Executive orders and AI: Why the future of energy is not what it used to be

Far from easing into the year, energy market followers could be forgiven for asking why we aren’t already on Easter break. Here, we analyse the root causes of recent volatility, and consider the future prospects – both good and bad – for companies across the energy landscape.

First, in a collection of executive orders titled ‘Unleashing American Energy’, US President Donald Trump announced sweeping changes affecting the energy value chain, from traditional to transition. Many of these measures were well-trailed during his presidential campaign, and few would disagree that political support for the energy transition has been under pressure for some time. In some respects though, the past few weeks may be looked back on as the point at which the dam finally broke.

Chinese artificial intelligence (AI) player DeepSeek also up-ended accepted ideas about the cost, chip requirements and therefore power consumption required to train large language models. From the perspective of energy producers and the related supply chain, the wild market moves which followed revealed two things. 

First, the huge dispersion in valuation sensitivity to AI-related news, even among energy supply chain companies that offer broadly the same products and services. Second, the fact that there is a difference between companies that are viewed by the market as structural beneficiaries of a theme, and those which may act as a bottleneck to the theme itself. In a rising tide of excitement, both sets of companies will perform, but it is when the tide goes out that the latter category is truly exposed.

Heightened dispersion calls for an active approach

Most of all, these events reiterate that an active approach and attention to detail are needed. Looking back at January 2025, the utilities sector provided both the best and the worst performing stocks in the S&P 500 index, with a dispersion of 65%. The index’s best performer, Constellation Energy, rising 34% is made all the more notable given the shares fell 20% on the day that DeepSeek’s AI model gained worldwide attention. 

Taking the aggregate picture, and judging by UBS’s AI value chain baskets, power-related names fell furthest in the week following DeepSeek. They underperformed AI beneficiaries, AI software pioneers and AI semiconductors, showing that technology themes are driving asset prices far beyond their conventional area of influence. However, clearly much is still to be determined in the coming months. The potential changes to US clean technology tax incentives will only be addressed when the budget reconciliation bill is finalised (planned for the Spring), while the implications for AI training-related power demand will likely take months, if not years, to fully understand. 

For now though, we can start to separate out some areas of the US energy landscape that have clearly changed, some that face uncertainty and, by elimination, some that remain broadly the same.

What has changed?

Positive for US midstream/liquefied natural gas (LNG)

What?

Accelerated permitting and planning approval for large-scale energy projects. An executive order dating back to May 1977, which saw President Jimmy Carter introduce Environmental Impact Statements, was revoked. Measures are now being put in place to allow agency heads to issue ‘permits by rule’. 

Our view

In practice, this removes some of the barriers to the construction of interstate energy pipelines and other critical energy infrastructure (such as LNG projects). While state-level consideration can still provide a block, the aim here is very clearly to reinvigorate large-scale pipeline investment which has been under increasing environmental scrutiny in recent years in the US.

Positive/neutral for US oil and gas producers

What?

Boosting oil and gas production. Under the pretext of an ‘energy emergency’, the executive order aims to encourage energy exploration and production on Federal lands, including the Outer Continental Shelf, which had been banned by the previous administration.

Our view

This is not necessarily a clear positive for oil and gas producers, because they invest based on the economics of their projects, while Trump’s very clear message from his Davos address and other speeches has been his desire for oil and gas prices to be lower. This plays into geopolitics and the interaction with Russia and Saudi Arabia in particular, and may turn into a net-negative for oil producers if international benchmarks are under pressure. However, based purely on the executive order, the desire is to boost US producers through de-regulation.

Negative for electric vehicle (EV) manufacturers 

What?

Eliminate the ‘EV mandate’ and promote consumer choice around vehicles. 

Our view

In practice, this involves scrapping the US$7,500 consumer EV subsidy and removing emissions-based sales quotas for autos. It is difficult to imagine this being anything other than negative for EV adoption, and will likely compound the slowdown in EV roll-out that has been evident in vehicle manufacturer capacity announcements over the past couple of years.

