How this is made, and how it stays honest.
dhawal.org is one person’s markets desk. These are the rules it holds itself to: what a model writes versus what a person writes, how mistakes get fixed, where every number comes from, and what the author does and does not have a stake in.
Use of AI
Some text here is drafted by a language model; most is not. The rule is that a model never originates judgement or fact: it only re-states data the site already computed or fetched, and every figure it prints is checked against that underlying data before it ships.
- Model-drafted, data-grounded. The daily brief on /today and the earnings-call summaries are drafted by a language model from the data shown beside them. Each edition labels its provenance: the model that wrote it, or “composed deterministically” when no model ran. When the model is unavailable, a deterministic template builds the brief from the same data: no model, no invention.
- Human. The research, the futures-trading book, the indicator code, the methodology, and anything resembling an opinion or a call are written by a person.
- Never. No model writes trading calls, track-record entries, audit or P&L figures, testimonials, or credentials. Those are owner-authored or they do not appear.
The full formula registry and source list live on /methodology.
Corrections
Errors are fixed in place, with a dated note added here when the correction is material (a wrong number, a broken claim, a misattributed source). Cosmetic or typographical fixes are not logged. If you spot something wrong, email [email protected] and it will be looked at.
No corrections logged yet.
Sourcing
Every external number is traceable to its source, and every headline links to its original publisher. The site computes its own composites (the macro regime, the risk dial, breadth, relative strength, the VIX term structure) from public data; it does not republish anyone’s proprietary feed.
- Public-domain sources. CFTC (Commitments of Traders), SEC EDGAR (filings and XBRL fundamentals), the U.S. Treasury (yield curve), Treasury FiscalData (auctions), the EIA (energy curves), FRED (macro series), and GDELT (news coverage velocity).
- Delayed market data. Quotes, movers, and dividends use unofficial endpoints (Yahoo Finance, Finnhub, CBOE). These are clearly labelled and treated as delayed, indicative data, not a redistribution feed.
- News. Aggregated from public RSS feeds; the wire shows the first mover and links every outlet’s version. No article body is ever scraped or republished.
Source cadence and cost for each feed are listed on /methodology.
Positions & conflicts of interest
The author trades the instruments this site covers. Until a detailed position-disclosure policy is published in this section, assume he may hold a position (long or short) in any instrument shown here at any time, and that nothing on the site is timed to his own entries or exits. The site earns nothing from any broker, exchange, or data vendor named here.
Funding
The site carries no ads, no cookies, and no paywall. It is funded out of pocket by its author; a fuller funding statement will appear here. If that ever changes (a sponsorship, a paid tier) it will be disclosed in this section before it goes live.
The only measurement is Cloudflare Web Analytics: privacy-first, cookieless, aggregate page-view counts with no cross-site tracking and no personal data. That is why the site says “cookieless analytics only” rather than “no analytics”: it counts visits to know what is useful, but it cannot identify you.
Not investment advice
Everything here is research, data, and tooling, not investment advice, not a recommendation, and not a solicitation. Markets carry risk; you are responsible for your own decisions. Do your own work.