Trade Ideas
Data centers will require another 200 gigawatts of installed capacity by 2030. Schneider confirmed 70% of its turnover is related to digital activity and is launching new R&D in India specifically for liquid cooling. The AI boom is physically constrained by power and heat. As compute density increases (Nvidia chips), traditional air cooling fails. Schneider is positioned as a "pick and shovel" play on the physical infrastructure required to run AI, regardless of which model wins. LONG SCHNEIDER as a primary beneficiary of the energy efficiency and cooling capex cycle. Supply chain bottlenecks for chips/components slowing down data center construction.
ServiceNow stock dropped ~20% in three weeks amid a broader SaaS sell-off. However, the CEO is personally buying $3M in stock, and the company is crossing $1B in AI-related revenue with free cash flow margins at record highs (Rule of 50+). The market is indiscriminately selling software stocks on fears that AI agents will replace SaaS seats ("SaaS-pocalypse"). ServiceNow's insider buying and tangible AI revenue growth contradict the "AI loser" narrative, offering a value entry point in a high-quality compounder. LONG NOW as a contrarian play against the software sentiment washout. Continued sector-wide multiple compression or slower enterprise AI adoption.
Microsoft is on pace to spend $50 billion by 2030 on infrastructure in the Global South alone. Copilot usage is "gaining ground" and getting better weekly. The sheer scale of capex creates a moat that smaller competitors cannot cross. Microsoft is diversifying its model dependencies (OpenAI, Mistral, internal models) while owning the distribution layer (Copilot/M365), insulating it from single-model risks. LONG MSFT as the dominant infrastructure and application layer player in enterprise AI. ROI on the massive $50B+ capex spend takes longer than expected to materialize.
Local media reports suggest Samsung could price its HBM4 chips at around $700 per unit, representing a 20-30% increase from the HBM3E. Bloomberg Intelligence analysis suggests this could drive operating margins to as much as 60%. The market previously penalized Samsung for lagging SK Hynix in high-bandwidth memory. A successful price hike and margin expansion indicate they are closing the gap and exercising pricing power in a supply-constrained environment. LONG Samsung as it catches up in the AI memory race with improved unit economics. Production yield issues or faster-than-expected commoditization of HBM chips.
The TOPIX is trading around 18x forward earnings, the highest valuation in 20 years. The Nikkei has rallied significantly following political consolidation. While the structural reform story (corporate governance, weak yen) remains, the valuation expansion has been rapid. The market is now priced for perfection, making it vulnerable to profit-taking or a reversal in the Yen. NEUTRAL/WATCH. The "easy money" in the Japan trade has been made; upside is now limited by stretched valuations. A sudden strengthening of the JPY would hurt exporter earnings and compress multiples.
Wesfarmers (parent of Bunnings/Kmart Australia) reported that "lower income households have less money to spend" and customers are becoming "more value conscious" due to persistent inflation and high interest rates. While the company is managing well, the macro commentary signals a tapped-out consumer in Australia. Retailers facing cost-of-living headwinds struggle to expand margins when volume growth depends on price cuts. AVOID WES (Wesfarmers) until inflation data stabilizes and consumer discretionary spending power returns. An unexpected rate cut by the RBA could reignite consumer spending.
This Bloomberg Markets video, published February 19, 2026,
features Olivier Blum, Amit Zavery, Brad Smith, Avril Hong, Alice French, Rob Scott
discussing SCHNEIDER, NOW, MSFT, SSNLF, EWJ, WES.
6 trade ideas extracted by AI with direction and confidence scoring.
Speakers:
Olivier Blum,
Amit Zavery,
Brad Smith,
Avril Hong,
Alice French,
Rob Scott
· Tickers:
SCHNEIDER,
NOW,
MSFT,
SSNLF,
EWJ,
WES