly targeted signaling with favorable safety profiles, validated by the GLP-1 revolution. Consumer demand for proactive health, telehealth access, and expanding pipelines drive adoption. Pure-play exposure is limited; only GLP provided real-time data, so related ecosystem leaders are included for context.
Thesis argues capital rotates from crowded AI trades into undervalued insurers with strong moats and TAM. Best candidates combine underwriting discipline, float leverage, and cheap valuations.
AI is compressing drug discovery timelines from years to months. This thesis maps compute, data, and AI-platform providers enabling a potential cancer cure.