Global X Uranium ETF
Global X (Mirae Asset)
Tracks the Solactive Global Uranium & Nuclear Components index — the broadest single-ticker uranium play, spanning miners plus nuclear-component and physical-uranium exposure.
- Reconstructed — Figures are third-party estimates rebuilt from filings, not reported results.
- Concentrated — Top disclosed positions are 50%+ of the disclosed book.
Top holdings
Dominant single position; world's #2 producer
Advanced/SMR reactor developer — extends URA beyond pure miners
Canadian Athabasca developer (Rook I)
US-focused producer
Direct physical-uranium exposure inside the equity ETF
World's largest producer (Kazakhstan)
Performance
A nuclear-equity reset rather than a commodity one: U3O8 spot held ~$86/lb through late July while URA fell 10.59% and the S&P 500 was -0.1%. Damage ran wider than the SMR sleeve — Cameco (23% of the fund) fell on a Q2 profit retreat, NuScale -23% and UEC -18% on the month. Index weights barely moved.
Recent moves
Rode the 2025-26 nuclear re-rating — uranium spot back above $100/lb, Cameco up sharply, supply cuts at McArthur River/Kazatomprom deepening a structural deficit, plus US reactor-funding and fuel-security policy driving strong inflows.
Our take
The default one-ticker entry to the AI-power-demand uranium trade — broad and liquid, but a ~25% Cameco weight plus the Oklo/physical-trust sleeve means you're buying the whole nuclear-fuel-cycle complex, not a clean miners basket.
The design decision that matters most is what 'nuclear components' was allowed to mean.
As of Jul 17, 2026 the fund holds BHP Group at 3.29% (a diversified iron-ore/copper major whose uranium output is a rounding error), Samsung C&T at 3.25% (a Korean construction and trading conglomerate), and Oklo at 5.91% (pre-revenue, no fully NRC-approved design, first commercial power targeted late 2027/early 2028). That is roughly 12% of the fund whose earnings are driven by something other than uranium. Meanwhile Cameco alone is 23.03% — above the index's own 22.50% pure-play ceiling, which is what between-rebalance drift looks like. So the honest description is: a quarter is one company, an eighth is not really uranium, and a twentieth is a closed-end trust that trades at its own premium or discount to NAV and layers a second fee under the 0.69%. That is a defensible portfolio, but it is not the clean miners basket the ticker implies, and it is not a spot proxy. A note on the standard leveraged-ETF critique: it does not apply here — URA has no daily reset, so there is no compounding decay from a leverage multiplier. What does apply is raw dispersion. A 52-week range of $35.64–$62.28 against a $38.67 close on Jul 20, 2026 means the fund gave back roughly 38% from its high, and at that volatility the compounded outcome an investor actually realizes sits far below the average of the annual prints. That is a separate point from the since-inception figure, though: the 10-year (+15.75%) and since-inception (-2.51%) numbers are both already compounded, so the gap between them is not volatility drag at all — it is entirely window selection. The 10-year window starts after the post-Fukushima bottom; the since-inception window contains the collapse itself. Any explanation that blames the negative long-run figure on volatility rather than on the 2011–2016 destruction is describing the wrong mechanism.
Thesis
URA is sold as the one-ticker uranium bull case, and on liquidity and breadth it delivers. Our read is the label oversells the exposure in two directions at once. It is not a uranium-price instrument — it is a high-beta equity basket whose largest position (Cameco, 23.03% as of Jul 17, 2026) is a liquid large-cap available commission-free at most brokerages, and whose most volatile 2026 positions (Oklo 5.91%, Centrus 3.16%) are reactor stories with almost no economic link to the price of U3O8. The number marketing never leads with is the one we think defines the vehicle: since its Nov 4, 2010 inception URA has compounded at roughly -2.5% to -3.4% a year (issuer NAV -2.51% through Jun 30, 2026; StockAnalysis -3.41% through Jul 20, 2026). Fifteen-plus years, negative. That is not an artifact — it is the strategy's signature: a commodity-equity basket launched four months before Fukushima, where the drawdowns have been deep enough that the violent recoveries have still not paid for them.
