Markets · The Setup Scan
The Setup Scan
Every session, the engine tests both sides. LONG needs bullish structure and supportive filings; SHORT needs bearish structure and filed deterioration; bearish tape without that confirmation becomes HEDGE, and mixed evidence fails closed to NO_TRADE.
The scan
The symmetric Setup Scan tests both directions after every session. As of the 2026-07-31 close, 81 names read LONG, 32 SHORT, 51 HEDGE and 323 NO_TRADE across 487 fully evaluated names. 2 additional technical reads lacked a complete fundamentals/evaluation record, so no side was inferred for them. An outright SHORT requires bearish price structure and deterioration in official filed numbers; valuation alone can never create it; the highest-confluence read is KALYANKJIL at 100 out of 100.
BazaarBaaziSource & method
Search the scanned universeClose 2026-07-31
Type any symbol or company, pick a sector from the dropdown, or walk the tiers. Every name opens its full engine read: the scores, the reasons, the red flags, the filed quarter behind it, and the exact date the analysis is current to.
82 of 487 scanned names · every read current to the 2026-07-31 close
How the engine judges a setupThe method
Most screens quietly assume the answer must be long. This one scores both sides and is allowed to withhold a trade. LONG needs bullish price structure, supportive filed fundamentals and positive directional flow. SHORT needs bearish structure, weak flow and filed deterioration. When the tape is bearish but the company evidence remains sound, the verdict is HEDGE. When evidence conflicts, it is NO_TRADE. The fail-closed state is part of the model, not an empty result.
The long technical axis reads stage-two trends, breakouts, pullbacks to rising averages and relative strength. The short axis reads stage-four trends, 52-week breakdowns, failed rallies under falling averages, distribution and relative weakness. Official NSE-XBRL filings remain a separate axis: acceleration and quality support LONG, while earnings deterioration and balance-sheet stress are required for SHORT. Valuation changes conviction, but expensive never means short by itself.
Then the tape gets a signed vote. Chaikin money flow, the up-day versus down-day volume ratio, MFI and accumulation-distribution days support one side and weaken the other. Only LONG and SHORT get a side-aware entry, stop and 2R or 3R frame. HEDGE gets a removal condition; NO_TRADE gets the exact missing gate. Every verdict can therefore be challenged on its evidence and invalidation.
Everything here is deterministic rule code over real data. The same inputs produce the same scan, every reason traces to a price series or an official filing, and every card carries the date its analysis is current to. When the scan is wrong, it is wrong in the open, which is exactly how the desk wants it.
FAQ7 reader questions · AEO-eligible
The editorial line, distilled. Schema-marked for AI Overview and reader search.
What makes a setup near-perfect?
A directional side must clear its own price-structure, official-filing and money-flow gates with no blocking red flag. LONG needs bullish structure plus supportive results. SHORT needs bearish structure plus actual filed deterioration. A rich valuation or high call OI is never enough to manufacture a short.
Why do only 82 names qualify out of 500?
Because conflicting or incomplete evidence fails closed. A bullish chart on rotten earnings is not LONG; a bearish chart on strong filings is HEDGE at most; a valuation concern by itself is NO_TRADE. A small directional count is information about evidence quality, not a reason to loosen the gates.
How does the engine choose between SHORT and HEDGE?
Both need bearish price structure and weak directional money flow. SHORT additionally requires deterioration in official filed fundamentals and a side-aware invalidation above the short stop. If the tape is bearish but the filings remain sound, the engine emits HEDGE and deliberately withholds an entry, target and outright short thesis.
What is the Setup Score?
A 0 to 100 composite the engine prints per name: the best setup on each axis plus a small confluence bonus when multiple setups fire together, with a money-flow boost when the volume tape confirms. It is deterministic rule code over real data, versioned and repeatable, which is why every card can show its reasons instead of asking you to trust a black box.
Where do the fundamentals come from?
Official NSE-XBRL filings, paisa-exact, ingested by the BazaarBaazi earnings lab. Never scraped third-party ratios. Filed EPS is quarterly, so the engine sums the trailing four quarters into a TTM figure before any valuation math (feeding a quarterly EPS into a P/E is roughly four times wrong). Every card links the official filing it read.
How fresh is each analysis?
Every card prints the exact close its analysis is current to, currently 2026-07-31. The scan refreshes on the desk's nightly run after each trading session. If a scan is ever delayed, the date on the card says so plainly; a read here is never silently stale.
Is a near-perfect tag a buy recommendation?
No. It is a transparent, deterministic screen over real price structure and official filings, with every reason and its source on the card. It is editorial framing and a research starting point, not SEBI-registered investment advice. The engine even prints its risk frame so you can disagree with it precisely.
Go deeper
Hub
The market board
Indices and marquee names, color-graded across timeframes
Screener
The signed-reads screener
Filter the covered set by desk stance, Crack Score and Event Heat
Stocks
Signed stock views
Every covered name's full verdict
Earnings
The earnings desk
Official filed results, season-wide
Brief
The market brief
The dated record of every session