Two Random Thematic ETF Ideas I Would Like to See Come to Life

One thing I have learned from analyzing ETFs for both retail investors and advisors is that it is surprisingly difficult to predict what will actually become successful.

You can have a genuinely thoughtful strategy, a reasonable fee, strong branding, and even decent early performance, and the ETF still might fail to gather assets. Meanwhile, some completely absurd product unexpectedly catches fire and accumulates billions.

Part of that is because the ETF market has become unbelievably crowded. The U.S. alone now has more than 5,000 listed ETFs, and new launches are hitting the market constantly. At this point, almost every conceivable category feels saturated.

And because of that, issuers only really have a handful of ways left to differentiate themselves. Fees are obviously one avenue, especially because investors have become hyper-sensitive to costs over the last decade (thanks, Vanguard).

But beyond pricing, ETF firms are now competing on strategy construction, tax efficiency, liquidity, branding, accessibility, derivatives overlays, distribution partnerships, and increasingly just plain novelty.

Still though, one of the fun things about having your own website is the ability to exercise a bit of editorial freedom. So today, I figured I would throw together a list of ETF ideas I personally would not mind seeing come to life.

Some of these are legitimate concepts that I genuinely think could fill an actual gap in the market. Others are probably terrible ideas from a commercial perspective but interesting thought exercises nonetheless. And honestly, if you happen to be an ETF entrepreneur, portfolio strategist, or product development person at an issuer reading this and want to steal one of these ideas, go ahead.

Ideas are cheap. Execution is the hard part. Launching an ETF requires market makers, seed capital, legal work, distribution, operations, index licensing, exchange relationships, compliance approval, and enough assets under management to keep the product alive long enough to gain traction.

So, with that in mind, here are a handful of ETF concepts I think would either be genuinely useful, mildly entertaining, or at the very least interesting enough to justify existing. I’ll also throw in what I think would be a reasonable expense ratio for each because in today’s ETF market, pricing is half the battle.

The Long AI / Short India ETF

This next one is admittedly a little unhinged, but I genuinely think there is a coherent macro thesis behind it. As I am writing this, the NIFTY 50 is down 9.55% year to date. That stands out because broader emerging markets have generally done very well over the same period.

And that got me thinking about a thematic long/short ETF structure. The idea would be pretty simple. On one side, the ETF would short the Nifty 50 through swaps or futures. That gives you bearish exposure against 50 of India’s largest blue-chip companies.

On the other side, the ETF would go long the Nasdaq-100. Now, I am sure there are more targeted AI indexes you could use for the long leg, but the Nasdaq-100 is extremely liquid, easy to access, and there is no shortage of counterparties willing to facilitate that trade.

The thesis here is basically an AI disruption trade. You are betting that AI continues to displace white-collar labor at an accelerating pace.

And when you look at India’s economic structure, that becomes pretty interesting because the country has increasingly positioned itself as the back office for the global economy through business process outsourcing, customer support, coding, IT services, accounting support, and administrative labor rather than through the kind of manufacturing-heavy industrialization path followed by countries like China.

At the same time, India has a massive glut of university graduates competing for what may become a shrinking pool of white-collar employment opportunities. Western corporations have happily outsourced labor-intensive knowledge work to India for years because labor arbitrage made economic sense.

But I genuinely think agentic AI changes that equation over time. If an AI system can replace entry-level coding, customer support, translation, administrative processing, or document review work at lower cost and higher speed, then the outsourcing advantage weakens substantially.

So, the trade structure becomes pretty straightforward. If AI-related firms continue outperforming, your Nasdaq long leg works. If global equities broadly sell off because AI valuations compress or the economy tanks, well, emerging markets like India would probably get hit hard too, which means the short leg potentially cushions some of the downside.

And if the more aggressive thesis plays out where AI meaningfully disrupts outsourced white-collar labor markets, then theoretically both sides of the trade could work simultaneously. Your U.S. technology exposure benefits from AI adoption while the India short leg suffers from weakening growth expectations tied to labor displacement.

Now obviously this would fall squarely into the “alternative ETF” category, and you would probably need to structure it using total return swaps or futures rather than physical holdings. But honestly, derivatives access is not particularly exotic anymore.

I think something like this could probably be launched with an expense ratio around 0.75%. As for who would actually issue it, maybe Roundhill Investments or Defiance ETFs would take a swing at something this weird. Who knows?

High Asset, Low Obsolescence ETF

One of the more notable ETF losers year-to-date has been the iShares Expanded Tech-Software Sector ETF (IGV), largely on the back of fears surrounding AI disruption. At one point, the ETF was down 27.24% year to date as of April 10 before recovering somewhat. As of May 18, it is now down closer to 9.8%.

I think the opposite side of that trade is interesting. If software companies are vulnerable to AI commoditization and displacement, then perhaps there is a case for businesses with hard physical assets, entrenched infrastructure, and operations that are difficult to automate or digitally replicate. Basically, businesses with high asset intensity and low obsolescence risk.

Ironically, I am a little late to this idea because Roundhill Investments already launched something fairly similar in the form of the Roundhill HALO ETF (LOHA). Apparently, the ticker HALO itself was already taken by biotech Halozyme Therapeutics.

Roundhill’s approach is actually pretty thoughtful. The ETF tracks the Acros U.S. Heavy Asset Low Obsolescence Index and charges a fairly reasonable 0.35% expense ratio. The benchmark consists of roughly 99 companies believed to possess hard-to-replicate physical infrastructure, durable operating footprints, and entrenched competitive moats.

Personally though, I would have gone active with this concept. I think there is a lot of nuance in identifying businesses that are resistant to technological disruption, and a discretionary manager could probably build something more interesting than a rigid rules-based index. For me, the portfolio would probably break down into a few broad themes.

First would be waste management and disposal businesses. Companies like Waste Management, Republic Services, Waste Connections, and Clean Harbors. These rely on physical networks, permits, landfill ownership, specialized fleets, environmental regulation expertise, and local monopolistic dynamics. Nobody is disrupting garbage collection with a chatbot. In many regions, these firms effectively operate as tollbooths on waste infrastructure.

The second bucket would be what I would broadly call “dirty businesses.” Things like Cintas, Rollins, and Service Corporation International. These are businesses tied to physical labor, recurring service needs, regulatory requirements, or uncomfortable industries people generally do not want to compete in. Uniform cleaning, pest control, funeral services, sanitation logistics. Not glamorous, but durable.

Then I would probably include industrial and storage REITs. Names like Public Storage and Prologis. Again, these own hard assets in strategically valuable locations. Warehouses, logistics hubs, and storage facilities become more important as e-commerce, supply chain complexity, and inventory management continue expanding. You cannot digitally replicate a well-positioned industrial logistics network.

And finally, I would absolutely include railroads. Union Pacific, Canadian Pacific Kansas City, Canadian National Railway, Norfolk Southern, and CSX. The rights-of-way alone are almost impossible to recreate politically or economically today. They own irreplaceable infrastructure networks connecting ports, cities, industrial hubs, and supply chains across entire continents. AI may improve scheduling and logistics efficiency, but it is not replacing the physical rail network itself.

Now obviously, going active would cost more. That is the trade-off. Roundhill’s passive approach at 0.35% is honestly pretty affordable already. An actively managed version probably comes in closer to 0.5%. Still though, I think the benefit of an active structure is flexibility. You have more discretion regarding what kinds of businesses truly fit the “high asset, low obsolescence” idea, for better or worse.