URA tracks the Solactive Global Uranium & Nuclear Components Total Return Index — a modified market-cap index spanning uranium miners, physical-uranium vehicles, and nuclear-component/fuel-cycle names, held in 52 positions as of Jul 17, 2026. The index caps any pure-play name at 22.50%, caps the aggregate of pure-plays weighted 5% or more at 47.50%, and caps all remaining pure-plays at 4.75%. It is fully unlevered and physically replicating: no daily reset, no swaps, no futures roll, no leverage multiplier. Its only quasi-direct commodity sleeve is Sprott Physical Uranium Trust at 5.41%. Expense ratio 0.69%; net assets $5.38B as of Jul 17, 2026.
Assessment
- Genuine breadth and liquidity: at $5.38B (Jul 17, 2026) it is the largest vehicle in the category by a wide margin, giving it tight spreads and real capacity — a structural advantage over sub-$1B peers in a sector prone to liquidity air pockets.
- Access solves a real problem. Kazatomprom (5.23%) and Canadian and Australian small-caps are awkward or impossible for a US retail account to hold directly. The wrapper does work here that a DIY basket cannot easily replicate.
- The concentration caps are a real discipline, not decoration — capping pure-plays at 22.50%, the aggregate of large pure-plays at 47.50%, and the remaining pure-plays at 4.75% prevents Cameco from swallowing the fund outright during a re-rating, which unconstrained cap-weighting would allow.
- The fuel-cycle breadth is defensible on its own terms: enrichment and conversion (Centrus 3.16%) and fabrication capture the bottleneck that Western fuel-security policy is actually trying to fix, which mining exposure alone misses.
- The since-inception record is the strongest evidence against the vehicle and it is almost never shown. Roughly -2.5% to -3.4% annualized since Nov 2010 tells you that thematic commodity-equity funds are launched at peak enthusiasm, and that in this sector the busts have consistently outweighed the booms over a full cycle.
- Theme drift dilutes the stated exposure. BHP (3.29%) and Samsung C&T (3.25%) are in a uranium fund for classification reasons, not economic ones. Someone seeking uranium exposure gets ~6.5% in an iron-ore major and a Korean conglomerate.
- 0.69% is expensive relative to what it buys. The single largest exposure, Cameco at 23%, is directly and cheaply accessible. VanEck's NLR sits materially cheaper. A holder is paying an active-adjacent fee for an index approximable with four or five liquid names.
- The Sprott trust sleeve is a fee-on-a-fee and a tracking wildcard — a closed-end vehicle whose premium or discount to NAV moves independently of both uranium and the rest of the portfolio, adding noise that is not priced by most holders.
- The SMR sleeve is a different asset class wearing the same label. Oklo carries no revenue, no fully approved design, and a 2026 cash outflow guide of $430–550M against March liquidity. It re-rates on narrative, not on uranium, and it made the fund more volatile without making it more uranium.
- 2026-07Jul 27, 2026: Cameco sits at 23.05%, up from 23.03% on Jul 17 — the 22.50% pure-play ceiling stayed breached straight through the July drawdown. Sustained overshoot reads as drift, not discipline.
Record
The trailing numbers are unusually dependent on the endpoint chosen, and that is itself the finding. Issuer NAV through Jun 30, 2026: 1yr +17.61%, 3yr +31.91% annualized, 5yr +20.17%, 10yr +15.75%, since inception -2.51%. Three weeks later, StockAnalysis shows 1yr at -0.92% through Jul 20, 2026, with the price at $38.67 against a 52-week high of $62.28. An 18-point swing in a trailing one-year figure over three weeks means any single performance print here describes the measurement date, not the strategy. The attribution is where the fund's real character shows. Over the same trailing year, uranium did not fall — spot was up about 20.7% year-over-year at $85.70/lb on Jul 17, 2026 ($85.74 on Jul 19), and the long-term contract price stood at $95.50/lb as of Jun 30, 2026, its highest since 2008 (Cameco, from UxC/TradeTech data). So the commodity rose while the equity basket went nowhere. The drag came from a sentiment reset in the reactor-technology sleeve as investors repriced timelines: Oklo down roughly 46% year-to-date and about 72% from its 52-week high as of mid-July 2026, and Cameco down about 37% from its 52-week high of $135.24 to $84.85 on Jul 20, 2026. Read plainly: this is equity-sentiment behavior, not commodity behavior — which is the sharpest available evidence for our thesis. Separately, the 3-, 5- and 10-year annualized figures are measured from post-Fukushima lows and flatter the strategy; the since-inception figure measures from a launch high and is unflattering. Both are true, which is why neither should be used alone.
