Matthew Smith — Natural Gas: The Next Bottleneck - [Invest Like the Best, EP.483]

2026-07-21 · Source: Invest Like the Best with Patrick O'Shaughnessy · Field: Energy & Utilities — Traditional Energy & Fossil Fuels, Energy Markets & Policy, Utilities & Infrastructure · Depth: Advanced, extended

Summary

Matthew Smith, founder and CIO of Chronometer Partners, forecasts a historic deficit in US natural gas supply starting in 2028. His 18-month analysis, modeling nearly every natural gas asset, concludes that surging demand from AI data centers and planned LNG exports will outstrip production and delivery capacity. The US is projected to exhaust its working natural gas storage by 2030, leading to "unbounded and convex" price increases, acutely impacting electricity costs. LNG exports are set to grow from 15 BCF/day to 35 BCF/day by 2030, while AI compute demand could add 5 BCF/day (base case) to 12-15 BCF/day. This situation was set in motion before AI's rise, with infrastructure constraints and contract law making export curtailment difficult.

Key takeaway

For Directors of AI/ML or Hyperscaler Executives planning future data center deployments, you must re-evaluate energy cost projections beyond 2027. The current flat forward curve for natural gas is complacent; expect significant price increases and supply scarcity. Secure long-term physical gas supply contracts now, or seriously consider distributed solar and large-scale nuclear solutions to mitigate substantial operational cost increases and potential service disruptions by 2028-2030.

Key insights

US natural gas supply faces an imminent, structural deficit by 2028, driven by AI and LNG exports, leading to unbounded price risk.

Principles

• Long-term energy planning must account for committed infrastructure projects.
• Complacency regarding abundant resources can lead to critical shortages.
• Marginal fuel costs dictate electricity prices, benefiting free-fuel sources.

Method

Smith's team modeled US natural gas assets from the well-level up, assigning probabilities to power generation ideas for compute, focusing on P50 assets with approvals or contracts.

In practice

• Prioritize large-scale nuclear for long-term energy stability.
• Invest in residential solar with batteries for electricity price protection.
• Secure physical natural gas supply contracts for future operations.

Topics

• Natural Gas Markets
• AI Data Centers
• LNG Exports
• Energy Infrastructure
• Nuclear Power
• Energy Security

Best for: CTO, VP of Engineering/Data, Executive, Investor, Director of AI/ML, Consultant

Related on AIssential

• Demand Push — US natural gas supply is robust, but rapidly increasing demand, particularly from AI, poses a significant risk of future price crises and shortages.
• The AI Data Center Delusion — The AI data center boom faces critical power grid and supply chain constraints, forcing reliance on fossil fuels despite "Net Zero" pledges.
• Blue Chips — The AI boom is driving demand for "behind the meter" natural gas data centers, despite regulatory hurdles and slow progress.
• What's really slowing down the AI buildout — The primary bottleneck for AI buildout is grid interconnection, not energy supply, due to outdated, inflexible processes.
• More than 100 UK datacentres plan to burn gas to generate electricity — UK datacentres are increasingly turning to gas power due to grid connection delays, jeopardizing climate targets.
• Tariffs, scarce power and a splintered map: the macro forces reshaping technology investment to 2035 — Geopolitical shifts and resource constraints are fundamentally reshaping technology investment and supply chain strategies.

See Counsel's argued verdicts on the open AI decisions leaders are weighing →

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Invest Like the Best with Patrick O'Shaughnessy.