I spent twenty years inside businesses — auditing them, running their finances, raising their capital across three countries. Now I’m doing the thing I prepared for all along: investing from India in the world’s best businesses. I built a system to read companies the way an insider reads them — and I’m learning as I go. Everything I read, conclude, and get wrong is published here, dated and graded, so we can learn it together.
The training was an accountant’s — chartered accountancy in India, then years of signing the audit reports other people trade on, an executive-education stint at Harvard Business School along the way. The education that mattered more was the one no institution grants: watching, from the CFO’s chair, exactly how the numbers are made before they are printed.
That vantage is the whole premise of this site. Most market writing reads companies from the outside in. This desk reads them the way the inside reads them — where cash actually moves, which disclosures are confessions, what a footnote costs.
- 1 · CORPUSPrimary sources: filings, call transcripts, trade data.
- 2 · EVIDENCEEvery figure gets a source and a date. Missing data is declared.
- 3 · THE READA written argument — mechanism over headline, anti-thesis included.
- 4 · THE CALLWhen a read becomes a call, it is dated at that day’s price, with the condition that proves it wrong.
- 5 · THE GRADERight or wrong, in public, permanently. Corrections append; nothing is deleted.
Those are the honesty rules, and they are mechanical: a source and date on every figure, unknowns declared instead of filled, and a ledger that keeps the wrong calls exactly as prominently as the right ones.
This desk is AI-assisted and human-judged. Every day the system reads hundreds of filings, disclosures, and news items across the world’s major markets. What you read here is what survived my judgment: curated, checked, and signed. The machine gives me breadth; the conclusions are mine.
- The method in the open
- Checklists, scan rules, and frameworks published in full — not summaries of them.
- Sources you can check
- Every figure tagged and dated, so the work can be re-done by anyone who cares to.
- Misses graded in public
- The lessons are real because the wrong calls stay on the page with their post-mortems.
If that sounds like a sensible way to learn markets, read with me — the Daily takes two minutes.
A working investor’s journal with a system behind it: daily context, deep research, published methods, and a permanent public ledger of every call — dated and graded.
Not advice, not tips, not a chatroom. Nothing here tells you what to buy — it shows you how one investor decides, and holds him to it.
Until SEBI investment-adviser registration completes, this site is a journal in the strict sense: no recommendations, on anything, to anyone.
The registration is in process. The discipline is already here.
Corrections are especially welcome — they append to the record, with credit.