$Apple(AAPL)$ The iPhone 17 Pro models are set to use 12GB of DRAM. Apple's estimated cost for that is around $39. For the base storage, like 256GB NAND, it's about $13. So the overall bill of materials for a base iPhone 17 Pro comes to roughly $582. This contributed to the modest price adjustments, like the $50 hikes on some Pro models where the base storage doubled from 128GB to 256GB. Looking ahead to the expected 2026 iPhone 18, projections show much higher memory costs due to global shortages and AI server demand. The DRAM for around 12GB could jump to about $145, nearly four times higher. The 256GB storage might rise to around $51. That puts the total BOM for a base iPhone 18 Pro potentially at $726, a 25% increase. Analysts have linked
$Microsoft(MSFT)$ At these prices, MSFT is definitely buying back shares. They've been doing it and will continue to, as they have a ton of cash set aside for that purpose.
Cash is becoming a weapon again. Balance sheet strength means optionality in this cycle. Here are the largest cash positions among the mega-cap leaders: $Berkshire Hathaway(BRK.B)$ - $397B $Amazon.com(AMZN)$ - $143B $Alphabet(GOOGL)$ - $127B $Meta Platforms, Inc.(META)$ - $81B $Microsoft(MSFT)$ - $78B In volatile regimes, liquidity isn't just defensive—it's offensive firepower for buybacks, AI capex, and M&A. That's why capital still rotates back to the strongest balance sheets first when stress hits the market.
The massive deal signed by $Microsoft(MSFT)$ and $Chevron(CVX)$ serves as a proof of concept for GEV. It validates that off-grid, “behind-the-meter” natural gas power plants can directly and reliably fuel massive AI data centers. The initial phase of this data center and power project (“Project Kilby”) is estimated to cost $7B. Under the 20-year Power Purchase Agreement, $Chevron(CVX)$ will develop an off-grid natural gas power plant in West Texas to supply dedicated electricity to a new $Microsoft(MSFT)$ AI data center campus. The specialized turbines required for Project Kilby are already factored
$Microsoft(MSFT)$ Microsoft is so oversold that even a broad market drop can't push it lower anymore. We've hit the floor, and the only way forward is up.
$Microsoft(MSFT)$ Owning Amazon, Meta, and Microsoft stock has been one of the simplest long-term ways to beat the market, and that basic setup still hasn't really changed. These names don't feel "cheap" in the traditional sense, but that's often the case when they're in the early stages of re-rating cycles. The structure looks similar to how Google behaved in 2025, where patience mattered more than perfect timing. It's not about overthinking the entry points. It's about staying aligned with companies that keep compounding while the market rotates around them. Simple doesn't mean easy, but it often works.