Thematic ETFs are appealing: a strong narrative and simplified exposure to major contemporary issues. But in reality they are fairly complex financial products, and it is better to read the documentation before taking a position. In particular, the document detailing the methodology for selecting and rebalancing the tracked index. The production of data about these ETFs could also be greatly improved. ETFs are rated on their fees or their tracking error, never on the financial statements of the companies they hold.

As a result, these ETFs that are supposed to provide exposure to a long-term trend become vessels for euphoria, fueling valuations disconnected from any economic reality.

This is particularly striking in the case of the uranium theme, one of the commodities at the center of attention in 2025.

One thematic ETF dominates the European market: the VanEck Uranium & Nuclear Technologies. And for the purposes of this analysis, we will also look at the two others that follow it in terms of assets under management: Global X Uranium UCITS ETF and HANetf Sprott Uranium Miners ETF Acc.

All aim to capture the same dynamic: nuclear power's return to prominence in energy debates, notably in anticipation of a surge in demand following massive investments by the GAFAM in data-center development.

But the index construction and monitoring mechanisms tracked by these ETFs differ greatly from what you are used to seeing... and explain why a company that has never generated a single dollar of revenue, Oklo Inc, can represent as much as 14% of some of them.

What the numbers say

Since hardly anyone seems interested in doing it, I took the time to rebuild a few key aggregated economic indicators to better understand what we are investing in when we buy units of these continuously traded funds. I relied only on the top 10 holdings, which each time represent nearly 70% of the total weight of the tracked index.

Indeed, the first takeaway is that these ETFs are very concentrated: 70% in 10 positions is a lot, but it is not surprising given the theme and the number of listed companies whose activities are correlated with the uranium price.

Global X Uranium

To be completely honest, I was expecting much worse in terms of aggregated profitability. In any case, on valuation, all the aggregated P/Es flirt with or far exceed 50x.

All right, let's now try to explain how a company like Mitsubishi Heavy Industries, worth $90 billion, is assigned a weight of 4.92% in the VanEck ETF, while small ones like Cameco, or even Oklo Inc, often exceed 10%.

Understand that the nuclear value chain is highly stratified, from exploration and extraction through to plant operation and decommissioning, including enrichment, turbine manufacturing and the development of nuclear technologies. Each layer involves very different players, and if you want exposure to the uranium theme-and only the uranium theme-you will naturally focus on the earliest stages, because that is where you can find pure-plays, i.e., companies whose vast majority of cash flows are generated by activities highly dependent on the uranium price.

And in terms of choice, you cannot say we are spoiled, since the number of listed pure-plays can be counted on two hands.

Note, moreover, that the first two ETFs presented mix miners, infrastructure and nuclear technologies, while the last focuses on producers and physical uranium holders. It is in the first two that Oklo Inc, a pre-revenue company developing small modular reactors (SMRs), reaches outsized weights.

How can these outsized weights be explained?

Put simply, the big problem with most of the indices tracked by these ETFs is that their providers chose to focus on pure-plays (perfectly normal-that is what investors are looking for), but for lack of enough constituents, they had to add some fairly outlandish rules to allow the inclusion of certain companies, so as not to end up with five holdings. For example, in the MarketVector index, tracked by VanEck's ETF, different rules apply to current constituents (and therefore, I imagine, the initial list) versus candidates. Companies already in the index keep their status even if they no longer meet the selection rules. Quite strange.

On top of that, a different weighting method is applied depending on whether the company is a pure-play (more than 50% of your revenues directly exposed to uranium activities, in this case) or a more diversified player. The relative weights of companies such as Cameco, Kazatomprom or Oklo depend on their free-float market cap, while conglomerates like Mitsubishi Heavy Industries are capped at 5%!

It is the combination of the following factors:

  • there are not many uranium pure-plays in the market,
  • they are small or mid-sized,
  • they set their own pace when it comes to weighting (since the biggest players are capped),
  • and those weights depend on their free-float market capitalization.

That makes it possible to understand how a small pure-play like Oklo Inc ends up accounting for 15% of the index. In the case of illiquid companies, these thematic ETFs also act as genuine bullish catalysts. 

This self-reinforcing effect is nonetheless generally limited by thresholds that no security can exceed. For MarketVector, the top holding cannot represent more than 15% of AUM, 10% for the second holding and 8% for the others. But that cap is much higher, at 22.5%, for the Solactive index (tracked by the Global X ETF)!

So you are going to tell me, yes but these ETFs are great, performance was there, you are talking nonsense, you absolutely need them in the portfolio (yes, yes, you will indeed find such comments under the YouTube video associated with the article-no surprise that in Europe the Momentum premium is the most attractive). Remember that anything that rises fast can fall just as fast. That is the problem. In fact it makes you really want to go long one of these ETFs and short Oklo Inc.

For those who are interested, here is a brief summary of the main selection and weighting rules for the indices tracked by these three ETFs (these days you can have AI summarise these methodology documents-you really have no excuses anymore):

MarketVector Uranium & Nuclear Infrastructure Index (tracked by the VanEck ETF)

  • Each security is weighted by its free-float market capitalization, with caps: 15% (top holding), 10% (second), 8% (others).
  • If ≥ 50% of revenues come from uranium or nuclear → pure-play companies.
  • If < 50% → weight capped at 5%.
  • "Grandfather” exception: companies already in the index keep their status even if they no longer meet the thresholds

Consequence: two worlds coexist:

  • pure-plays (Cameco, Kazatomprom, Oklo) weighted by their market cap and therefore by their share-price performance,
  • the others (Mitsubishi Heavy, Jacobs Solutions…) capped at 5%.

Solactive Global Uranium & Nuclear Components Index (tracked by the Global X ETF)

  • Selection: companies with, or expected to have, activities linked to uranium, exploration, or nuclear technologies.
  • Weighting: minimum of Free-float Market Cap and ADV × 2000) → depends on both free-float market cap and average liquidity.
    • Uranium pure-play → cap at 22.5%;
    • Nuclear Component Producers → cap at 2%

Sprott North Shore Uranium Miners Index (tracked by the HANetf ETF)

  • Selection: only companies for which ≥ 50% of assets relate to uranium exploration, development or production.
  • Each security is weighted by its free-float market capitalization, capped at 20%
  • New thresholds (2025): market cap ≥ $125 million, liquidity ≥ $100,000/day.

Here, no nuclear tech, no SMRs, so no Oklo. The least convoluted index in my view.