America's power grid faces its biggest test yet

Key Concepts

• Energy Infrastructure: The physical systems for producing and delivering energy, including power plants, pipelines, transmission lines, storage, and on-site systems for data centers.
• AI Economy: An economy driven by artificial intelligence, encompassing technologies like driverless cars, personal robots, and flying taxis.
• Electricity as the New Oil: The concept that electricity is becoming the most critical resource for powering the AI economy, analogous to oil's historical importance.
• Data Centers: Facilities that house computer systems and associated components, such as telecommunications and storage systems.
• Total Return: The overall performance of an investment, including both capital appreciation (stock price increase) and dividend income.
• Dividend Yield: The annual dividend payment per share divided by the stock's current market price, expressed as a percentage.
• Economic Moat: A sustainable competitive advantage that protects a company's long-term profits and market share from competitors.
• Behind the Meter Power: Energy generated on-site for a specific facility (like a data center) and not fed into the public grid, with the facility owner paying for it directly.
• Price-to-Earnings (P/E) Ratio: A valuation ratio of a company's current share price compared to its per-share earnings.
• Hyperscalers: Large cloud computing providers that operate massive data centers.
• Capital Expenditures (CapEx): Funds used by a company to acquire, upgrade, and maintain physical assets like property, buildings, and equipment.

AI Economy and Energy Demand

The discussion centers on how the burgeoning AI economy is poised to dramatically increase electricity demand, shifting the energy landscape. Host Jared Blicker introduces the concept of "energy infrastructure" as a critical, albeit less glamorous, component of the AI trade, contrasting it with the more visible NVIDIA GPUs.

Key Points:

• Electricity is the New Oil: Rob Thumbl, a senior portfolio manager at Tortois Capital, argues that electricity is becoming the most crucial resource for AI, akin to oil's historical significance.
• Data and Electricity are Essential for AI: Both data and electricity are fundamental requirements for AI technologies like ChatGPT to function.
• Projected Surge in Electricity Demand: US electricity demand has been flat for two decades but is expected to accelerate significantly due to AI.
• Sam Altman's Projection: OpenAI CEO Sam Altman estimates that data centers could require 250 gigawatts of power by the early 2030s. This is approximately one-fifth of the current US electricity capacity.
• AI Development Bottleneck: Thumbl and other tech executives suggest that electricity supply, not computing power, is the primary constraint on AI development.
• Comparison to Shale Revolution: Thumbl likens the current AI-driven electricity demand surge to the shale revolution, which transformed the US into the world's largest oil and gas producer.

Energy Infrastructure: The Nuts and Bolts of AI

The phrase of the day, "energy infrastructure," is defined as the physical systems that produce and deliver energy, from power plants to transmission lines and storage. This includes crucial on-site systems for data centers, such as cooling and backup power.

Key Points:

• Criticality of Infrastructure: There is no AI without robust infrastructure.
• Investor Opportunities: The evolving energy sector presents significant opportunities for investors.
• Data Center Cooling and Power: Specialized systems for data centers are a key area of focus.

Investment Ideas in Energy Infrastructure

The podcast highlights specific companies and sectors poised to benefit from the increased demand for energy infrastructure driven by AI.

Real Estate Plays and Company Transformations

Riot Platforms (RIOT):

• Transformation from Bitcoin Miner to Data Center Operator: Riot Platforms, initially focused on Bitcoin mining, is pivoting to become a data center operator due to its access to significant power.
• Business Model Shift: The company is transitioning from a commodity-based business to a long-duration, fee-based model.
• Investment Needs: Riot needs to build physical data centers.
• Future Plans: The company is expected to announce plans for massive data centers near its electricity generation sources in Texas, likely backed by contracts with hyperscalers.

Evergy Corporation (EVRG):

• Powering Data Centers in Kansas City: Evergy, a utility provider in Kansas City, is experiencing accelerated growth due to the influx of data center development.
• Slow Growth Utility to Growth Accelerator: Historically a slow-growth, "boring" utility, Evergy is set to significantly increase its growth rate.
• Capacity Expansion: The company may double its electricity generation capacity to power these new data centers.
• Projected Growth Rate: Evergy's growth rate is projected to rise by approximately 50%, potentially leading to a higher stock valuation multiple.

Dividends and Stable Returns

The discussion shifts to dividends as a key component of total return, often overlooked in favor of stock price appreciation.

Key Points:

• Dividends as a "Second Cousin" to Buybacks: Dividends are cash payments from a company to shareholders, usually from profits, and are taxable.
• Total Return vs. Share Price Performance: Total return, which includes dividends, provides a more accurate picture of a stock's performance than just share price gains.
• Energy Sector as a Dividend Haven: The energy sector has historically been a strong source of dividends.
• Dual Benefit of AI and Dividends: Investors can potentially benefit from both income generation through dividends and exposure to AI growth.

