Why are utility companies critical to the energy transition?

The short answer

Utilities are not merely beneficiaries of the energy transition — they have become its operational bottleneck. Their grids must absorb intermittent generation, integrate distributed resources, and deliver power to data centers whose load is growing at a pace utilities have never planned for.

The capital reallocation is unprecedented. According to S&P Global Market Intelligence, US energy utility capex rose from $146 billion in 2022 to $173 billion in 2024 and is projected at $215 billion in 2025, with cumulative 2025-2029 spending exceeding $1 trillion across just 47 investor-owned utilities. Yet in 2024, four hyperscalers alone spent more than $200 billion on capex — surpassing the entire utility sector for the first time in history.

The transition has shifted utilities from regulated infrastructure managers to capacity-rationing institutions.

→ New to financial education? Financial education hub

What the data shows

The pace of utility capital reallocation is documented across S&P Global, Deloitte, and Morgan Stanley research:

• US utility capex 2022-2025: $146 Bn → $164 Bn → $173 Bn → ~$215 Bn (+47% over 3 years)
• Cumulative 2025-2029 forecast: over $1 trillion across 47 IOUs (S&P RRA)
• 2026-2030 forecast: approaching $1.3 trillion (S&P, April 2026)
• Renewables share of utility capex: $25 Bn in 2024 rising toward $33 Bn by 2027
• Hyperscaler capex 2024 (Amazon, Microsoft, Google, Meta combined): over $200 Bn, +62% YoY
• Hyperscaler capex 2025 forecast: $371 Bn across the eight largest hyperscalers (Deloitte)
• 2025 utility rate hike requests: $31 Bn (more than double 2024)

The exception worth noting: not all utility capex goes to clean energy. Natural gas remains in expansion plans because data center demand growth exceeds what intermittent renewables plus storage can deliver in current grid configurations. Dominion Energy’s 2024 resource plan adds 5.9 GW of gas alongside 21 GW of renewables and SMRs by 2039.

→ Dataset: Natural gas price history

Why it happens — the macro mechanism

The utility bottleneck has three structural drivers.

Channel 1 — Generation-grid imbalance. Variable renewables require more transmission per MWh delivered than dispatchable thermal plants. The IEA documents annual global grid spending of about $400 billion against roughly $1 trillion on generation — a ratio that needs to converge for the transition to deliver. In the US, transmission permitting can take 10-15 years against 2-4 years for utility-scale solar, creating a structural lag that no amount of capex can bridge instantly.

Channel 2 — Data center demand shock. AI training and inference loads grow at a pace utilities have never planned for. Morgan Stanley projects global power demand rising by more than 1 trillion kWh annually through 2030, with data centers contributing nearly 20% of that growth — about 126 GW per year. The US utility planning model, calibrated for 1-2% annual demand growth, is being stress-tested by 5-10x faster increases in specific corridors (Virginia’s Data Center Alley, Ohio, Texas). This is the most underappreciated dimension of the transition: utilities are simultaneously decarbonizing their generation mix and rebuilding for a demand profile that did not exist five years ago.

A short note on regulatory friction. State-level rate cases must approve cost recovery for new infrastructure, and in 2025 utilities sought $31 billion in rate increases — more than double the 2024 figure — affecting 56 million Americans (PowerLines).

Channel 3 — Cost of capital and regulated returns. Utilities operate under regulated rate of return models. When real interest rates rose sharply in 2022-2024, the cost of new capex rose faster than allowed equity returns could be repriced through rate cases. This compressed forward earnings and made some projects financially marginal even when physically necessary. Bearing that lag rather than merely observing it is what turns a regulated network into infrastructure as an asset class rather than a public works programme.

Synthesis by regime: in the 1990s deregulation era, utilities optimized for cost-cutting under flat demand; in the 2010s low-rate era with stagnant electricity demand (US peak demand barely grew 2008-2020), capex was minimal and dividends robust; in the 2024+ AI-data-center era, demand is surging at 5-10x historical rates, real rates are positive, and the utility model is being stress-tested to deliver capacity at unprecedented pace under its existing regulatory architecture.

Utilities are no longer beneficiaries of the energy transition — they are its bottleneck, and the bottleneck has been priced in.

→ Framework: Macro-financial regimes

What it means for different economic actors

Savers. Electricity bills are rising in regions absorbing data center loads — the Belfer Center documents Dominion’s first base-rate increase since 1992. The transition is materially affecting household energy budgets in specific corridors.

Investors. Utilities sit at the intersection of regulated returns, rising capex, and rate-case approval risk. Their bond issuance is projected to exceed $150 billion in 2026 (Sage Advisory), with implications for the broader investment-grade credit market. Equity investors face the trade-off between earnings growth driven by rate-base expansion and execution risk on permitting and grid construction.

Industrial customers. Hyperscalers and large industrial loads increasingly bypass the public utility grid by signing direct power purchase agreements with renewable developers, or by contracting future SMR capacity (Amazon, Google, Microsoft). This privatizes the grid for those who can afford it and externalizes the cost to public-grid customers.

A common error is to assume utilities are simply passive infrastructure managers. The transition has thrust them into a role that combines capital allocation, capacity rationing, and political negotiation with regulators and large customers — a role for which the institutional model was never designed.

Practical observation

Go deeper

Related questions

• How much investment does the energy transition require?
• Why does intermittency matter for renewable electricity?
• How does battery storage change energy economics?
• What is the nuclear renaissance and will it happen?

Frequently asked questions

How are regulated returns set, and why does this matter for the transition?

State public utility commissions approve allowed rates of return on equity (typically 9-11% currently in the US) plus cost recovery for capital investments. The regulatory lag between when a utility commits capex and when it can recover it through rates can stretch 12-24 months. In a rising-rate environment with surging demand, this lag mechanically compresses utility returns and slows new project commitment, even when grid expansion is physically necessary. Several states have introduced “forward test years” and “trackers” to accelerate cost recovery, but the underlying mechanism remains slow.

Why are hyperscalers building their own power infrastructure?

Public utility grids cannot deliver multi-GW loads on the timelines that AI data center deployment requires. Microsoft, Amazon, and Google have signed power purchase agreements directly with nuclear operators and renewables developers, sometimes including SMR commitments for the early 2030s. This effectively bypasses the regulatory queue and shifts risk from utility ratepayers to the hyperscaler balance sheet. The downside: it fragments grid planning and may externalize transmission costs onto residential customers.

Are utility stocks defensive in this environment?

The traditional view of utilities as defensive bond-proxies has weakened materially. Capex intensity is at historic highs, regulatory approval rates are uncertain, and earnings are increasingly sensitive to rate-case outcomes rather than stable demand. Some investors continue to treat them as defensive due to dividend yields and rate-base growth, while others view them as cyclical capex stories with execution risk. The empirical correlation between utility stocks and 10-year Treasury yields has weakened post-2022 — historically near -0.5, now closer to -0.2 in some periods.

Last updated — 23 July 2026