# PatternsRadar

> Screen 3,500+ NSE stocks on price, volume, delivery and technicals across 30 years of split-adjusted history, then test the scan against a year of sessions before you trust it.

PatternsRadar (https://dev.patternsradar.com/) is a stock screener for the National Stock Exchange of India (NSE), built by Bitmask (https://bitmask.in). A scan is written in Sift, a readable query language (`close > ema(21) and delivery_pct > 60`), or clicked together from filter chips. The two are views of one query. Every scan can be replayed across any window from 10 to 7,500 sessions (every session since 1995) to measure what its matches did next: win rate, median move, payoff and excess return against one of seven NSE indices, over up to four forward horizons of 1-60 sessions each, entering at the signal close or the next session's open, with an optional stop-loss and target that label each exit stop/target/time. A replay returns an equity curve, monthly breakdown, return histogram and full trade list, each exportable as CSV, in 90-140 ms. The replay's full configuration is available on every plan including Free; plans set a daily allowance, not an annual quota. Caveats: close-to-close, no brokerage, slippage or position sizing, and universe membership measured as of today (a survivorship lean). Data: 3,700+ instruments, 153 scannable fields, indicators and candlestick patterns, daily bars back to 1995, split- and bonus-adjusted, including delivery percentage: the share of traded volume actually taken to demat, which most screeners do not carry. Also scannable: per-stock F&O aggregates (futures OI, OI change, basis, put-call ratios; `from fno` scopes to the ~200 derivative stocks), point-in-time fundamentals (P/E, market cap, growth, shareholding, each valued as known on the scan date, coverage still backfilling), dividends by ex-date (yield, growth, CAGR, paying years and streak over any span of 1–10 years: `dividend_yield(1y) > 3% and dividend_years(5y) == 5`), and NSE sector/industry classification (`sector is "Information Technology"`). End-of-day only: no intraday data, no live feed, NSE only.

Screening is free with no account. Paid plans (Pro ₹400/month, Max ₹2,000/month) add end-of-day email alerts, webhooks and higher API limits. Every plan, including Free, has REST API and MCP access. Latest session in the dataset: 2026-09-25.

Every link below is a markdown document. Each is the twin of an HTML page at the same URL without `.md`, and every public page on the site has one — append `.md` to its URL (e.g. /learn/rsi.md, /s/RELIANCE.md) or request the page with `Accept: text/markdown`. Every page declares its twin with `rel="alternate" type="text/markdown"` and this file with `rel="describedby"`, both as a <link> in its head and as an HTTP Link header. The whole set, by section: https://dev.patternsradar.com/sitemap.md. The exhaustive list of HTML pages, including every instrument, is the sitemap index at https://dev.patternsradar.com/sitemap.xml.

## Docs

- [Sift language reference](https://dev.patternsradar.com/docs/sift.md): every field, indicator, window function, operator and candlestick pattern; generated from the same catalog the compiler reads
- [API & MCP reference](https://dev.patternsradar.com/docs/api.md): REST endpoints with examples, and the MCP server for agents
- [MCP server](https://dev.patternsradar.com/docs/mcp.md): connect Claude, Codex or any MCP client to NSE end-of-day data — the server hands the agent the whole query language
- [Methodology](https://dev.patternsradar.com/methodology.md): where the data comes from, how it is refreshed and verified, what a universe tier is, and exactly what the hit-rate replay measures and leaves out
- [Glossary](https://dev.patternsradar.com/learn.md): 57 indicators, fields and candlestick patterns, each defined with a working scan — e.g. [Multi-year highs, lows and returns](https://dev.patternsradar.com/learn/multi-year-highs-and-lows.md), [52-week high and low](https://dev.patternsradar.com/learn/52-week-high-low.md), [Pivot points](https://dev.patternsradar.com/learn/pivot-points.md)

## Scan library

142 prebuilt scans, each on its own page with its query, an explanation and a backtest. By category:

