Ordinary people buy stocks by looking at prices, financial reports, profits, sector popularity, and whether institutions are buying.

Serenity's investment approach involves breaking down the upstream and downstream of the industry chain. First, decompose a major trend, then find the most difficult-to-replace, easiest-to-block others, and the key nodes most dependent on giants.
After Nvidia's rise, truly skilled people will continue to break it down: why AI needs GPUs, why GPUs need HBM, why data centers need optical modules, why larger models require more storage, networks, energy, packaging, and cooling.
In early March, I also noticed a very interesting phenomenon: many industry leaders are starting to invest in nuclear power plants. The reason is simple: AI lacks energy.
This is a typical downstream logic.
If you are optimistic about AI, you can't just look at ChatGPT, Nvidia, Microsoft, and these names. You need to continue breaking down the AI industry chain:
Power supply, equipment manufacturing chips, chips generate computing power, storage and networks release computing power, cloud platforms distribute computing power, model companies train intelligence, transfer stations aggregate models, application layers reach users, agents ultimately complete tasks.
Serenity's investment logic is roughly like this. She looks down the industry chain layer by layer, identifying bottlenecks, choke points, overlooked areas by the market, where the next round of capital rotation might enter.
She looks at AXTI, focusing on InP substrates and other optoelectronic base materials. She looks at SIVE, focusing on next-generation CPOs and external light sources. She looks at AAOI, LITE, COHR, MRVL, focusing on optical modules, silicon photonics, networks, and the transmission between AI data centers.
The most valuable part of this approach is being ahead of where capital will flow in the future.
The market's first phase was speculating on Nvidia; the second phase is speculating on HBM and storage; the third phase is speculating on optical modules and CPOs. Going further, it may sink into materials, equipment, packaging, energy, and more niche supply chain nodes.
The explosion of AI models is the first order. The rising demand for GPUs is the second order. The shortage of HBM, optical modules, storage, networks, and energy is the third order. Going further down, certain materials, equipment, substrates, packaging, and light source suppliers are re-priced—that's the fourth order.
This is her logic: key nodes where the market has not yet fully priced in as the story continues to unfold.
Serenity's most valuable skill is this industry chain mapping ability—tracing from end-user demand back to infrastructure, from visible leaders back to hidden bottlenecks, from market consensus back to where capital will go next. This is her source of alpha. I don't know if I'm right—@aleabitoreddit, please advise!