Matt Smith, an energy analyst who spent 15 months modeling nearly every asset in the U.S. natural gas system, warns that the price of natural gas could start moving higher in as little as six months, and the setup behind that move is worse than anything in the modern record. Smith laid out the case on the latest Invest Like the Best podcast and in a letter titled “Got Gas.” His base case: “A shortage of natural gas for which there is no precedent starting in the second half of 2028… The fuel everyone thinks is abundant is not… A crisis is coming. Unfortunately, the die is cast.”
That claim lands as Washington celebrates a cooler inflation reading.
Why the Inflation Story Just Changed
The Consumer Price Index sat at 332.6 in June 2026, down 1.4 points from May, a monthly drop of 0.4%. West Texas Intermediate crude cratered from an April peak of $114.58 a barrel to a July low of $69.60, and pump prices followed, sliding from $4.31 a gallon on June 1 to $3.78 by July 6 before ticking back to $4.00. Energy giveth, energy taketh away. Smith’s point is that the giving is nearly done.
Here is the mechanism regular Americans need to understand: Natural gas was the largest source of US electricity in 2025 at roughly 41% of utility-scale generation, with nuclear a distant second near 18%. Gas is the marginal fuel on most of the grid, which means it sets the price at the meter. It also touches food, fertilizer, plastics, steel, cement, and pharmaceuticals. When gas moves, everything downstream of an electric bill or a factory floor moves with it.
The AI-Powered Squeeze
Smith argues the shortage is structural, driven by two demand walls hitting a production ceiling. The first is liquefied natural gas exports, which the Energy Information Administration forecasts will average 17.0 billion cubic feet per day in 2026 and 18.2 Bcf/d in 2027, with capacity heading to 27.7 Bcf/d by 2030. The second is data centers. The Department of Energy projects data centers could reach 12% of U.S. electrical demand by 2028, up from 4.4% of consumption in 2023. That electricity must come from somewhere, and mostly from gas turbines.
Smith’s modeling concludes that the U.S. will begin to deplete natural gas in storage in an unprecedented way in 2028, and likely exhaust working storage by 2030. He is blunt that “this does not portend a modest price increase.” The real bottleneck is upstream deliverability, not power plants.
Current Henry Hub prices offer false calm. Gas sits at $3.15 per million BTU in June 2026, down from a January 2026 spike of $7.72. The EIA forecasts an average around $3.50/MMBtu in 2026 and $3.18/MMBtu in 2027. Smith’s letter argues those forecasts do not account for the demand cliff.
What to Watch
Consumer sentiment already reflects economic anxiety, with the University of Michigan index at 44.8 in May 2026, down from 61.7 a year earlier. A gas-driven electricity and manufacturing shock would hit an audience already flinching. The signal to watch over the next two quarters is the winter Henry Hub strip and storage injection reports heading into October. If injections come in light while LNG export terminals ramp and hyperscaler power purchase agreements pile up, Smith’s six-month clock is running. If gas ticks toward $5 by spring, the CPI story flips from cooling to burning, and the White House will own the next inflation cycle.
Contact [email protected] for any questions or corrections.