Bank of America heavyweight report: AI "water sellers" are winning! Where might global funds flow next?

The Federal Reserve is set to cut interest rates, American households have $19 trillion in cash waiting to be allocated, and "water sellers" in the AI industry chain are outperforming tech stocks... Recently, Bank of America released "The RIC Report: The AI enablers are winning," which dissects the current macro trends, investment opportunities, and risks in the AI industry chain. Today, we will help you understand the core viewpoints of this report and see where global funds may flow next.

I. Macro Background: Interest rates are set to fall, $19 trillion in cash is "urgently seeking a place"

Bank of America's latest forecast indicates that, influenced by the weak non-farm data in August, the Federal Reserve may cut interest rates twice this year and three more times in 2026, ultimately bringing rates down to around 3.25%. Currently, American households hold $19 trillion in cash, which is 30% higher than the pre-pandemic trend—cash returns are being "eroded" by inflation and taxes, and this portion of funds urgently needs to find better allocation directions.

From the economic data (see the chart below), August's "soft data" (investor sentiment, consumer confidence, etc.) showed a slight recovery, with the ISM new orders index and the Conference Board Consumer Confidence Index rising to 0.4 standard deviations below the mean; "hard data" (actual economic activity indicators) remained stable, above the long-term mean by 0.3 standard deviations, overall presenting a "weak recovery but not recession" trend.

In the fixed income sector, Bank of America suggests: long-term bonds can serve as a trading allocation, but AAA-rated loan funds, which have lower volatility, are more prudent; the stock market should focus on "AI enablers"—over the past two years, these assets have significantly outperformed the market.

II. AI Enablers: "Invisible champions" outperforming the Nasdaq, but correlation risks are rising

The so-called "AI enablers" refer to sectors that provide foundational support for the AI industry, including power generation companies, infrastructure and industrial enterprises, nuclear energy companies, and pipeline MLPs (master limited partnerships) in the "water seller" sector.

According to the report, over the past two years, these sectors have comprehensively outperformed the tech-heavy Nasdaq 100: for example, utilities, industrial equipment, nuclear energy, and pipeline MLPs not only have high absolute returns, but some areas also show better risk-adjusted returns (such as Sharpe ratios).

However, it is important to be cautious as the correlation between these "AI enablers" and tech stocks is reaching new highs: unregulated power generation companies and natural gas pipeline companies have a correlation of 71% with the Nasdaq 100; The correlation of small and mid-cap (SMID) industrial stocks is 86%; the correlation with grid infrastructure is even higher at 92%; only uranium and nuclear energy companies have a lower correlation (56%).

Bank of America warns that this high correlation means the market is becoming "highly tied to AI demand"—if AI customer spending or hyperscaler capital expenditures slow down, these sectors may face synchronized pressure, posing the risk of "shared prosperity and shared losses."

III. Key Sector Breakdown: Investment Logic in Energy, Industry, Utilities, and Nuclear Energy

To analyze the opportunities driven by "AI enablers" more deeply, the Bank of America analyst team held a roundtable discussion with experts in energy, industry, utilities, and other fields to outline the core logic of each sector:

1. Energy: Natural Gas Welcomes "Dual Benefits," LNG Exports Become a New Engine

Demand Explosion: The demand for natural gas from data centers, combined with the U.S. lifting the liquefied natural gas (LNG) export ban, is driving a reevaluation of natural gas sector valuations. Previously, the market believed that natural gas demand would decline after 2030, but the current acceleration in LNG export facility construction will create a "second growth curve" for the natural gas sector over the next five years.

Geographical Constraints: Data centers need to be located in areas rich in natural gas resources (such as Ohio and Pennsylvania), which limits the risk of oversupply, but the industry is concerned that they may be replaced by nuclear energy or renewable energy in the future, leading to valuation pressure (current contract terms are mostly 10 years, below the industry expectation of 15-20 years).

Policy Dividends: The current government is more friendly to the energy industry, with several pipeline projects in the Appalachian region recently approved, continuously releasing positive signals from the policy side.

2. Industry: AI + Reshoring Dual Drivers, Order Growth Reaches Record Highs

Investment Heating Up: The industrial sector has become one of the most "overweight" areas in the market, with valuations reaching historical highs, driven by both AI (such as data center cooling equipment) and "manufacturing reshoring." Clear Demand: In the next 2-3 years, there will be high visibility of corporate orders, especially in the semiconductor, pharmaceutical, and defense sectors, with clear government strategic support (such as the industrial policies promoted by both parties in the United States).

Truth about Electricity Demand: Many investors mistakenly believe that AI is the main driver of the surge in electricity demand, but in reality, AI only contributes about 20-25% of the demand growth, with more coming from electrification policies (such as heat pumps), the proliferation of electric vehicles, and the return of manufacturing — Bank of America predicts that after 2025, the annual growth rate of electricity demand in the United States will rise from a historical 0.5% to 2.5%.

