How this data is built

Primary sources only. Financial statements come from the XBRL filings companies submit to the exchanges. Prices come from the exchanges’ own end-of-day bhavcopy files. Shareholding, insider dealing, pledges, bulk and block deals, open offers, AGM voting results and IPO offer documents each come from the exchange or SEBI filing that publishes them.

We do not license a consolidated data feed. That is a deliberate cost: it means building and maintaining a parser per source. It is also the reason figures here can be traced to a specific filing on a specific date rather than to a vendor’s database.

Companies are keyed on ISIN, and the NSE and BSE master lists are merged on it. Ticker symbols and BSE scrip codes change — companies rename, re-list, and move between series — so a symbol is a label, not an identity. Keying on the symbol silently splits one company’s history in two the day it renames.

A company listed on only one exchange is normal and not treated as an error. Where the two masters disagree, the NSE name is preferred and BSE fills the gaps. Symbol changes are detected on every run and recorded, so historical data survives a rename rather than being orphaned behind the old ticker.

A 1:1 bonus issue halves the share price overnight without anyone losing money. Charting the raw price would show a 50% crash. So price history is adjusted: for splits, bonus issues and rights issues, a factor is computed per event and applied to every bar before the ex-date, and the cumulative product of those factors gives the adjusted close.

Dividends are deliberately excluded from the default adjusted series. That follows Indian market convention, where quoted history is not dividend-adjusted — including them would make our chart disagree with every other chart a reader compares it against. Dividends feed a separate total-return calculation instead.

Where a corporate action is filed without the ratio or face-value fields needed to compute its factor, the factor is 1 — the event is recorded but no adjustment is invented from an incomplete filing.

Companies restate results. When a revised filing arrives, the exchange issues it as a new document rather than replacing the original, and we do the same: reported figures are stored append-only, with each version tagged by its filing date and the latest marked as current. Nothing is overwritten.

This matters more than it sounds. Overwriting would mean a figure you read here last quarter could silently change with no record that it ever differed — and revisions are not rare. Keeping every version means the history of what was reported, and when, stays recoverable.

Companies with subsidiaries file both consolidated and standalone accounts, and the two can differ substantially. Consolidated figures are shown by default, because they describe the whole group a shareholder owns. Where a company has filed only standalone results for a period, those are shown and the page says so — rather than leaving a gap or quietly mixing the two bases across a trend line.

Some filings are internally inconsistent. A minority of fund-utilisation statements declare amounts in units that disagree with the figures by a clean power of ten. A plausible-looking fix would be to rescale them. We do not: the read path refuses the rupee figures and says so, because silently rescaling a filer’s numbers means publishing a figure nobody filed.

The same rule applies to voting tallies that do not reconcile, and to subscription figures where the exchange’s own category rows would double-count if naively summed. A visible gap is a smaller error than a confident wrong number.

These are current and stated plainly rather than discovered by a reader:

• Prices are end-of-day, not live. There is no intraday or real-time quote anywhere on this site, and no tick data.
• No analyst estimates or consensus. Nothing here compares a result against what analysts expected, because we hold no estimates data.
• XBRL history has a floor. Structured filings begin around 2019. Longer history, where shown, is extracted from annual-report PDFs and is labelled with its verification status.
• Draft IPO documents are mostly blank. At DRHP stage the price-dependent fields are literal placeholders in the document itself, so those issues genuinely have no price band to show.
• Extraction errors exist. Parsing thousands of filings from several sources is not error-free. Where a figure looks wrong, the primary filing governs — and we would like to hear about it.

EquityTale is not registered with SEBI as an Investment Adviser or a Research Analyst, and it publishes no advice. Concretely, and by design rather than omission:

• No buy, sell or hold calls, and no price targets.
• No grey-market premium for IPOs. GMP figures are unofficial quotes with no verifiable source, and we will not publish a price expectation we cannot attribute.
• No allotment-status lookup, and no “should I apply” verdict on an issue.
• No listing-gain estimates for unlisted issues, and no index predictions.

What the site does instead is show what was filed, by whom, and when — and leave the conclusion to you.

If a figure here disagrees with the filing it claims to come from, that is a bug and we want to fix it. Email contact@equitytale.com with the company and the figure. Corrections are made at the source — the extraction rule — so the same error is fixed everywhere it occurred, not patched on one page.