$PPIH, $ASYS, $SILC - 3 AI names I like. Don’t have much debt, strong balance sheets, and should see strong earnings growth the next 12-18 months. I’d be super cautious buying companies with a lot of debt (think neoclouds), especially when interest rates will surely go up the next few quarters, and many of them need to raise a lot more capital, which will be expensive.

I started a position in $ASYS. I really like the setup here. The business is increasingly shifting toward AI-driven semiconductor advanced packaging, while the weaker SiC-related business continues to shrink. AI-related revenue was up ~120% YoY last quarter, TPS backlog is up 69%, and I think 2027 could be an inflection year as this demand really starts hitting the P&L. I’m also a big fan of new CEO Guy Shechter. He has deep semiconductor experience, strong industry relationships, and has been very visible meeting customers, partners and investors based on LinkedIn. Add in $83M+ of cash, and Amtech has plenty of flexibility to invest behind its strongest businesses or pursue an accretive acquisition. There’s also some longer-term optionality around thermal/cooling applications that I’m watching. Only one initial deal has been done here. Too early to tell though, so not a main part of the thesis. The stock isn’t optically cheap on current earnings, but 2027 estimates paint a much different picture. Around $15/share, I think the risk/reward is compelling enough to own an initial position. Looking forward to digging deeper into this one.

$VST remains one of my favorite ways to play rising U.S. power demand. Vistra is already benefiting from stronger pricing. Q2 adjusted EBITDA increased more than 30% YoY to $1.77B, driven primarily by higher realized energy and capacity prices along with contributions from acquired assets. Management continues to expect major load growth through 2030, which is roughly 4-6% annually in ERCOT and 2-3% in PJM driven by data centers, reshoring, electrification and population growth. The nuclear story is even more compelling. Vistra signed 20-year agreements with Meta covering 2,609 MW across its PJM nuclear fleet, including 433 MW of new uprates. It also has a 20-year agreement with AWS for up to 1,200 MW from Comanche Peak. The DOE has announced a conditional loan commitment of up to $4.2B to help finance nuclear uprates and modernization across Vistra's Pennsylvania and Ohio fleet. If finalized, that financing would help preserve nearly 4 GW of existing nuclear generation while adding 433 MW of new capacity. And the financial setup remains strong. Vistra's 2027 adjusted EBITDA midpoint opportunity is $7.4B-$7.8B and importantly, that figure excludes potential contributions from the pending Cogentrix acquisition AND the Meta PPAs. Vistra has also repurchased roughly $6.5B of stock since 2021, reducing its share count by ~30%. Management summed up the opportunity well on the latest earnings call: “the long-term expected improvement in power market fundamentals is underway.” TLDR; Higher realized power prices. Rapid electricity demand growth. 20-year nuclear contracts. Potentially billions in federal financing. Aggressive buybacks. I really like the long-term setup for $VST.