INSIGHTS

On August 10, 2026

AAM Viewpoints — The Discipline Behind Durable AI Exposure

An Inflection Point for AI Investing

As the dust settles from several incidents that appear to mark an inflection point in artificial intelligence (AI) investing, market participants may be wondering what’s next for one of the most transformative themes of this era. Many investors are likely reassessing risk appetite after a string of blowups in recent weeks: South Korean Leveraged ETFs (exchange-traded funds) tied to individual semiconductor stocks unraveled when rallies in underlying names lost steam, and hedge fund Situational Awareness faced massive liquidations after concentrated, highly leveraged bets on AI infrastructure names sharply reversed. We believe these blowups are clear warnings against overconcentration, but investors shouldn’t assume a broad-based index will solve the problem either. The AI expansion is real and ongoing, but in our view, investors should be deliberate about how they participate in it. Balancing diversification with selectivity may be the key to potential rewards in the next phase of this super-cycle.

Two Blowups, One Avoidable Lesson

The recent unwinds may have triggered flashbacks to the dotcom bust, but the two are fundamentally different, and both blowups were avoidable with basic risk management. In the South Korean ETF case, spiking volatility in a handful of semiconductor names caused the daily-resetting leveraged returns to work quickly against investors. Poor retail understanding of how these products function compounded the damage and prompted regulators to halt new listings. Situational Awareness ran into similar trouble. Concentration and leverage of up to 4x forced a block liquidation after the fund reportedly lost over 70% of its asset value. The takeaway is clear: exuberance in this space can generate outsized returns quickly, but concentrated, leveraged bets can just as quickly turn disastrous.

Source: FactSet (12/31/2025 – 07/31/2026) | Past performance is not indicative of future results.

The Buildout Extends Far Beyond the Usual Names

As the AI buildout continues, we think investors should consider widening their scope beyond companies at the core of the supply chain to include peripheral contributors likely to play a growing role. In the United States, sectors well outside technology are essential to building and operating data centers. Electrical equipment, cooling, industrial automation, real estate, infrastructure, and vast amounts of energy are all critical to this shift. The expansion is also global. While AI is often associated with American hyperscalers, companies across Asia, Europe and the Middle East are contributing meaningfully to the infrastructure layer. Just as user adoption of this technology has expanded across demographics, geographies, and industries, exposure to it shouldn’t be limited to U.S. tech.

Source: AAM

Why Broad Indices Don’t Solve the Problem

Many investors assume that broad-based indices, such as the S&P 500, Nasdaq-100 or Russell 1000 Growth, offer a way to avoid concentration risk. To the contrary, these indices have become increasingly concentrated in the same handful of mega-cap names, as their top 10 holdings show. Worse yet, they may not deliver the central value chain exposure investors are actually seeking. Key players in the AI ecosystem, including networking and hardware firms, memory and storage companies, semiconductor equipment subsystems providers, and many more are largely absent from the benchmarks. Investors looking for exposure across the full AI value chain, from energy inputs through compute hardware, to end applications, may find broad-based indices currently come up short.

Source: FactSet, Index Fact Sheets (06/30/2026)

Finding the Middle Ground

Resilient AI exposure isn’t found at either extreme. It isn’t found in concentrated leveraged bets, or in cap-weighted indices that quietly concentrate a few names. It requires deliberate spreading across the full value chain, sectors and geographies, while staying selective about which companies actually merit inclusion. As the AI buildout matures and widens, a thoughtfully diversified approach to this shift may prove far more durable than chasing the next short-term rally.

Source: AAM | For illustrative purposes only

CRN: 2026-0803-13666 R