According to @DowdEdward, companies leasing assets on a 3-year depreciation schedule are aiming to keep revenue momentum, which he says indicates the AI capex cycle is in its late innings, making revenue quality and financing structures key trading variables. Source: https://twitter.com/DowdEdward/status/1971983286554841314 Under US GAAP ASC 842, shifting from upfront equipment sales to 3-year leases recognizes income over the lease term and can smooth reported revenue while changing cash flow timing, a pattern traders monitor in late-cycle hardware markets. Source: FASB ASC 842 Major hyperscalers have recently increased server useful lives to 5–6 years, so a 3-year depreciation horizon is comparatively short and would front-load expense recognition versus peers, a useful benchmark for equity screening. Source: Amazon.com, Inc. 2023 Form 10-K; Microsoft Corporation 2023 Form 10-K; Alphabet Inc. 2023 Form 10-K For crypto markets, the state of AI capex can influence power and data center availability for Bitcoin miners, as US miners have signed AI/HPC hosting deals that tie mining infrastructure to AI compute demand. Source: Core Scientific company announcements 2024; Hut 8 Corp. investor materials 2024 Traders should track disclosures on lease mix versus capex in AI hardware vendors and data center operators, and watch any changes in hosting rates and capacity allocations that could affect BTC mining economics. Source: Edward Dowd on X; FASB ASC 842; public company filings cited above

According to @stocktalkweekly, Jefferies initiated Entergy (NYSE: ETR) at Buy with a $109 price target, calling it the clearest regulated beneficiary of hyperscaler demand and the best-positioned AI load play in U.S. utilities (source: Jefferies initiation note as relayed by @stocktalkweekly on X, Sep 4, 2025). Jefferies projects 11–15% EPS CAGR from 2024 to 2030 and estimates ETR’s earnings will run 7–11% above Street in 2028 and beyond, outpacing every utility peer on growth, which forms the core of the upside case for the stock (source: Jefferies initiation note as relayed by @stocktalkweekly on X, Sep 4, 2025). For traders, the thesis hinges on sustained AI and data center load growth driving regulated rate base expansion and EPS outperformance versus the utility sector, with ETR identified as the most levered regulated name to that trend by Jefferies (source: Jefferies initiation note as relayed by @stocktalkweekly on X, Sep 4, 2025). The AI-driven power demand backdrop is corroborated by U.S. EIA analysis showing rising data center electricity load in the South, a region overlapping ETR’s footprint, reinforcing the structural demand case for utilities exposed to hyperscalers (source: U.S. Energy Information Administration, “Some U.S. states see growing data center electricity demand,” May 2024). For crypto markets, ERCOT has documented accelerating large flexible loads from data centers and crypto mining that affect system planning and pricing dynamics, implying potential pressure on power availability and costs for Bitcoin (BTC) miners as AI data center load scales, which can influence mining margins and hash rate economics (source: ERCOT Long-Term System Assessment 2024 and related planning materials).