- 2026-07Jul 31, 2026: URA fell 10.59% in July and 30.75% over three months while U3O8 spot held near $86/lb. The basket did not go nowhere against a flat commodity — it fell hard. The decoupling is sharper than described.
- 2026-07Jul 31, 2026: Oklo closed $38.83 against a 52-week high of $193.84 — about 80% below, not 72%, and within 6% of its $36.61 low. The reactor-sleeve drawdown is roughly 8pp deeper than the mid-July reading.
Risks & fit
- Equity beta amplifies in both directions, and decouples. Uranium spot rose ~20.7% year-over-year while URA's one-year return went slightly negative — miners and reactor names carry operating, financing, and narrative leverage on top of the commodity, so a rising uranium tape does not guarantee a rising fund.
- Single-name concentration. At 23% of assets, a Cameco-specific event — an operational failure at McArthur River, a contract-book disappointment, a Canadian tax or regulatory decision — moves the fund independently of the uranium thesis.
- Sovereign and geopolitical exposure. Kazatomprom (5.23%) is a Kazakh state-controlled producer with Russian logistics dependencies; sanctions, export routing, or a change in state production policy is a live and largely unhedgeable risk.
- Timeline risk in the SMR sleeve. Only two SMR systems are operable worldwide today (China and Russia). If Western first-power dates slip past 2028, the reactor-technology names re-rate down again without any change in uranium fundamentals.
- Utility contracting behavior. Most uranium trades under long-term contracts; if utilities stay covered and stay out of the spot market, the headline price everyone watches can stagnate even with a real structural deficit.
- Realized-return dispersion. With a 52-week range of $35.64–$62.28, the compounded outcome over any holding period sits well below the average of the annual prints — a fund of this volatility requires a much larger cumulative gain to recover a given drawdown.
Our skepticism is that URA behaves as a sentiment-driven equity basket rather than a claim on a structural uranium deficit. The cleanest evidence for that read right now is the decoupling: spot up ~20.7% year-over-year and long-term contract at a post-2008 high of $95.50/lb (Jun 30, 2026), while the fund's one-year return is roughly flat-to-negative. What would break our read: long-term contract prices ratcheting further above the $95.50/lb level on rising utility contracting volumes, with Cameco converting that into realized-price and earnings growth — such that the fund's largest weight rises on delivered cash flow while the reactor-technology sleeve is flat or down. If URA advances while Oklo-type names do not, the thesis is earnings-driven and our critique is wrong. Second falsifier: a full cycle from here in which the since-inception annualized figure turns durably positive would show the busts no longer outweigh the booms. Conversely, if the fund's next leg up is again led by pre-revenue reactor names while contract prices are flat, our read is confirmed.
This exposure suits an investor who explicitly wants the whole nuclear fuel-cycle complex — miners plus enrichment plus reactor developers plus some non-uranium industrials — in one liquid line, and who treats single-name concentration in Cameco and drawdowns in excess of 50% as a normal feature rather than a failure. It does not match an objective centered on the uranium price itself (a physical trust tracks that more directly), a pure miners basket (URNM is tighter), nuclear generation with utility ballast and income (NLR is built for that), or a view specific to Cameco, which is individually liquid and available commission-free at most brokerages.
0.69% net expense ratio (issuer, Jul 2026) on $5.38B in net assets as of Jul 17, 2026. That is above VanEck's NLR (~0.5–0.6%, secondary-source) and below Range's NUKZ (~0.85%, secondary-source), roughly in line with Sprott's URNM (~0.75%, secondary-source). The fee is not egregious for a global multi-listing thematic index, but it is worth sizing against the portfolio: on a fund where ~23% is a single US-listed large-cap and another ~6.5% sits in BHP and Samsung C&T, a meaningful share of the fee is being paid on exposure that is directly accessible at no fund-level cost. The Sprott Physical Uranium Trust sleeve (5.41%) carries its own management fee beneath the 0.69%, so the all-in look-through cost is modestly higher than the headline.