The energy security trade the market has not priced yet

•       The AI infrastructure bottleneck has shifted from chips to power — and the companies supplying that power layer are now the more compelling investment.

•       Even if AI spending slows, electrification of transport, buildings and industry creates an independent, multi-decade demand base for grid infrastructure that requires no further policy catalyst to activate.

•       The Hormuz closure has done in three months what a decade of climate policy could not: made domestic grid investment a national security priority for every major economy simultaneously — and that conclusion is now permanently baked into capital allocation.

•       GRID is currently trading at 21x 2028 earnings against a historical average of 22x — before the market has fully priced a dual-driver earnings cycle with better visibility than at any prior point in the fund's history.

The IEA's Fatih Birol said it plainly in April: "The vase is broken." Not cracked. Not chipped. Broken. Even if a US-Iran deal is signed tomorrow, the world's energy system will not return to what it was. The question for investors is what the new world looks like — and which assets are built for it.

The First Trust NASDAQ Clean Edge Smart Grid Infrastructure Index Fund (GRID) is one answer. It provides exposure across electrical equipment, grid automation, utilities, cables and grid construction — the physical layer that every electron must pass through, regardless of how it is generated or where it is going.

AI data centres: A visible near-term catalyst

Global electricity consumption is rising at its fastest pace in more than a decade — and data centres, the core infrastructure behind artificial intelligence, have become one of the most visible triggers for this acceleration.

AI's power demand is shifting from one-off model training to recurring inference at scale. Inference refers to the computing required each time users generate text, images, code or analysis; each new wave of AI adoption increases the number of queries processed, and even as chips and models become more efficient, rapid growth in usage means aggregate electricity demand continues to rise. The IEA expects global data-centre electricity consumption to roughly double to around 945 TWh by 2030, growing at approximately 15% per year from 2024 — more than four times faster than electricity demand from other sectors.

This is not only a volume story but a structural shift. Next-generation AI chips deliver higher performance but require higher power density and cooling capacity, pushing existing power supply and cooling systems beyond their design limits and forcing upgrades in power management equipment, liquid cooling solutions and data-centre electrical infrastructure. As a result, the core bottleneck in AI infrastructure is shifting from chip supply to power supply and supporting systems — and capital expenditure is expanding accordingly.

Order data is already validating this shift. GE Vernova reported Q1 2026 orders of USD 18.3 billion, up 71% organically year-on-year — with the Electrification segment alone booking USD 2.4 billion of data-centre equipment orders in the quarter, exceeding its full-year 2025 data-centre-related total. The company pulled its USD 200 billion backlog target forward from 2028 to 2027, and its electrification segment backlog has grown from USD 9 billion in 2022 to USD 42 billion today — a 4.7-fold increase in four years that reflects locked-in future revenue, not pipeline speculation.

The clearest sign that power — not processing power — is now the real bottleneck for AI came in June 2026. SoftBank committed EUR 75 billion to build AI data centres in France, choosing it over Germany and the UK for one reason: France generates approximately 70% of its electricity from nuclear power, giving it some of the most stable and affordable energy in Europe. Industrial electricity in Europe costs roughly double what it does in the US — and SoftBank's decision to build there rather than elsewhere tells you exactly how much that gap matters to the economics of running AI at scale.

Perplexity's CEO put it simply at a major technology conference in June: the company that delivers the most value from every unit of electricity it uses will win the AI race. Not the company with the most chips. The one with the cheapest, most reliable power. Grid infrastructure is not a side bet on the AI trade. It is the precondition for it.

This shift in investment logic is already visible in capital flows. Investment decisions in Gulf data centre projects have been paused or delayed because of rising geopolitical risk — two Amazon data centres in the UAE were targeted early in the conflict — and future Gulf facilities will require physical hardening, anti-drone technology and higher insurance premiums, all of which erode the economics that made the region attractive. The war has demonstrated that data centre economics built on cheap, stable Gulf energy are no longer reliable, accelerating the shift of AI infrastructure investment toward jurisdictions with diversified, domestically controlled power grids.

Electrification: The broader structural driver

Data centres have received the most market attention, but they are not the entire logic behind power infrastructure upgrades. The IEA provides an important calibration: data centres are expected to account for around one-fifth of the increase in global electricity demand before 2030 in advanced economies, and closer to half in the US. This means that even in the most data-centre-intensive market, a significant share of incremental electricity demand still comes from other structural forces.

