How we research, write, and correct

We publish what the historical data supports, and nothing beyond it. Every figure on this site traces back to a market price. Where we interpret those numbers, we say so plainly and show the arithmetic. Where the data cannot answer a question — and it very often cannot — we say that too, rather than filling the gap with speculation.

The single rule that governs everything here: we never predict.No price targets, no "top picks for 2027", no buy or sell calls. Those are the stock-in-trade of Indian market media and they are precisely what ReturnScreener was built as an alternative to.

All analytics are computed from publicly available market data for companies in the Nifty 500 universe, with price history going back up to 20 years. Prices and every derived metric refresh each trading day.

Returns are price-only: they exclude dividends, bonuses adjusted outside the price series, brokerage, and taxes. Rankings cover companies currently in our universe, which means they carry survivorship bias — businesses that delisted or failed are not represented. We state this on the ranking pages themselves rather than burying it here. The full calculation method, with worked examples and known limitations, is on our Methodology page.

Written guides are researched and authored by a named human being, not generated in bulk. Each one is written against our own dataset — where a guide cites an example, that example is a real company with real numbers pulled live from the platform, not an invented illustration.

We use AI tools for drafting assistance, editing, and code, in the same way we use a spellchecker or an IDE. No guide is published without a human reading it end to end, verifying every factual claim against the data, and taking responsibility for it. We do not publish content at a volume that would make that impossible.

ReturnScreener is free to read and is funded by advertising. Advertisers have no influence whatsoever over our rankings, our data, or what we write. Rankings are produced mechanically from price data — no company can pay to appear in one, be removed from one, or change its position.

If we ever introduce affiliate links or sponsored content, they will be labelled clearly and unambiguously at the point of appearance. We have no paid placements today.

We will get things wrong. When we do, we want to know quickly and fix it publicly rather than quietly. If a return, price, CAGR, or written claim looks wrong, email hello@returnscreener.com with the page and the figure — corrections are the first thing we look at each day.

• Data errors are investigated at source and fixed in the next daily refresh.
• Errors in written guides are corrected in place, and the "last updated" date changes.
• Where a correction materially changes a guide's conclusion, we note what changed at the foot of the page rather than silently rewriting it.

• Publish price targets, forecasts, or buy/sell recommendations.
• Accept payment to feature, promote, or rank a company.
• Present historical performance as evidence of future performance.
• Publish mass-generated articles without human research and review.
• Claim qualifications or registrations we do not hold.