Pinned Tweet
We have been running a series called 20 Unique Businesses on this page.
The filter is simple :-
→ High barriers to entry
→ Cannot be replicated just via blank cheque
→ Good return ratios
→ Limited competitors
→ Doing something cutting edge or structurally different
Not stocks to buy. Not targets to chase. Just businesses worth understanding, the kind where the moat is not a number on a screener but something you feel once you study the value chain.
If you missed any of them, this thread is your single window. Every business we have covered so far, in one place -
SOIC Research retweeted
A detailed 4 min that explains complete BESS industry in detail and how money flows in this value chain.
Could BESS be the next growth frontier?
Disclaimer - Not a recommendation to buy/sell
SOIC Research retweeted
So what is this chain, really?
It looks like an oil-price trade. It isn't. It's a capex pipeline with a policy valve in the middle. The price decides who earns this quarter; the capex decides who earns for the next five years.
Here is the complete flow of money that happens across the value chain
> Brent: $61 in January, $126 by the end of April
> HPCL: ₹11,526 cr standalone loss in a single quarter
> Reliance's O2C EBITDA: up 17%
> Petronet: terminals 58% utilised, and still a record profit
All this because Strait of Hormuz went shut
So we went through the latest (Q1 FY27) concalls and filings of all listed names in India's hydrocarbon value chain, from ONGC's wells to the cylinder in a CNG car, to answer three questions:
1. Where does the money actually flow?
2. Who carried the 2026 shock, and who got paid by it?
3. What are the growth triggers in each layer now?
And here are the growth triggers across the value chain
Source - @stockscansin
docs.google.com/spreadsheets…
What's common between an Austrian circuit-board maker and a 114-year-old Japanese company that started life as a hydroelectric power utility?
Neither of them makes a single chip.
They make the small square of layered plastic, glass and copper that every AI chip sits on. It's called an ABF substrate, and the market has suddenly realised that the world cannot make enough of them.
ABF substrates are AI's next bottleneck. Here is what is happening, how the value chain works, and where AT&S and Ibiden sit in it.
ABF substrate demand is growing much faster than the industry can add usable capacity, GPUs started it, agentic AI is now making server CPU demand explode, networking chips are piling on, and 2028 will be the worst year.
Glass-core substrates replace the organic core with glass, which stays flatter at large sizes. SKC's Absolics targets mass production around end-2026 (after slipping before), Samsung Electro-Mechanics' JV plant is slated for 2H 2027, DNP for FY2028, Intel around 2030 (now licensing rather than manufacturing), and TSMC after 2030.
Two points: glass cores still carry build-up layers on top, so ABF film does not disappear; and whoever masters glass at yield could reshuffle the substrate league table in the 2030s.




