Known limitations

DocStoX would rather tell you where its data is thin than paper over the gap with a guess. This page lists the specific limits worth knowing before you rely on a number.

Quarterly and shareholding data lags the company's own filing#

Shareholding patterns, quarterly results and similar disclosures update once a company has actually filed them and DocStoX has processed the filing, not on the date the quarter ended. A company that reports late shows its previous quarter's numbers until the new filing lands. This is a lag in when the company disclosed the figure, not something unique to DocStoX. See Freshness and update timings for typical refresh cadence.

Banks and NBFCs do not report a comparable "revenue" line#

A lender's income statement does not map onto standard revenue and PAT growth metrics the way a manufacturer's or a services company's does; there is no single "revenue" line to compute a growth streak from. Metrics that depend on it, such as multi-year revenue growth streaks, show as unavailable for banks and NBFCs rather than a misleading number computed from the wrong line item. Fair value and the AI score both use a separate approach for lenders for the same reason.

Price history depth varies by company and by year#

Daily price history is usable from 2020 onward, and broadens as more companies list. A "52-week high" or a multi-year chart is only ever as deep as the history DocStoX actually holds for that company: for a stock that listed after 2020, that is its full history; for one that has traded much longer, DocStoX's own record still starts in 2020. See Coverage for the detail by year.

A small number of price series can also still carry a split or bonus that DocStoX has not been able to confirm and adjust for. Where that happens, an unadjusted corporate action can look like a large one-day price move that never actually happened to a holder. See How prices are adjusted for how this is detected and handled, and where it can still show through.

The AI score and fair value are not each individually fact-checked#

Both are computed the same deterministic way for every company from DocStoX's own data; a human does not review each individual company's score before it is shown. Where the underlying data for a company is missing or looks unreliable, the score treats that input as absent rather than guessing at it, and fair value withholds a number entirely for the categories of company its models have not been measured to handle well. See how each degrades in How the AI score is built and How fair value is calculated.

AI answers can be wrong#

DoXy and the MCP tools answer from DocStoX's own data, not from a model's memory, and show which data they used. That does not make an answer infallible: check the numbers behind an answer the same way you would check any analyst's work, especially before acting on it.

Quotes are delayed, and DocStoX is not a trading terminal#

During market hours, a quote is roughly 15 minutes behind the exchange, on every plan. DocStoX is built for research, not for executing or timing an order to the minute.

DocStoX does not give investment advice#

DocStoX is a research and information platform, not a SEBI-registered investment adviser. Scores, verdicts, fair value estimates and setups describe data; none of them is a recommendation to buy, hold or sell. For advice tailored to your situation, speak to a SEBI-registered investment adviser.