Williams Companies (WMB):

• Pipeline Operations and Stable Cash Flows: Williams is a pipeline company with stable, consistent cash flows generated from transporting natural gas.
• Dividend Payouts: These consistent cash flows enable the company to pay regular dividends to shareholders.
• AI Infrastructure Involvement: Williams is building electric generation facilities (behind the meter) to support Meta's Socrates data center in New Albany, Ohio.
• Dual Investment Appeal: Investors receive a dividend yield from its pipeline business and exposure to growth potential from its AI-related electric generation.

Dividend Investment Strategy:

• Long-Term Horizon: The ideal timeline for dividend stock investment is "forever," provided the dividend is consistent and covered.
• Dividend Coverage and Growth: Investors must ensure the dividend is sustainable and ideally growing.
• Cash Flow Volatility: Companies with less volatile cash flows are more likely to maintain steady dividends.
• Economic Moats: Thumbl emphasizes looking for companies with strong "economic moats" – sustainable competitive advantages that protect their cash flows. Infrastructure assets, often regulated, tend to have these moats.
• Dividend Yield: This is the annual dividend as a percentage of the stock price, similar to an interest payment from a bond.

The Jaw-Dropping Numbers of AI Power Needs

The conversation revisits the immense power requirements for AI, emphasizing the scale of the challenge.

Key Points:

• 250 Gigawatts by Early 2030s: This projection from Sam Altman highlights the massive future demand for electricity for data centers.
• Realistic Projections: Thumbl believes these numbers are realistic, as electricity supply is the current bottleneck for AI development.
• Comparison to "Back to the Future": The 250 GW figure is 200 times the 1.21 gigawatts mentioned in the movie.
• Avoiding an Overbuild: Unlike the 2001 fiber optic buildout, Thumbl believes an overbuild is unlikely due to the strong cash flows of hyperscalers, who are funding these investments with cash rather than debt.
• Hyperscaler Cash Flow: Major tech companies are generating significant free cash flow (e.g., $250 billion this year) to fund data center expenditures.

Consumer Impact and Pricing

The potential for consumers to bear the cost of these energy buildouts is addressed.

Key Points:

• Avoiding Retail Price Hikes: Utilities and data center companies are aware of the need to avoid passing increased costs onto retail consumers.
• "Behind the Meter" Solutions: This model, where data centers pay directly for their energy generation, shields regular consumers from higher prices.
• Special Rate Structures: Utilities may implement special rate structures for large electricity users like data centers.
• Complexity and Nuance: The pricing mechanisms are complex but aim to protect the retail consumer.

Other Investment Avenues

Beyond direct energy infrastructure, other related sectors are discussed.

Key Points:

• Pure Play Natural Resources: While gold is not directly part of the AI trade, silver might be. Natural gas and nuclear power are seen as future energy supply sources.
• Natural Gas Companies: Companies like EQT and Expand Energy are mentioned as direct plays on natural gas.
• AI Enablers: Companies that produce data storage devices, cooling systems, cables, and network switches are also expected to see significant growth tied to data center buildouts.
• Massive Data Center Scale: Meta's planned data centers, the size of Central Park, illustrate the immense infrastructure needs.

Market Strategy: Reading the Green vs. Reading the Table

A segment comparing investment strategies to golf and poker skills is presented.

• Reading the Green (Fundamental Analysis): Studying the business, demand growth, operating costs, and cash flow sustainability to support dividends. This is Thumbl's preferred approach.
• Reading the Table (Market Sentiment/Technical Analysis): Observing investor behavior, crowd dynamics, and momentum to inform entry and exit points. This is seen as more tactical.

Key Points:

• Thumbl's Preference: Thumbl favors "reading the green" (fundamental analysis) and focuses on long-term bets.
• AI Enablers' Growth: The growth of AI enablers mirrors the capital spending growth of hyperscalers, projected to reach $3-4 trillion by 2030.
• AI Tech Bubble Concerns: The podcast addresses concerns about an AI tech bubble.
• P/E Ratio as an Indicator: The P/E ratio is used to assess whether stock prices are outpacing earnings. Current P/E ratios are not at bubble levels seen in 2001.
• Earnings Keeping Pace: Earnings have kept pace with stock price performance in the current AI-driven market.
• Investment Strategy: Look for stocks with low P/E ratios and growing earnings rates.

Conclusion and Takeaways

The episode concludes by reiterating the importance of energy infrastructure and dividends in the AI economy.

Key Takeaways:

• Energy Infrastructure is Key: Focus on the less glamorous but essential energy infrastructure plays.
• Dividends Matter: Recognize the importance of dividends for stable returns and compounding.
• No AI Overbuild: The current AI buildout is being funded by cash, not debt, mitigating the risk of an overbuild similar to 2001.
• Fundamental Analysis is Crucial: Prioritize understanding the underlying business and its cash flows for long-term investment success.