Thematic Investing Examples: Two Themes With Real Numbers

What actually counts as a theme

A theme is a claim about how the world is changing, specific enough to check. "Technology is important" is not a theme. "Companies are rebuilding their computing around AI chips, and a handful of suppliers collect most of that spending" is a theme. So is "retail keeps moving from stores to the internet."

Sector investing says own technology. Thematic investing says own the specific change, wherever it sits. The claim comes first and the tickers follow from it.

That ordering matters because it tells you what evidence to watch. If the claim is about data-center spending, you check data-center revenue in filings. If the claim is about shopping habits, you check the retail sales data the government publishes every quarter. The stock chart is the last thing you look at, not the first.

Example 1: AI infrastructure, the loud one

The cleanest recent example of a theme showing up in actual filings is the AI buildout, and NVIDIA (NVDA) is where it shows up first.

All three figures from NVIDIA's Q4 FY2026 results, filed on SEC EDGAR.

The company's fourth-quarter press release, filed on SEC EDGAR, reports record full-year revenue of $215.9 billion, up 65%. Data-center revenue alone was $193.7 billion for the year, up 68%. The fourth quarter's data-center figure was $62.3 billion, up 22% from Q3 and 75% from a year earlier.

Notice what those numbers are not. They are not a sentiment reading, a chart pattern, or an analyst's price target. They are revenue, reported in a securities filing. When people say a theme is "real," this is what real is supposed to mean: the money shows up in the income statements of exactly the companies the theme says it should.

Now the uncomfortable part. Most of the public proof for the AI theme lives in a small number of income statements, and a lot of it lives in this one. When that happens, a theme bet and a single-stock bet start to look like the same trade. Buying "the AI theme" through one supplier is concentration risk wearing a theme costume.

Whether the buildout keeps growing at anything near that pace is the open question, and the filing does not answer it. Nobody's filing does. If you want to watch the argument play out, the NVDA symbol page on MarketPlays collects the news, sentiment, and community research in one place.

Example 2: e-commerce, the slow one

In the fourth quarter of 1999, online sales were 0.6% of US retail. As of the first quarter of 2026 they are 16.9%, per the FRED series ECOMPCTSA.

That is the whole theme in two numbers. Retail moved online, slowly, for a little over twenty-six years, through two crashes, a pandemic, and every flavor of market mood. Anyone who claimed in 1999 that shopping was headed to the internet turned out to be right, and kept being right for a quarter century.

And yet plenty of the people who bet on that exact idea in 1999 lost most of their money.

Here is an illustrative example, not a recommendation or a real portfolio. Picture an investor in early 2000 who was completely right about online retail and expressed the view through the flashiest dot-com storefronts of that era. The claim kept getting confirmed, quarter after quarter, in the same government data series. Many of the tickers did not live to see it, and their shareholders got nothing when the growth they had correctly predicted finally arrived.

The slow theme is more forgiving in one way: it hands you decades of chances to get on board. It is less forgiving in another: it demands you sit through long stretches where nothing seems to happen. Most people can't. That, more than idea generation, is where investing in themes actually gets hard.

A good theme is not the same as a good trade

Set the two examples side by side and the differences do most of the teaching.

Before putting money behind any theme, three questions are worth answering in writing:

• What specific, checkable number proves this theme is happening, and where is it published?
• If the theme is right, which companies actually collect the money, and how many of them exist?
• What would make me admit the theme broke, other than the stock going down?

If the answers are fuzzy, the idea is not ready to be a position yet. As for sizing, there is no cited standard and this article does not invent one; the honest framing is that a theme is usually a small overlay on a diversified base, held at a size you could watch fall out of favor without flinching.

Browse the AI tag on MarketPlays to see which symbols other investors group together. Checking someone else's grouping against your own list is a fast way to find the names your version of the theme forgot.

Key takeaways

• A theme is a checkable claim about change. If you can't name the number that would prove or break it, it's a mood.
• The AI theme is visible in filings: NVIDIA reported $215.9 billion in fiscal 2026 revenue, $193.7 billion of it from data centers (SEC EDGAR, February 2026).
• E-commerce went from 0.6% of US retail in late 1999 to 16.9% in early 2026 (FRED series ECOMPCTSA). Slow themes offer more entry points and demand more patience.
• Being right about a theme and getting paid for it are separate problems. Survivorship and concentration sit between them.

Open a MarketPlays account and set up your own hub portfolio in under two minutes. If you'd rather look around first, the explore page shows trending stocks, big movers, and the full tag index.

Whichever theme you pick, expect stretches where it looks dead. E-commerce spent years of its twenty-six year climb out of favor, so size the position at a level you can hold through those years.

FAQ