Three Uranium ETFs Pitch the AI-Power Trade as Talen and Constellation Lead the Weights

Three Uranium ETFs Pitch the AI-Power Trade as Talen and Constellation Lead the Weights A late-May pitch frames URA, URAN and NUKZ as the cleanest way to play data-center electricity demand, but the underlying signals are running bearish. A widely circulated note published Thursday (2026-05-28) made a blunt case: nuclear power is the only real answer to AI electricity demand, and three exchange-traded funds own the trade. It named the Global X Uranium ETF (URA), the Uranium & Nuclear ETF (URAN), and a third fund trading under NUKZ.³ That matters because the demand side is no longer a forecast traders can wave away. US power generation drawn by data centers is projected to climb from roughly 5% of the total to about 15% over a five-year span, a step change on a grid that has barely grown since 2000.³ The vehicles differ in what they actually hold. URA offers the deepest liquidity and the purest uranium-price exposure, with $6.86 billion in assets.³ NUKZ has returned about 52% over the past year and roughly 11% year to date through late May (2026-05-28), with shares near $71 and an expense ratio of 0.85% against assets of about $841 million.³ Its top weights have included Talen Energy at roughly 3% and Dominion Energy near 3%, alongside Cameco, GE Vernova, and Constellation.³ The clearest evidence that capital is chasing power supply came earlier in the month. Fluence Energy closed at $24.16 on May 8 (2026-05-08), up 98.2% in a single week, after the company disclosed master supply agreements with two hyperscalers and a record $5.6 billion backlog.¹ That is the appetite the ETF pitch is built on.¹ But the Fluence move cuts both ways. The shares are still down roughly 39% year to date, leaving a micro-cap in turnaround territory rather than a clean growth story.¹ The balance sheet carries the strain: stockholders' equity of negative $265.88 million and cash of just $36.59 million.¹ A 98% week does not erase that. The operational signal is steadier than the share price. Fluence delivered positive adjusted EBITDA of $2.0 million in the first quarter of 2026, its fourth consecutive quarter in the black, with non-GAAP gross margin widening to 52%.¹ Chief executive Arun Narayanan said the operational discipline and margin profile established in 2025 are proving durable, and pointed to a PowerTrack platform managing 37.5 GW of solar assets with annual recurring revenue guided to $65 million to $70 million by year-end.¹ The physical logic behind the nuclear case is hard to argue with. A 1 GW reactor occupies a fraction of the land of an equivalent solar build and runs at capacity factors north of 90%, the kind of firm baseload a data center cannot get from intermittent supply alone.³ The contracts are already being signed. Microsoft agreed a 20-year, 835 MW power purchase agreement with Constellation Energy in September 2024 to restart Three Mile Island Unit 1, a $1.6 billion project now targeting a 2027 startup.³ That deal sits behind Constellation's place in the ETF weightings, and it is the template the bulls expect others to copy.³ Policy points the same way over a longer horizon. The US government wants to quadruple nuclear capacity from roughly 100 GW in 2024 to 400 GW by 2050.² Surging demand for clean, reliable power and advances in small modular reactors have revived an industry that not long ago looked like a relic.² Yet the market signals underlying this story are running bearish, not bullish. The packet's consensus read is bearish at full strength across nine signals, with no offsetting bullish weight.³ A 2027 reactor restart and a 2050 capacity target do not pay off a position taken this week (2026-06-02), and the gap between a multi-decade build-out and a fund priced daily is where the risk lives. The thesis also leans on names whose fundamentals are uneven. NUKZ has run hard, but its largest holdings carry their own exposures, and a fund built on Talen, Dominion, Cameco and Constellation is a bet on equity multiples as much as on uranium pounds.³ The pure-uranium exposure in URA is a different trade from the reactor-builder and utility basket in NUKZ, and conflating them is how investors get the direction right and the instrument wrong.³ What to watch is whether the contracting pace picks up from here. One restart deal and one hyperscaler supply agreement do not yet make a trend; a second Three Mile Island-style PPA, or another Fluence-scale backlog disclosure, would.^3,1 Until then, the demand chart is convincing and the entry point is not.