MAXIMIZATIONS

NVIDIA is increasingly becoming the foundational compute infrastructure layer powering the global AI supercycle. Beyond GPUs, its true moat lies in the deeply embedded CUDA ecosystem, which creates high switching costs and multi-year dependency across hyperscalers and enterprises. As AI factories scale globally, NVIDIA remains one of the key bottleneck owners benefiting from the accelerating demand for AI compute.

ResMed operates in a structurally underdiagnosed market where demand is driven by medical necessity rather than economic cycles. Its hardware-led model evolves into a recurring revenue ecosystem supported by patient adherence and provider integration. While GLP-1 concerns may moderate growth at the margin, the long-term demand outlook remains intact. At ~22× P/E, returns are likely to be driven by steady earnings compounding rather than multiple expansion.

Earnings growth is a key driver of long-term stock returns, and cost optimization is often an underappreciated lever. By analyzing earnings transcripts and filings, investors can identify whether management has concrete plans to improve profitability. Going further, mapping these initiatives to COGS, SG&A, or R&D provides a clearer view of margin expansion and earnings quality — turning qualitative insights into a real modelling edge.