Positioning for 2026: AI Infrastructure and Cloud Leaders with Scalable TAM

The setup for 2026 is clear: a bull market with a new, more modest target. According to Wall Street strategists, the S&P 500 is projected to end next year at a level representing 11.6% to 13% upside from current levels. That's a step down from the explosive gains of recent years, but the market's momentum remains intact. The real question for growth investors isn't whether the market will rise, but which companies will lead the next leg of the rally.

That leadership is likely to shift. The past two years have been dominated by a handful of tech giants, creating a dangerous concentration risk. As one portfolio manager noted, the market's heavy reliance on a few tech giants has created a "dangerous single point of failure". This isn't just a theoretical concern; it's a practical one for portfolio construction. When a handful of stocks drive the entire index, diversification becomes essential to manage volatility and capture broader growth.

The opportunity lies in sectors where demand is accelerating and the Total Addressable Market is vast and still being penetrated. The clear winners in this narrative are AI infrastructure and cloud software. These are not speculative bets but foundational layers for the AI economy, with companies scaling rapidly to meet unprecedented demand. The thesis is straightforward: the next wave of high-growth, scalable businesses will emerge from these sectors, offering investors a way to participate in the AI revolution without being fully exposed to the concentrated valuations of the megacaps.

AI Infrastructure: Broadcom's Scalable Chip Advantage

Broadcom's ascent in 2025 was powered by a clear winner: its custom AI accelerator chips. While the company's overall growth was solid, the real story is in the explosive scaling of its AI semiconductor division, which is set to drive outsized returns in 2026. The key differentiator is its expertise in building application-specific integrated circuits (ASICs). These chips are designed to outperform general-purpose alternatives like Nvidia's GPUs in specific, high-volume workloads, delivering better performance at a lower cost. This proposition is so compelling that it has become a strategic imperative for hyperscalers and AI startups alike.

The financial results underscore this shift. In its fiscal 2025 fourth quarter, Broadcom's total revenue grew 28% year-over-year, with earnings per share up 37%. But the standout was the AI semiconductor segment, which grew at a blistering-fast pace of 74% year over year to reach $6.5 billion. For context, that growth rate already surpassed Nvidia's data center segment growth for its latest quarter. The momentum is accelerating further, with the company projecting AI semiconductor revenue of $8.2 billion for Q1, more than double the year-ago figure. This isn't just growth; it's a ramp-up that is scaling at a rate faster than the industry leader.

The durability of Broadcom's moat lies in its unique position as a design partner for companies that lack the in-house chip expertise. It combines its own engineering prowess with a client's specific workload requirements to produce a potent, custom solution. This model is proving scalable, as evidenced by the projected trajectory: Wall Street analysts expect 50% revenue growth to $96 billion in fiscal 2026, followed by another 36% leap to $130 billion in 2027. The primary catalyst for 2026 is the continued scaling of custom AI chip demand, fueled by robust hyperscaler spending on data centers. For a growth investor, Broadcom represents a play on the foundational layer of AI infrastructure, where its scalable model and performance advantages position it to capture a large and growing share of the expanding market.

Cloud Software: Oracle's AI-Driven Reacceleration and Backlog

Oracle's cloud story in 2026 is defined by a massive backlog and a capital plan that signals deep conviction in its growth trajectory. The company's strategic pivot to a chip-neutral, multicloud infrastructure model is now translating into explosive financial execution, positioning it to capture a larger share of the cloud market.

The most striking metric is the surge in its Total Remaining Performance Obligations (RPO), which jumped 438% year-over-year to $523 billion. This isn't just a number; it's a forward-looking commitment tied directly to AI capacity. New deals with major tech players like Meta and NVIDIA are fueling this backlog, which grew by $68 billion in a single quarter. For a growth investor, this is a powerful indicator of demand visibility and the scalability of Oracle's cloud infrastructure (IaaS) business. It suggests the company is not just selling services but locking in long-term revenue streams for its expanding data center footprint.

That footprint is being built at a rapid pace. Oracle's Cloud Infrastructure segment, which includes its core IaaS offering, grew 68% year-over-year in Q2. This growth rate far outpaces the overall cloud revenue increase and demonstrates the strength of its AI data platform. The company is delivering on this demand, handing over roughly 400 megawatts of data center capacity in the quarter and delivering 50% more GPU capacity than the prior period. This operational ramp is critical for converting backlog into future revenue.

To meet this near-term demand, Oracle is making a significant capital commitment. The company has raised its FY 2026 capital expenditure plan to about $50 billion, an increase of roughly $15 billion from its prior forecast. This isn't a speculative bet; it's a direct response to the AI capacity backlog and a signal of confidence in the sustainability of its growth. The plan includes innovative financing models, like customer-provided chips and supplier leases, to manage cash flow while scaling rapidly.

The bottom line is a company reaccelerating. Oracle's financials show it: non-GAAP earnings per share grew 54% last quarter, and the company is investing heavily to build the infrastructure that will drive future profit. For a growth investor, Oracle represents a play on the foundational layer of cloud software, where its massive AI-backed backlog, rapid IaaS growth, and aggressive capex plan indicate a scalable model poised to capture a larger share of the expanding Total Addressable Market.

Catalysts, Risks, and What to Watch

The growth thesis for these AI infrastructure leaders hinges on near-term execution. For investors, the path forward is defined by specific catalysts that will validate the scalability narrative and a set of risks that could challenge it.

For Broadcom, the primary catalyst is the commercialization timeline and volume of custom AI chip orders from major hyperscalers in the second half of 2026. The company's AI semiconductor segment is already growing faster than Nvidia's data center business, and its Q1 guidance of $8.2 billion in revenue points to a blistering ramp. The real test will be whether this momentum translates into sustained, high-volume production contracts. The stock's recent pullback on margin guidance suggests the market is already pricing in near-term execution risks, making the second-half order flow critical for re-rating the stock.

Oracle's main risk is execution on its aggressive capex plans and the conversion of its massive backlog into consistent quarterly revenue growth. The company has raised its FY 2026 capital expenditure plan to about $50 billion, a move that signals deep conviction but also introduces complexity. The key will be managing this build-out efficiently while maintaining its investment-grade balance sheet. The backlog, which surged to $523 billion, is a powerful forward indicator, but the company must now demonstrate it can convert those multi-year commitments into smooth, predictable revenue streams each quarter.

The overarching risk for both companies is a macroeconomic slowdown that could pressure IT spending. While current guidance suggests demand for AI infrastructure remains resilient, a broader economic downturn could force enterprises to defer or scale back capital investments. This is a sector-specific vulnerability that could dampen the growth trajectory for both chipmakers and cloud providers. For now, the evidence points to strong demand, but it is a risk that must be monitored as the year progresses.

In a year where growth is the primary driver, the catalysts and risks for Broadcom and Oracle are clear. The former needs to prove its custom chip orders are durable; the latter must show its massive capex translates into flawless revenue execution. For a growth investor, navigating 2026 means watching these specific milestones while keeping an eye on the broader economic environment.