Myuex

Long term energy storage solutions

The thesis that AI-driven electricity demand will boost long-duration energy storage stocks is structurally sound and supported by a cascade of confirming news: the IEA projects data center power demand will double by 2030, BloombergNEF doubled its US forecast to 194 GW by 2035, and grid connection bottlenecks are already delaying projects. These forces create a genuine need for storage to manage intermittency and grid congestion. However, the thesis faces two powerful headwinds. First, the regime is hostile: inflation is reaccelerating at 3.7% CPI, the Fed hiked in September and signals further tightening, and the 10-year yield is above 5%, which raises capital costs for capital-intensive storage projects and compresses valuations for pre-profit companies. Second, the basket's composition is problematic: BEP (20% weight) is primarily a renewable power generator with only 0.30 exposure to storage, diluting the pure-play thesis, while GWH, EOSE, NRGV, FLNC, and IESVF are all pre-profit, cash-burning companies with high short floats (EOSE 32%, FLNC 29%, NRGV 21%) that are acutely vulnerable to rising rates and the hostile risk-off macro environment. The invalidated 'AI Expansion Boosts Energy Use' pillar—triggered by the October stock crash and rising Fed hike odds—removes a key near-term demand catalyst. The upside case rests on the multi-year structural demand story, but the near-term macro and financing headwinds are severe enough to cap conviction. The strongest upside path runs through FLNC and NRGV, which have the most credible passthrough from grid-scale storage procurement, but even they trade at high multiples (FLNC P/E 56) that already price in some optimism. The downside is concentrated in the basket's vulnerability to rising rates and a recession that would delay data center buildout. The score reflects a thesis that is directionally correct over a multi-year horizon but faces a hostile near-term regime that limits conviction.

GaN is the future of semiconductors

The thesis rests on GaN's physical efficiency advantage driving adoption in AI data centers, with a non-Chinese gallium supply chain as the bottleneck bypass. The regime is hostile: reaccelerating inflation and a hawkish Fed keep capital costs high, which slows enterprise procurement cycles and data center buildout timelines, directly undercutting the adoption speed the thesis needs. The invalidated 'Limited Supply of High Purity Gallium' pillar is actually a net positive for the thesis—China's dominance and export controls are confirmed, not disproven, so the supply-risk argument for Western alternatives like AA and FPLSF is reinforced, not broken. The strongest supporting evidence is the string of confirmed events: Navitas' US GaN shipments, Alcoa's gallium project FID, and 5N Plus' defense role all validate the non-Chinese pipeline. However, the basket's realized performance (-36.9% vs S&P +0.4%) and NVTS's extreme valuation (P/S 103x, negative earnings) show the market has already priced in a lot of optimism that the thesis has not yet delivered. The key risk is timing: enterprise procurement cycles are slow, and the Fed's tightening bias delays the data center spending that would trigger GaN adoption at scale. The upside case is real but distant, while the downside from execution delays and macro headwinds is near-term and tangible. Conviction is moderate: the structural story is sound, but the horizon is long and the path is obstructed by the current regime.