Wedbush is betting that there is still an investment-friendly economy beyond artificial intelligence.

The firm recently filed for the Wedbush Analog Economy ETF, which would track an index of roughly 50 U.S.-listed companies with businesses that depend on physical assets, human labor and tangible output. The proposed portfolio would span building products, construction, materials, packaging, machinery, vehicle and equipment manufacturing, industrial distribution and environmental and commercial services.

But the more interesting part of the filing may be what doesn't make the cut.

No AI, No Data Centers, No Semiconductors

The Solactive index behind the ETF is designed to screen out companies with significant exposure to AI, semiconductors, data center infrastructure and power generation or utility infrastructure.

It goes further than a simple sector exclusion. Companies must meet an asset-intensity test, with net property, plant and equipment accounting for at least 7.5% of total assets on average over three years. The methodology can also apply a labor-intensity screen. Companies are then subjected to an AI exposure test that uses Solactive's proprietary ARTIS natural-language-processing system to assess AI-related activity in corporate disclosures.

That creates an unusual investment proposition: own companies that are physically intensive, but deliberately avoid the parts of the physical economy most directly tied to the AI boom.

And that's where things get a little complex.

AI Is Moving Into the Real Economy

Construction companies, industrial distributors, machinery makers and materials producers may look decidedly "analog." But many of these businesses are also benefiting from the enormous physical buildout associated with AI.

Data centers need construction materials. They need cooling equipment, electrical infrastructure and machinery. They consume enormous amounts of power. Industrial companies increasingly use automation and AI themselves.

In other words, drawing a clean line between the AI economy and the physical economy is becoming harder.

The prospectus itself acknowledges the problem. Wedbush says the index's AI screen could fail to identify exposure accurately because the underlying data and AI-relevance models may be incomplete or inaccurate.

Wedbush Isn't Alone

The filing also arrives as ETF issuers are turning the physical economy into a distinct investment theme.

The Roundhill HALO ETF (BATS:LOHA), launched in May, tracks an index of 100 companies selected for heavy physical assets and lower exposure to AI displacement. Its methodology specifically looks at physical-asset intensity and AI-displacement immunity.

The Tuttle Capital Heavy Asset Low Obsolescence ETF (BATS:HALX) takes a similar route, targeting companies with tangible assets and asset-backed cash flows. Its index generally holds 30 to 50 stocks.

Investors can also get more conventional exposure through sector funds such as the State Street Materials Select Sector SPDR ETF (NYSE:XLB) and the Industrial Select Sector SPDR ETF (NYSE:XLI). XLB, for example, holds companies across chemicals, metals and mining, packaging and construction materials.

That leaves Wedbush with a slightly different challenge.

Can "not AI" become a sufficiently differentiated ETF strategy when so much of the real economy is already being reshaped by AI?

The answer may depend less on the novelty of the "analog" label and more on whether its screening rules produce a portfolio meaningfully different from the industrial, materials and HALO (heavy assets, low obsolescence) ETFs already available to investors.