How to Make Money From the Booming Demand for Energy

Episode Summary

Executive Summary: The episode examines how AI, data centers, and global reindustrialization are reshaping infrastructure and natural resources investing. Guest Tyler Rosenlicht argues that energy demand is structurally rising, the market underestimates bottlenecks in power and regulation, and opportunities increasingly lie in utilities, pipelines, nuclear, and “picks-and-shovels” industrials rather than only direct commodity bets.

Main Topics: AI-driven infrastructure demand (Priority: 5/5): The hosts open with Jensen Huang’s comments about improving chip efficiency potentially reducing cooling needs, highlighting how AI investment is now tied to physical infrastructure and can rapidly shift winners and losers. Utilities, rate base, and data-center economics (Priority: 5/5): Rosenlicht explains how utilities can benefit or suffer from data-center load growth depending on regulation, generation mix, and whether added demand lowers or raises local bills. Secular rise in global energy demand (Priority: 5/5): The guest argues that population growth, economic growth, and even slower gains in energy efficiency still point to much higher global energy demand by 2040, forcing more supply buildout. Nuclear as baseload solution (Priority: 4/5): Nuclear is presented as the best fit for the ‘more, stable, clean’ energy triangle because it can provide 24/7 baseload power without the intermittency of wind and solar. Supply discipline, scarcity, and government intervention (Priority: 4/5): Natural resources and mining are described as capital-starved and consolidated; investors still demand discipline, so government catalyst/backstops may be required to unlock supply in uranium, copper, and critical minerals. Infrastructure bottlenecks and regulatory risk (Priority: 5/5): The conversation emphasizes that local politics, utility commissions, permitting, and expropriation risk can make apparently attractive projects fail, especially in Venezuela and in contentious U.S. jurisdictions. Picks-and-shovels winners (Priority: 4/5): Rather than only direct exposure to energy or commodity prices, the guest sees attractive risk-adjusted returns in engineering, construction, and equipment companies that enable buildouts.

Key Arguments: AI and data centers are turning infrastructure into a high-volatility, tech-linked investment theme rather than a stable old-economy one. Utility outcomes are highly location-specific: some service territories can see lower bills when data centers absorb surplus power, while others face affordability backlash and stranded-asset risk. Global energy demand is rising even under optimistic efficiency assumptions, so the world needs substantially more power generation and grid buildout. The market has likely underpriced the growth dispersion among utilities; best-in-class names are not trading at much of a premium despite materially stronger growth prospects. Nuclear is likely to expand because it uniquely combines baseload reliability and low emissions, but the buildout will require government support rather than pure market forces. Capital discipline in mining and shale means supply responses will be muted unless incentives change, which supports higher returns for incumbents and selective new projects. Regulatory risk is the central risk in infrastructure investing, often more important than commodity price direction. Data-center growth and reindustrialization create long-run demand for natural gas pipelines, electrical equipment, and industrial build-out services. Public-private partnerships and direct government intervention are becoming more important as a way to overcome permitting, financing, and overcapacity-cycle constraints.

Data Points: Global energy demand (2040 baseline): 178,000 to 220,000 terawatt hours - Rosenlicht’s model for global energy demand growth over the next two decades, assuming rising efficiency Renewable supply needed to replace coal: 55,000 to 60,000 terawatt hours - Estimated additional renewable generation needed as coal consumption falls while total demand rises U.S. electricity demand growth: 0% from 2007 to 2020; recently ~1.5%; expected ~2.5% - Rosenlicht says demand is reaccelerating across data centers, EVs, and industrial uses Utility rate-base example: $10 billion rate base, 10% ROE = $1 billion earnings - Simplified illustration of how utilities earn returns and recover costs from customers Fastest-growing utilities valuation premium: 6% premium today vs. 11% premium eight years ago - Shows growth is cheaper relative to the average utility than in the past Growth differential in utilities: ~2 percentage points higher growth for top utilities (8.5% vs. 6.5%) - Best-in-class utilities have better growth but are not priced much higher than peers Midwest utility size: 11 gigawatts existing system - Guest uses this as a scale example for a utility with a century of development Potential new data-center demand at that utility: 15 gigawatts - Illustrates how quickly data-center load can exceed historical system scale One gigawatt analogy: About the electricity needs of a city like Denver / about one million people - Used to make the scale of gigawatt capacity easier to visualize U.S. oil price reference: $57 per barrel - Mentioned in discussion of whether Venezuelan or new pipeline projects pencil out

Pivotal Quotes: "“the world needs more energy. We think the world needs more power.”" — Tyler Rosenlicht: Summarizing the overarching demand thesis across data centers, industrials, EVs, and households "“the three factors have changed.”" — Tyler Rosenlicht: Referring to the energy system priorities shifting from clean/stable/more to more/stable/clean "“We’ve exited what we talked about as the era of abundance and we’ve entered the era of scarcity.”" — Tyler Rosenlicht: Explaining why natural resources and infrastructure may generate above-average returns for longer

Implications: Energy, utilities, and infrastructure are becoming central to AI and reindustrialization. Expect more regulation, more government support, selective winners in utilities and nuclear, and strong opportunities in enabling industries rather than broad beta exposure.