The core logic of electrification is that processes previously powered by burning fossil fuels are increasingly powered by electricity. Electric vehicles are the most visible example, but the transition extends across residential and commercial heating — where natural gas boilers are increasingly being replaced by heat pumps that use electricity to move heat rather than generate it directly, and are typically three to five times more efficient than their gas equivalents (IEA). US air-source heat pump shipments exceeded gas furnace shipments for the fourth consecutive year in 2025 (AHRI), suggesting the heating equipment market is already past an inflection point.

From the generation side, low-cost renewable energy is strengthening the economic foundation of electrification. On an unsubsidised basis, utility-scale solar (USD 38–78/MWh) and onshore wind (USD 37–86/MWh) have ranked among the most cost-competitive new generation options for ten consecutive years, now below gas combined cycle in most scenarios (Lazard, 2025). In 2024, 91% of newly commissioned utility-scale renewable projects globally produced electricity at lower cost than the cheapest new fossil fuel alternative, with solar 41% cheaper and onshore wind 53% cheaper than the lowest-cost fossil fuel option (IRENA, 2024).

The result is a broader investment cycle that extends beyond AI data centres. The IEA characterises the current period as an "Age of Electricity," with electricity demand expected to grow faster than overall energy demand as transport, buildings, industry and data centres become more power-intensive. Even if AI capital expenditure slows, demand for power infrastructure has independent and sustained structural support from electrification, industrial growth, renewable integration and energy security. The Hormuz closure has made that conclusion inescapable for every major economy simultaneously: not as a policy preference, but as a strategic reality baked into capital allocation.

Why more power generation is not enough: the urgency to build grid infrastructure

Faced with rising electricity demand, the intuitive answer is to generate more power — but that answer is only half right. The constraint is not only power generation but also grid infrastructure.

In high-renewables markets, electricity is often generated far from where it is consumed. Solar farms may be located in inland areas where sunlight and land availability are favourable, while load centres are typically cities or industrial hubs elsewhere. When transmission channels are insufficient, the system experiences congestion — electricity prices in generation-rich regions fall sharply or turn negative while load centres face elevated prices because power cannot be delivered efficiently.

This also explains the holding logic of GRID. Its holdings are concentrated across critical parts of the power infrastructure chain: ABB, Eaton and Schneider Electric produce electrical equipment, energy management and automation solutions; Quanta Services undertakes transmission engineering and construction; Prysmian manufactures high-voltage cables; and National Grid and E.ON operate regulated transmission and distribution networks. Regardless of who generates the electricity or how, every electron must pass through this infrastructure layer.  

The cost of skipping grid investment is already showing up in real losses.

In June 2026, BlackRock froze more than USD 1 billion of planned renewable energy investments in Brazil. The reason was not a lack of demand for clean power — it was a grid that could not absorb the electricity being generated. When too much solar power floods the system at once and the grid cannot move it to where it is needed, operators simply switch off the solar farms. This is called curtailment. Projects that get curtailed still have to buy replacement power to honour their delivery contracts — in Brazil's case, at roughly twice the price they were selling their own power for. The economics collapse.

The lesson is simple: renewable energy built without matching grid investment does not produce clean power. It produces losses. 

This is precisely the gap that GE Vernova's electrification business is filling. Its USD 42 billion backlog is not wind turbines or solar panels — it is the grid integration work that connects generation to the places that actually need the power. Without that layer, the energy transition stalls. With it, the investment compounds.

Energy security: The permanent policy shift

The deepest consequence of the Strait of Hormuz crisis is not the oil price spike. It is a permanent shift in how governments think about energy supply.

Before the conflict, around 140 tankers and gas carriers transited the Strait daily. That number fell to roughly 33. Iran has since established a standing institution with authority over vessel transits — this is not a negotiating position that disappears when talks resume. It is a functioning bureaucracy, and it will continue to operate.

Even if the situation eases, the damage is done. ECB Executive Board member Isabel Schnabel said it plainly in May: "even if the war ended today, a lot of damage has already been done to energy infrastructure and global supply chains." Industry research reaches the same conclusion — even in the most optimistic scenario, the global energy system will remain tighter and more fragile than it was before the shock. Full normalisation is not on the table.

Governments have moved from statements to action. The UK has formally designated energy infrastructure procurement as a matter of national security. The head of Ofgem, the UK's electricity regulator, has called grid investment "the biggest transformation of our lifetime." These are not analysts making forecasts. They are the regulators and policymakers who set the terms of investment — and their conclusion is the demand driver.

For investors, the timing matters. Qatar LNG repairs and alternative supply projects carry construction timelines of four to five years at minimum. Until 2030, the only structural substitutes available for displaced Middle Eastern energy are grid infrastructure and storage. This is not a cyclical opportunity. It is a policy-mandated, supply-constrained multi-year build cycle with no shortcut around it.