- [Breakout stock scanners for NSE](https://dev.patternsradar.com/scans/breakouts.md): 17 scans — Free breakout screeners for NSE stocks. 52-week highs, volume breakouts, Donchian and Bollinger channel breaks, each on its own page with a hit-rate replay.
- [Reversal stock scanners for NSE](https://dev.patternsradar.com/scans/reversals.md): 32 scans — Oversold and overbought screeners for NSE stocks. RSI, stochastic, CCI, MACD crosses and the golden cross, each paired with the trend filter that makes it usable.
- [Momentum stock scanners for NSE](https://dev.patternsradar.com/scans/momentum.md): 20 scans — Momentum screeners for NSE stocks. Relative strength leaders, stacked moving averages, ADX trends, OHL scans, and pullback entries inside confirmed uptrends.
- [Delivery percentage scanners for NSE](https://dev.patternsradar.com/scans/delivery.md): 11 scans — Delivery-based screeners for NSE stocks. Delivery surges, quiet accumulation, institutional-grade delivery days, and the churn that follows a volume spike.
- [Volatility stock scanners for NSE](https://dev.patternsradar.com/scans/volatility.md): 16 scans — Volatility screeners for NSE stocks. Bollinger squeezes, NR7 days, range expansions, gaps that held and gaps that failed, plus both 52-week extremes.
- [Candlestick pattern scanners for NSE](https://dev.patternsradar.com/scans/candlestick-patterns.md): 15 scans — Candlestick pattern screeners for NSE stocks. Engulfing bars, hammers, harami, morning and evening stars, three white soldiers, three black crows, each filtered by the context that gives it meaning.
- [F&O derivatives scanners for NSE](https://dev.patternsradar.com/scans/derivatives.md): 11 scans — Open interest screeners for NSE F&O stocks. Long and short buildups, futures premium and discount, put-call ratio extremes: positioning the cash tape cannot show.
- [Fundamental stock scanners for NSE](https://dev.patternsradar.com/scans/fundamentals.md): 20 scans — Fundamental screeners for NSE stocks. Dividend yield, growth and consistency, promoter buying and selling, shareholding patterns, market-cap filters and sector scans, all on point-in-time data.