Negative for offshore wind developers and supply chain

What?

All federally controlled Outer Continental Shelf sea-bed is withdrawn from future leasing for wind projects. While the executive order states that existing sea-bed leases are unaffected, it also introduces an ecological, economic and environmental review of all planned and under-construction projects.

Our view

Although heavily trailed during the election campaign, the apparent threat to planned or under-construction projects is, at the margin, incrementally negative for developers, who had messaged strongly that existing approvals would be honoured.

What remains the same?

Structural growth in utility-scale solar

What? 

A decade ago, the US had installed only a few gigawatts (GW) each of utility-scale and residential (rooftop) solar. Since then, falling costs have driven a 26% decade-long compound annual growth rate in utility-scale installations, which now make up approximately 80% of the US solar market.

Our view

Given the longstanding bi-partisan support for the baseline tax credits supporting utility-scale solar, at this point it seems most likely that they will remain in place, implying the growth runway for US solar, and associated storage, remains intact.

Structural growth in investment in networks, driven by electrification and grid hardening

What?

The independent bodies that forecast the US’s power load are expecting a boom in demand. The North American Electric Reliability Corporation (NERC) recently noted that its 2028 load growth forecast for the US has doubled versus 2022 levels. Even the most conservative estimates expect this to drive an approximate doubling of US power grid transmission miles over the next 25 years.

Our view

Effectively every major economy is being forced to re-think the level of attention it applies to its power networks. This is partly a result of anticipated load growth from electric vehicles, and more recently, coping with increasing AI data centre demand. However, at its core, it is about renewing ageing infrastructure, which is critical to the functioning of the economy but has broadly been neglected. Judged on the statistics, the US has among the least reliable grids of the world’s major economies, and is geographically prone to damaging natural disasters, as evidenced by the recent tragic wildfires in California. Grid investment appears to be one of the most durable themes of the coming decades.

Gravitating towards natural gas as the ‘least-bad’ power generation fuel

What? 

Five years ago, the aggregate expectation of US natural gas-fired power plant operators was that their fleet would shrink by the end of the decade. Each year since, they have revised up their expectations, such that they now expect a net 9.6GW of capacity expansion.

Our view

All power generation technologies have their drawbacks. Renewables are intermittent, nuclear is expensive and takes an age to build, coal is ruinously polluting. For countries with abundant domestic reserves, such as the US, natural gas is increasingly looking like the ‘least bad’ option to meet demand growth. It is comparatively quick to construct, reactive to demand, low cost and less polluting than coal. The natural gas equipment supply chain has been sending this message for some time, and recent evidence only serves to support it.

Tax credits for sustainable fuels

What?

Somewhat bizarrely, ‘45Z’ tax credits to support sustainable fuels were only just finalised in the last days of the Biden administration. They provide a sliding-scale deduction for different types of clean transport fuels, such as first and second-generation biofuels, and e-fuels.

Our view

Partly because of the link between agriculture and bio-fuels, and partly because of their comparative complexity and low economic cost, there is likely to be enough bi-partisan support for these credits to remain in place. This means the US renewable diesel market – which transforms used cooking oil, among other things, into transport fuel – can continue to grow.

Uncertainty: the only certainty?

This list is not designed to be exhaustive, and it is doubtful that any can be, given the uncertainties raised by the executive orders of 20 January themselves. The full phrase referenced in the title of this article is “the problem with the future is that it is no longer what is used to be”, from a 1937 essay by French philosopher Paul Valéry. 

He was writing about how rising global geopolitical tension in the wake of the First World War was combining with technological and scientific acceleration to create new types of volatility, which were hard to understand because they had never been seen before. Does this sound familiar? His ideas seem as relevant now as any time in the last century, and energy market observers in particular should take note of his central idea: the only thing we can be certain about is uncertainty.

 

The value of investments will fluctuate, which will cause prices to fall as well as rise and investors may not get back the original amount they invested. Past performance is not a guide to future performance. The views expressed in this document should not be taken as a recommendation, advice or forecast.