Marcos Milla
Former Analyst at the Federal Reserve, ETF/Stock Newsletter Author
Roundhill's HALO ETF is a far safer play than a memory ETF because it holds companies AI is unlikely to displace. HALO stands for Heavy Assets, Low Obsolescence: businesses built on tangible, hard-to-replicate physical infrastructure and entrenched operating footprints, which make their moats resistant to AI breakthroughs. Top holdings include Cummins, AutoZone, TFI International, J.B. Hunt and Lennox International. x.comÂ
The Roundhill HALO ETF flips AI disruption risk by owning heavy-asset, low-obsolescence businesses. Its holdings rest on real assets, essential demand and cash flows that are here to stay. x.comÂ
Memory chip makers look promising as a long-term hold because memory supply is constrained until 2028 and most likely beyond, and demand will only become more constrained. The value still looks insane even after the run. x.com x.com x.comÂ
Amazon's logistics network protects its investment case against AI agents that could hurt its platform: agents cannot rebuild its delivery infrastructure, which has reportedly become America's largest parcel carrier. x.comÂ
SK hynix sits at the center of every major tech trend from robotics to AI, yet trades at a significant discount even on conservative estimates, offering very favorable risk/reward. Memory remains cyclical, but the floor has moved significantly higher. Even on the 15x terminal multiple @Kennyz8e prefers, fair value would be roughly $210, a 15% annualized return. x.com x.comÂ
Heavy-asset, low-obsolescence businesses such as ExxonMobil, Chevron, Waste Management, Safran and GE Aerospace are where the next decade's wealth will be built, as AI turns code into a commodity. Their moats are physics, regulation and scarcity: pipelines, landfill permits and jet engine certification take permits, capital and decades to build, and cannot be replicated by a founder with a laptop and an API key. Software's asset-light story is over. x.comÂ
If recursive AI makes intelligence cheaper, demand explodes and the bottleneck moves to the physical world, which favors suppliers of GPUs, HBM, interconnect and power such as Nvidia, Bloom Energy, Eaton, Vertiv and Wolfspeed. Recursive self-improvement is debated as a software question, but that framing is wrong. x.comÂ
Applied Digital is a safer AI infrastructure bet than CoreWeave, with fundamentals aligning for a technical breakout. Its weekly chart is breaking out of a multi-year accumulation base on strong volume, and it has outlived doubts about its crypto past to trade as a neocloud. Powered campuses give it a hard-asset profile, while CoreWeave leans more on aggressive utilization and valuation assumptions. A quoted post notes a 31% short float that fuels covering on new highs. x.com x.comÂ
SK hynix's buyback signals that management believes memory has become less cyclical, which the market doesn't yet price at 4x forward earnings. It announced a $29B buyback and cancellation and changed its policy to return at least 50% of free cash flow, from up to 50%. With long-term contracts, AI as a commodity with constant demand, and demand from AI devices and humanoids, memory should stay in high demand for years, and bottom-cycle multiples should be much higher than in the past. x.comÂ
Semiconductor stocks' easy-money phase is over as the sector shifts from hype to show-me-the-money. The SOX index is down over 15% from its June peak, chip funds saw $11B of outflows in one week, the largest this century, short interest is at a three-year high and earnings growth is projected to slow by 2027. Memory and logic chips remain commodity-like. This is cyclical cooling, not a death knell; the AI build-out continues, but expect more turbulence and favor companies with defensive moats. … Show more