## Guides

- [Scans that say "most of the last ten days"](https://dev.patternsradar.com/blog/count-the-bars-scans-for-most-of-the-time.md): How count() and countstreak() let a scan ask for seven of the last ten days instead of all ten, with four NSE scans built on them and the limits of both functions.
- [Scanning the candle body, not the close](https://dev.patternsradar.com/blog/scanning-the-candle-body-not-the-close.md): Most scans read the close and ignore the rest of the candle. abs(), greatest() and least() in Sift let a scan describe the body: its size against the stock's own norm, and where the whole of it sits. Three ready NSE scans.
- [How to find momentum stocks on the NSE](https://dev.patternsradar.com/blog/momentum-stocks-screener-nse.md): The five fields a screener uses to measure momentum, the filters that stop a strength ranking turning into a chase list, and four ready NSE scans.
- [RSI divergence screener for NSE stocks](https://dev.patternsradar.com/blog/rsi-divergence-screener-nse.md): What regular bullish and bearish RSI divergence measure, why a single-bar scan has to approximate them with offsets, what that approximation costs, and two ready NSE scans.
- [Stock screening in India: a complete guide](https://dev.patternsradar.com/blog/stock-screening-india-guide.md): What a stock screener does and does not do, end-of-day versus realtime, the eight setup families for NSE stocks, why delivery percentage matters here, and how to check a scan before trusting it.
- [Candlestick patterns need context](https://dev.patternsradar.com/blog/candlestick-pattern-screener-nse.md): Why a raw candlestick scan returns noise on the NSE, the three context conditions that make a pattern worth looking at, and the scans with each one built in.
- [Translating Chartink formulas to Sift](https://dev.patternsradar.com/blog/chartink-formulas-to-sift.md): Side-by-side translations of the four Chartink idioms: the crossover, the streak, the N-day high and the volume multiple, plus what has no equivalent here.
- [How to read a scan's hit rate](https://dev.patternsradar.com/blog/how-to-read-scan-hit-rate.md): What the 1, 5 and 20-day columns in the hit-rate panel measure, the two misreadings that flatter bad scans, and what the replay deliberately leaves out.
- [Screen NSE stocks with Claude over MCP](https://dev.patternsradar.com/blog/screen-nse-stocks-with-claude-mcp.md): Connect Claude Code or any MCP client to 30 years of NSE end-of-day data: run screener scans in plain English, check them against history in the same conversation, and save the ones that survive.
- [The 52-week high strategy on the NSE](https://dev.patternsradar.com/blog/52-week-high-strategy-nse.md): The overhead-supply logic behind buying 52-week highs, three entry variants with the scans that run them, and the three ways the strategy loses money.
- [Bollinger band squeeze screener for NSE](https://dev.patternsradar.com/blog/bollinger-band-squeeze-screener.md): How to measure a squeeze with normalised band width, why compression says nothing about direction, the NR7 single-bar version, and the NSE scans for both.
- [Choosing a stock screener: what to look for](https://dev.patternsradar.com/blog/chartink-alternative-with-backtesting.md): Five capabilities that separate stock screeners: outcome measurement, event conditions, delivery data, data depth and an API, with what each one is worth.
- [Delivery percentage explained](https://dev.patternsradar.com/blog/delivery-percentage-explained.md): What delivery percentage measures, why a fixed threshold like 60% is the wrong way to read it, and the NSE scans built on deviation from a stock's own baseline.
- [Golden cross and death cross on NSE stocks](https://dev.patternsradar.com/blog/golden-cross-death-cross-nse.md): What the 50/200-day crosses record, why they fire late by design, the whipsaw tax in sideways markets, and how to check their hit rate on NSE stocks.
- [How to backtest a stock scan, no code needed](https://dev.patternsradar.com/blog/how-to-backtest-a-stock-scan.md): A five-step method for testing a screener scan against history, from stating the claim to spotting the three biases that fake a good result.
- [The Open High Low (OHL) scan for NSE stocks](https://dev.patternsradar.com/blog/open-high-low-scan-nse.md): What the OHL condition actually measures, why open = low reads as buyer control, and how to run both sides of it across the NSE on end-of-day data, free.
- [RSI screener guide for NSE stocks](https://dev.patternsradar.com/blog/rsi-screener-guide.md): Why a bare RSI-below-30 scan returns falling knives, the one condition that fixes it, the RSI levels that matter inside trending stocks, and four ready NSE scans.
- [A swing trading screener for India](https://dev.patternsradar.com/blog/swing-trading-screener-india.md): A working swing screen in three layers: trend, setup, trigger. How to build each one for NSE stocks, which presets express them, and how to check the screen against history.
- [Volume breakout screener for Indian stocks](https://dev.patternsradar.com/blog/volume-breakout-screener-india.md): Why volume is the one input that can't be faked, what 2× average volume actually implies, three ready NSE volume breakout scans, and the mistakes that make volume screeners useless.