Why GRID?

GRID provides exposure across the power infrastructure value chain — electrical equipment, energy management and automation, cable manufacturing, grid contractors and regulated utilities. This matters because the power infrastructure opportunity is not limited to one part of the system: AI data centres require power distribution, cooling, transformers and grid connections; electrification requires stronger distribution networks and intelligent load management; renewable integration requires transmission lines, substations and grid automation; and energy security requires more resilient domestic electricity networks.

Chart 1: Top three holdings' backlog continues to show robust growth.

Table 1: Key holdings — Q1 2026 earnings summary

Company	Revenue	EPS	Guidance	Margin change
ABB Ltd	Strong growth	EPS up	Raised	Up 320bps; 250bps from real estate gains — underlying improvement ~70bps
Eaton Corporation	Record sales	Record adj. EPS	Raised	Down 120bps — temporary price-cost lag and factory ramp-up costs
Johnson Controls	Strong growth	Adj. EPS +~40%	Raised	Up 190bps to 12.4% adjusted EBIT margin

Backlog growth across key holdings strengthened through 2025 and into early 2026. Margin trends should be interpreted carefully — ABB's reported margin expansion was partly driven by a one-off real estate gain, while Eaton's margin declined due to temporary price-cost lag and upfront factory ramp-up costs. The sector is in an active scaling phase, not a simple across-the-board margin expansion cycle.

The most advanced real-world proof that this investment cycle is operational rather than theoretical is in Australia — where approximately 60% of global household-scale battery capacity installed this financial year will be deployed, with 415,000 units connected since July 2025. Gas-fired generation fell 24% in one quarter year-on-year as batteries absorbed the evening peak; electricity benchmark prices fell up to 10% in parts of the country. The Hormuz crisis has accelerated this transition globally by making gas dramatically more expensive, and GEV's electrification backlog growth from USD 9 billion in 2022 to USD 42 billion now is the institutional capital response to what Australia has already proven at the consumer level.

The rate headwind: Acknowledged and addressed

Infrastructure equities are rate-sensitive assets. A 30-year US Treasury at 5.06% and a newly installed Federal Reserve chair whose first public remarks emphasised price stability over growth are not a neutral backdrop for a fund with significant regulated utility exposure. We say this directly because the investment case for GRID should survive the objection rather than ignore it.

The counter-argument is structural rather than cyclical. UK grid network charges are contractually committed to rise from GBP7.6 billion to GBP12.1 billion annually by 2029–30, regardless of the rate environment — confirmed by Ofgem. EDF UK's CEO has stated that "even if the wholesale price were to halve, bills will rise." Grid capex is not rate-cycle-dependent capex: it is policy-mandated, contractually locked and driven by physical infrastructure deficits that do not respond to monetary policy. As Jeff Currie of Carlyle framed the broader energy market at the UBS Asian Investment Conference in Singapore: "sell the tweet, buy the molecule." The physical reality does not resolve because a rate decision is delayed.

Valuation

GRID ETF is currently trading at approximately 21x 2028 earnings estimates, below its historical average of around 22x. On this basis, we use a fair P/E of 25x as a scenario-based working assumption, modestly above the historical average. Applying 25x to 2028 estimated EPS of 98.3 implies an estimated target price of USD 228, or approximately 16% upside from the current multiple.

Historically, GRID was primarily driven by a single grid modernisation cycle with more limited earnings visibility and greater cyclicality. The current environment is different: AI-driven data-centre investment, electrification, renewable integration and energy security-related infrastructure spending are now operating simultaneously, with drivers that are not fully dependent on the same cycle. Backlog growth across key holdings has accelerated through 2025 and into 2026, and multiple companies have beaten expectations and raised guidance. This combination of visibility and momentum suggests that earnings quality is improving, not just the earnings level — and we acknowledge that higher long-duration rates compress infrastructure valuations, as addressed in the section above.

For investors with a two-to-three year horizon, the physical reality of the energy transition does not resolve on a rate cycle or a diplomatic timeline — and that is precisely the environment GRID is built for.

Table 2: GRID projection

GRID	FY25	FY26	FY27	FY28
PE Ratio (X)	32.90	27.65	24.20	21.25
Expected Earnings Growth	16.63%	18.98%	14.23%	14.30%
Earning Per Share (EPS)	63.27	75.3	86	98.3
Dividend Yield	1.74%	1.38%	1.48%	1.62%
Target Price (USD)				228
(Based on fair PE ratio of 25X)
Potential Upside (%)				15.97%
Source: Bloomberg Finance L.P., iFAST Estimates
Data as of 3 June 2026