If you followed 2026 through headlines, you heard one story: cyclicals are back. Money rotated out of AI-heavy technology names and into the so-called real economy. That summary is not wrong. It is just incomplete in a way that cost anyone who traded the label instead of the industries inside it.
Look at the three groups usually lumped together under that label and you find three different years.
The industrial complex has been genuinely strong, gaining more than 16% year to date. The leadership is concrete: Caterpillar rose roughly 32%, riding capital spending tied to electricity capacity and AI data center construction. Sector research from Schwab points to the same driver, with spillover into materials. This is not a sentiment trade. It is spending showing up in order books.
Energy did even better, up over 22%, with gains concentrated in the majors. Chevron climbed nearly 22% on rising production plans and cost-cutting targets, while Exxon committed to higher spending aimed at earnings growth through 2030. Again: identifiable drivers, identifiable names.
Then there is defense.
Defense sits inside the same broad industrial category as Caterpillar in most classification systems. It did not share the ride. An interim peace agreement between the United States and Iran took near-term conflict risk off the table this year, and expectations for new weapons demand reset across the group. Lockheed Martin lost more than 17% over a 90-day stretch on that shift alone.
Even inside that struggling group, the picture splits again. Northrop Grumman reported a record backlog above $95 billion and reaffirmed strong full-year guidance while the industry narrative around it soured. Company fundamentals and industry headlines can point in opposite directions at the same time.
Here is what makes this worth studying rather than just noting. A trader thinking in sectors saw one blended number that averaged a strong industrial rally against a defense reversal. A trader watching industry-level relative strength saw defense leadership fade months before it dented any broad index. The divergence was visible. It was just not visible at the resolution most people watch.
That is the practical argument for working top-down: market first, then industry, then the stock. Sector labels compress information until the interesting part disappears. Industries are where the actual stories live, and in 2026, three industries wearing the same label told three different ones.
The takeaway is not a call on energy, industrials, or defense. It is a habit: before accepting any theme as a single trade, check whether the industries inside it agree. This year, they did not.
Full research and methodology at imgeld.com