Vivek

#OlaElectric perennially loss making mints Rs 367 Crores from Taxpayers for shoddy products PLI Scheme is the worst idea of this government. Any Tax Break should be freely available to all, not allotted to specific companies for sloppy low grade or low value addition products economictimes.indiatimes.com/markets/stocks…

Market hasn't reacted negatively to Jupiter Wagons, likely due to a new order received from Indian Railways. I am still cautious but holding. #jupiterwagons #jwl

The next five years are going to be very interesting for natural gas… Revolution #1: Massive increases to U.S. compute infrastructure driving huge increases to natural gas sourced grid power. Large data center completions were doubling every 4 years, but AI has compressed this to every 2 years. Estimated US hyperscale completions since 2020: 300. Leading operators: Amazon (150+ US sites total), Microsoft (60+), Google (30+ new since 2020). By 2030, expect 500+ more large US facilities. These centers are primarily in: Northern Virginia (largest market, +83% capacity since 2020), Texas/Arizona (power availability), and Atlanta/Phoenix. Energy demands: 20-25 GW from new large data centers completed since 2020. That’s the equivalent to the output of ~15-20 new natural gas plants / ~16-20 million U.S. homes annually. If trends hold, new builds 2026-2030 add another 50-60 GW (more than double existing demand). Energy sourcing for new builds reveals a dominant reliance on natural gas. This is driven by the need for reliable, scalable baseload power to meet 24/7 demands. Bottomline: Today, if all large/hyperscale data centers were run at 100% capacity and were powered only by U.S. natural gas, they would consume roughly 6% of total existing natural gas production. By 2035, that figure will be over 20%. Revolution #2: Natural gas export demand is about to almost triple. In 2024, natural gas exports (11.9 Bcf/d / 4.35 Tcf annually) consumed 11.5% of total U.S. production. There are currently seven major LNG construction projects near-completion, underway, or planned and approved: ExxonQatar’s Golden Pass (2.4 Bcf/d); VG’s Plaquemines (1.6 Bcf/d); Cheniere’s Corpus Christi Stage 3 (1.7 Bcf/d); NextDecade’s Rio Grande (3.6 Bcf/d); Sempra’s Port Arthur (1.8 Bcf/d); VG’s CP2 (3.3 Bcf/d); and EnergyTransfer’s Lake Charles (1.6 Bcf/d). When completed (2029) these facilities will add another 18 Bcf/d of export capacity, taking total export capacity to 30 Bcf/d (11 Tcf annually) – roughly 3x current export capacity. This represents roughly 30% of U.S. production. That’s more than the entire U.S. electrical generation sector uses today. By 2030, LNG exports + AI compute energy demand will consume 50+Bcf/d of natural gas – or roughly half of total production. Supply is expected to grow, at most, by 10%. The coming natural gas price spike will be extreme at times (like in the winter) when demand surges. Pipeline capacity is very tight. Permian takeaway is already at 95% of capacity. And new pipelines take forever to permit and built. The other major issue is, unlike LNG, data centers never turn off. There’s zero seasonal flexibility. We’ll see consistent $10+ natural gas prices by 2030. And, when there's a significant demand spike in the winter, we'll see $50 gas.