Data Sources

BazaarBaazi runs a two-layer data architecture: a public-data display layer that any reader can verify, and a proprietary insight layer derived from licensed market-data feeds. The insight layer informs the editorial verdict; the display layer carries the numbers a reader can check. We do not redistribute licensed feed data.

Display layer (public sources)

Numbers shown to the reader come from these primary public sources:

• NSE bhavcopy and the NSE archives. End-of-day price, volume, OI, and FII/DII activity. The exchange URL is cited where the number is unusual.
• BSE filings and corporate announcements. Company disclosures, board meeting outcomes, regulatory disclosures, anchor allocations, postal ballots, related-party-transaction reports.
• SEBI filings. DRHPs, RHPs, prospectuses, takeover disclosures, insider-trading disclosures, regulatory enforcement orders.
• RBI publications and notifications. Monetary policy statements, financial stability reports, foreign-exchange advisory frameworks.
• Government statistical sources. CMIE-equivalent open data, MOSPI releases, data.gov.in datasets, CSO publications, where applicable.
• FRED (Federal Reserve Bank of St. Louis). Global macro context where India-impacting (US 10y, DXY, oil, gold).
• Marketaux, Tavily, and primary newswires. News aggregation with primary-source verification before any quantitative claim is published.
• CoinGecko and exchange APIs. Cryptocurrency display layer where a story touches the asset class.
• Wikimedia, Wikipedia, public-domain photo pools. Editorial images sourced primarily from Creative Commons or public-domain archives, embedded via the original publisher's URL where the publisher allows.

Insight layer (licensed feeds, not redistributed)

BazaarBaazi's editorial team uses licensed real-time market-data feeds to inform the editorial verdict. The published story does not reproduce or redistribute licensed data; it cites equivalent end-of-day public sources where the reader needs to verify a number.

• Zerodha Kite Connect. Real-time tick data, market depth, OI changes, and trade-by-trade information. License terms prohibit public redistribution of tick-level data; we comply. Stories may reference patterns observed in this layer where the reader can verify the pattern at end-of-day from NSE bhavcopy.
• Internal regime models. Quantitative models built on public and licensed data are described editorially; model outputs are characterised in plain English, not published as raw numbers.

Source-line conventions

• A specific quantitative claim (a price, a percentage, a count) is attributed inline to the primary source where the number is contested or unusual.
• A widely available number (e.g., the Nifty 50 close on a given day) is not source-line cited but the underlying source is reproducible.
• A number described as "our estimate" or "our model read" is the editor's estimate, not a primary-source figure, and is labelled as such in the running copy.

What we do not publish

• Tick-by-tick quotes from licensed real-time feeds
• Order-book depth from licensed real-time feeds
• Member-level position data not disclosed by the exchange
• Insider information shared with us in confidence outside a regulator-disclosed channel

Image sourcing

Editorial images are sourced from Creative Commons or public-domain archives where possible (Wikimedia, government photo pools), and otherwise embedded via the original publisher's host URL with attribution. We do not redistribute images from our own CDN by default. Where the original host drops the image, we may serve a cached copy from our own infrastructure with a clear "cached from publisher" note. Photo credits and source URLs are visible on every story. Read the editorial policy for the rule in detail.