For years, electricity prices tended to move in line with natural gas prices because gas-fired power generation often set the market rate. While that relationship still exists, new factors — like grid conditions — are causing power prices to rise even when gas prices fall. Fig. 1-1 shows 2027 future power prices at a 3-year high, highlighting that electricity is increasingly priced based on the condition of the grid itself — rather than fuel costs.

At the core of this is how power markets are designed:

• Electricity is priced using a marginal system where generators submit bids and the system operator dispatches them from lowest to highest cost.
• The final, most expensive unit needed to meet demand sets the overall price for all generation.
• These bids include not only fuel costs but also costs that reflect the efficiency of the specific generator, operating expenses, start-up risk, margin targets, and environmental compliance costs.
• Prices increase when higher-cost units are used, regardless of where natural gas prices are trading.

Due to the rapid load growth we’re experiencing in the U.S. and a tightening supply stack, higher-cost, less efficient generators are now needed more frequently. Rather than modern, lower-cost units consistently setting electricity prices, older or less efficient resources are clearing more often, which pushes power prices up — even when fuel costs are relatively subdued.

As data centers, electrification, and industrial growth drive load increases, demand is rising faster than new generation and transmission infrastructure can be built. This has led to a more constrained system, with less supply flexibility and prices that react sharply to grid stress points. Consequently, electricity prices are increasingly shaped by factors like capacity, location, and reliability, rather than just fuel costs.

Reserve generation plays a critical role in managing sudden demand increases or unexpected outages. However, this safety buffer is shrinking in part due to:

• Retirement of older generators
• Delays in bringing new generation online
• Accelerating demand growth

As reserve margins tighten, the system becomes more reliant on high-cost backup generation and faces a greater risk of shortages. These conditions not only drive up electricity prices but also make them more unpredictable, as markets start to account for risks to reliability along with standard production costs.

When the system becomes especially tight, market rules allow for scarcity pricing, where prices can rise sharply to signal the need for additional supply to ensure reliability. These price spikes are intentional, but they further reinforce the disconnect with natural gas, as they’re triggered by system stress rather than fuel movements.

The physical constraints of the grid also play a larger role in driving electricity prices. Electricity can’t be transported flawlessly, and when transmission lines are congested, cheaper sources of power may not be able to reach the areas that require them. In those moments, the system must rely on more expensive local resources, causing prices to spike in constrained areas and diverge across regions. Consequently, the cost of delivering electricity has become an increasingly significant factor when determining prices.

Policy-related costs add another layer of complexity. For example, carbon market programs like the Regional Greenhouse Gas Initiative (RGGI) — which Virginia recently joined — require generators to buy allowances for their emissions, and those costs are incorporated directly into their market offers. This raises electricity clearing prices independently from fuel, further distancing gas and power markets.

These structural changes are now being embedded into forward markets. Power curves are pricing in the risks associated with congestion, tight reserves, and future demand growth, while natural gas curves are still more closely tied to commodity fundamentals. That divergence is why the two markets are no longer moving in tandem.

This marks a meaningful shift that isn’t likely to change any time soon. Electricity prices can no longer be understood or forecasted through a gas-only lens. Instead, they must be viewed through the broader context of grid conditions, infrastructure constraints, and evolving market design.

During the recent contract expiry period, the May 2026 contract showed minimal volatility, with an absolute price spread of 15 cents over the final 3 days of trading. The contract reached a high of $2.63 and a low of $2.48 before settling at $2.559.

The current price structure has a strongly bearish set-up. The front of the curve is in a strong contango, as the cash market searches for demand and signals producers to hold back supply.

Cash prices in the Appalachian region are nearing levels below $2, prompting EQT and Chesapeake Energy to “strategically curtail” production in areas where both demand and prices are low.

As a result:

• U.S. natural gas production has declined by about 2 billion cubic feet per day (Bcf/d).
• Output fell from an average of 108.6 Bcf/d during the final week of March to 106.4 Bcf/d in the final week of April.
• While much of this decline is due to maintenance activity, a meaningful portion appears to be driven by deliberate production curtailments.

Across most of the U.S., only a few regions saw April's first-of-month index prices settle above the Henry Hub price, mostly limited to the Southeast, from eastern Louisiana through Florida. Because temperatures were mild throughout April, Florida stood out as the sole market consistently keeping a premium over Henry Hub. In stark contrast, West Texas pricing remained distressed, with spot prices averaging under minus $8 — meaning producers had to pay others to take their excess gas supplies.

We began the natural gas injection season with 1.828 trillion cubic feet (Tcf) in storage, close to the 5-year average of 1.815 Tcf.

As of early May:

• Storage balances stand at 2.205 Tcf.
• This is 139 Bcf above the 5-year average.
• The surplus is largely driven by mild weather conditions in March and April.

April 2026 extended the trend of warmer-than-normal spring temperatures. The month ranks as the second warmest since 1950 based on gas-weighted heating degree days (HDDs), surpassed only by April 2017.

Key highlights:

• Temperatures averaged 5 to 7 degrees above normal across parts of the central and eastern U.S.
• Only the far northern edge of the central plains finished the month cooler than normal.

Temperatures have been noticeably cooler than normal in early May:

• The first half of the month is on track to be in the top 10 coolest since 1950 in terms of power-weighted cooling degree days (CDDS) across the Midwest.
• Eastern parts of the U.S. are averaging as much as 5 to 8 degrees below normal to start the month.

However, current weather models suggest a shift:

• A warmer pattern may develop in the second half of May.
• This could push highs into the lower 80s across the Midwest and East.

El Niño forecasts have been a notable story in the weather world recently. Probabilities of an El Niño developing later this year are pushing northward of 90% based on the latest updates from the Climate Prediction Center. Some models are suggesting a rather strong event, which has historically correlated warmer than normal in winter months across the U.S