(From 2015 to 2025, the compound annual growth rate of transmission and distribution capital expenditures for U.S. utility companies is expected to reach 8.7%)

3. Utilities: 6-8% High Growth, Comparable to "Bond-like Stocks"

Supply-Demand Gap: From 2010 to 2020, the electricity supply in the United States grew almost zero, and now the demand for industrial electricity driven by AI and reshoring exceeds GDP growth, while the aging grid and insufficient new capacity (it takes 3 years to build a natural gas power plant, and turbines are queued until 2030) lead to supply tightness.

Investment Value Highlighted: The growth rate of regulated utility companies (such as traditional grid operators) has jumped from 2-4% to 6-8%, with a current price-to-earnings ratio of about 17 times, a dividend yield of 3-4%, and combined with a 7% EPS growth, the annual total return can reach 10%, with a market beta of only 0.5, making it a model of "low volatility and high yield."

(The average household electricity cost rises by 4% annually, currently approaching $18 per kilowatt-hour)

Risk Points: Policy pressure (such as consumer dissatisfaction with rising electricity prices may trigger regulatory intervention), management execution capability (60% of CEOs/CFOs have been in office for less than 4 years, leaving little room for error in large projects) 4. Nuclear Energy: Low Technology Correlation, Strong Long-term Growth Certainty

AI + Carbon Neutrality Dual Catalysis: Data centers require stable clean energy, making nuclear energy the preferred choice; at the same time, under the global carbon neutrality goals, the market size for Small Modular Reactors (SMR) is expected to reach USD 1 trillion by 2050, which can meet about 1/4 of the current global electricity demand.

Low Correlation Advantage: Among all "AI drivers," nuclear energy has the lowest correlation with the Nasdaq 100 (56%), making it a quality target for diversifying technology sector risks.

Key Targets: Bank of America is optimistic about NuScale (signed a 66GW SMR agreement with the Tennessee Valley Authority), Oklo (vertically integrated SMR deployment, 14GW project reserve), and BWXT (U.S. Navy nuclear reactor supplier, received a USD 2.6 billion order).

4. Investment Recommendations: These ETFs Are Worth Attention, Balancing Returns and Risks

For investors with different risk preferences, Bank of America has selected two core ETFs that cover "AI driver" opportunities while controlling risks:

1. Industrial and Infrastructure: High Returns + Low Volatility

AIRR (Small-Cap Industrial ETF): Focuses on small-cap industrial stocks, with a 1-year return of 37.3%, a 3-year return of 31.2%, a risk-adjusted return (Sortino ratio) of 1.35, and a maximum drawdown of 27.9%. It has an "upside beta" of 0.78 and a "downside beta" of 0.68, with greater upside elasticity than downside risk.

PAVE (U.S. Infrastructure ETF): Focuses on infrastructure construction, with a 1-year return of 26.8%, a 3-year return of 23.4%, a Sortino ratio of 1.05, and a maximum drawdown of 26.2%, suitable for investors who prefer "steady growth."

2. Nuclear Energy: Low Technology Correlation, Preferred for Long-term Allocation

URA (Uranium Mining ETF): 1-year return of 80.6%, 3-year return of 25.3%, Sortino ratio of 1.48. Although it has a maximum drawdown of 37.8%, it has the lowest correlation with tech stocks, making it suitable for diversified allocation.

NLR (Nuclear Energy ETF): 1-year return of 71.5%, 3-year return of 32.3%, Sortino ratio of 1.58, with holdings covering nuclear energy operators and equipment suppliers, benefiting from SMR development.

5. Risk Warning: These Three Major Pitfalls Cannot Be Ignored

AI capital expenditure normalization: If ultra-large technology companies (such as Amazon and Microsoft) slow down their AI-related spending, it will directly impact the demand for "AI drivers."

Policy uncertainty: Federal or state governments may adjust energy policies (such as restricting natural gas or increasing renewable energy) or impose limits on utility companies' electricity prices and ROE.

High correlation risk: With the exception of nuclear energy, most "AI drivers" have a correlation with technology stocks exceeding 70%. If the Nasdaq index corrects, these sectors may decline simultaneously.

Summary

The current macro environment (interest rate cuts + excess cash) provides a funding soil for "AI drivers," while the supply-demand gap and policy dividends in sectors such as energy, industrials, utilities, and nuclear energy make them "more certain" opportunities than technology stocks. However, investors should be cautious: avoid overly concentrated allocations in high-tech correlated assets, prioritize ETFs like AIRR, PAVE, URA to diversify risks, and remain vigilant about the potential impacts of policy and capital expenditure fluctuations