Power is the constraint
The AI trade is sliding downstream from chips into electricity. Solaris Energy (SOLR) moved higher after projecting $600M+ in pro forma earnings, explicitly tied to rising data-center power demand. This isn’t about a story you can’t model. Load growth you can see and contract is easier to underwrite than “AI monetization,” so capital keeps rotating toward the parts of the stack where demand is measurable and capacity is finite.
Two things in the framing did the work:
- Visibility is valuation. SOLR talked about data-center power as an earnings engine, not a concept. In this tape, clarity on how capex turns into EBITDA gets rewarded.
- Bottlenecks are a premium. Grid and delivery constraints aren’t theoretical. Scarce capacity is starting to show up in pricing—and in who gets funded.
Solar isn’t one trade
Solar didn’t move as a sector. It traded like a set of exposure charts.
Enphase (ENPH) was up after a Jefferies upgrade to Buy, citing stronger demand and better margin visibility. When incentives and channel data are noisy, “visibility” is basically a product: investors want someone who can draw a straight line from shipments to gross margin and defend it.
First Solar (FSLR) was down on weaker demand and persistent import tariff pressure, including on aluminum components. When demand softens but cost/policy headwinds stick, the equity story compresses quickly. Names with cross-border sourcing risk and trade-driven pricing uncertainty don’t get the benefit of the doubt.
The split was clean: the market isn’t paying for “clean energy” as a single duration bet. It’s sorting the group by (1) demand elasticity, (2) tariff/input exposure, and (3) margin resilience. ENPH got credit for perceived control. FSLR got hit for getting squeezed from both sides.
Tariffs create a wedge
A useful non-ticker datapoint: Emirates Global Aluminium (EGA)ramped U.S. aluminum sales amid domestic shortages and tariff-driven price increases, capturing the tariff premium. The headline matters less than the mechanism, because it’s the same machinery that shows up in downstream manufacturers.
- Tariffs don’t just reduce flows; they create a pricing wedge. If U.S. supply is tight enough, imports still clear because end prices rise enough to cover the tariff.
- The pain usually lands downstream. Component-heavy manufacturers eat higher inputs and more headline risk, while producers/exporters monetize the spread.
As long as tariffs stay in place alongside constrained supply, expect the same pattern: occasional margin lift for sellers into tight U.S. pricing, and ongoing cost pressure for anyone building physical products with aluminum in the bill of materials.
The rest of the tape
Cava (CAVA) jumped on an optimistic full-year sales outlook backed by better traffic. That’s a stock-level demand signal the market still pays for. In restaurants, traffic plus a guide-up reads as higher quality than a vague “consumer is fine” narrative, because the operating leverage cuts both ways.
Elsewhere was mostly maintenance. New Mountain BDC / New Mountain Finance (NEWT) was flat despite Coller Capital agreeing to buy $477M of assets to improve flexibility; investors seem to want clarity on reinvestment yields and distribution impact before re-rating it. Ionis Pharmaceuticals reiterated $2B+ peak sales potential for olezarsen tied to 2026 launches—a useful timeline marker, not a new catalyst today. In Washington, shutdown noise persisted with DHS funding unresolved and TSA pay disruption still in the background.
- SOLR: data-center power demand is becoming an earnings bridge investors can fund
- ENPH vs FSLR: solar traded on margin visibility and tariff/input exposure
- EGA: tight supply + tariffs = premium pricing that flows downstream
- CAVA: traffic and a higher guide still cut through macro haze
The common thread: the market is paying for control—of capacity, of inputs, and of the path from demand to margins.