## Comparisons

- [Chartink alternative](https://dev.patternsradar.com/compare/chartink-alternative.md): Where Chartink's backtest stops, what it still does better, and what a forward-outcome replay adds.
- [PatternsRadar vs Chartink](https://dev.patternsradar.com/compare/patternsradar-vs-chartink.md): The two screeners feature by feature: backtesting, event conditions, data depth, alerts, API.
- [Trendlyne alternative](https://dev.patternsradar.com/compare/trendlyne-alternative.md): Trendlyne meters backtesting at roughly 200 runs a year on GuruQ. Here the replay runs on every scan, at full depth, from ₹0.
- [PatternsRadar vs Trendlyne](https://dev.patternsradar.com/compare/patternsradar-vs-trendlyne.md): Trendlyne is a research platform with a screener inside. PatternsRadar is a technical screener with testing built in. Backtest quotas, query models, delivery data and APIs, compared honestly.
- [StockEdge alternative](https://dev.patternsradar.com/compare/stockedge-alternative.md): Write any scan instead of picking from 500, replay every one across up to thirty years, and screen delivery percentage with evidence. Free, no account.
- [PatternsRadar vs StockEdge](https://dev.patternsradar.com/compare/patternsradar-vs-stockedge.md): StockEdge curates 500+ readymade end-of-day scans. PatternsRadar lets you write and backtest your own. Scan models, delivery analysis, custom-scan quotas and pricing, compared honestly.
- [TradingView alternative](https://dev.patternsradar.com/compare/tradingview-screener-alternative.md): An NSE-focused alternative to TradingView's global screener: scan-level backtesting, delivery percentage, event conditions and thirty years of daily history.
- [PatternsRadar vs TradingView](https://dev.patternsradar.com/compare/patternsradar-vs-tradingview.md): TradingView's screener is real-time, global and chart-first. PatternsRadar is end-of-day, NSE-only and scan-first. Backtesting scope, delivery data, history depth and pricing, compared honestly.
- [Screener.in alternative](https://dev.patternsradar.com/compare/screener-in-alternative.md): If you want fundamentals, stay on Screener.in; it is the best there is. If you have hit its technical ceiling, that is the half PatternsRadar covers.
- [PatternsRadar vs Screener.in](https://dev.patternsradar.com/compare/patternsradar-vs-screener-in.md): Screener.in screens businesses from ten years of financials. PatternsRadar screens price behaviour from thirty years of daily bars, with backtesting. Mostly different jobs, including the narrow overlap.
- [Tickertape alternative](https://dev.patternsradar.com/compare/tickertape-alternative.md): Tickertape's screener stops at menu filters: SMA, EMA and RSI thresholds, partly Pro-gated, untestable. PatternsRadar picks up there, free.
- [PatternsRadar vs Tickertape](https://dev.patternsradar.com/compare/patternsradar-vs-tickertape.md): Tickertape is a broad research app whose screener is menu filters. PatternsRadar is a technical screener with a query language and backtesting. Filters against sentences, scorecards against evidence.
- [Strike alternative](https://dev.patternsradar.com/compare/strike-money-alternative.md): Strike is a real-time trading terminal, paid after a 7-day trial. PatternsRadar is the end-of-day half: an NSE screener with a documented replay on every scan, free to use.
- [PatternsRadar vs Strike](https://dev.patternsradar.com/compare/patternsradar-vs-strike-money.md): Strike.money scans live with options analytics and delivery scanners. PatternsRadar scans after the close with a configurable replay on every scan. Timing, testing, query models and pricing, compared honestly.
- [Intradayscreener alternative](https://dev.patternsradar.com/compare/intradayscreener-alternative.md): Intradayscreener.com is a live scanner for the session. PatternsRadar is the end-of-day counterpart: write a swing setup, replay it across up to thirty years, screen delivery.
- [PatternsRadar vs Intradayscreener](https://dev.patternsradar.com/compare/patternsradar-vs-intradayscreener.md): Intradayscreener scans NSE live with breakout alerts and option tools. PatternsRadar scans after the close with a configurable replay on every scan. Timing, testing, query models and pricing.
- [MarketsMojo alternative](https://dev.patternsradar.com/compare/marketsmojo-alternative.md): MarketsMojo rates stocks and sells the calls, and its screener is for paying subscribers. PatternsRadar rates nothing: write a technical scan, test it, screen delivery, free.
- [PatternsRadar vs MarketsMojo](https://dev.patternsradar.com/compare/patternsradar-vs-marketsmojo.md): MarketsMojo is a research and advisory platform built around a composite score. PatternsRadar is a screener that never scores a stock. Scores against conditions, calls against replays, paywalled against free.
- [Moneycontrol alternative](https://dev.patternsradar.com/compare/moneycontrol-screener-alternative.md): Moneycontrol's screener is free threshold filters beside the news. PatternsRadar is the next step: write the setup, replay it across up to thirty years, screen delivery. Also free.
- [PatternsRadar vs Moneycontrol](https://dev.patternsradar.com/compare/patternsradar-vs-moneycontrol-screener.md): Moneycontrol's screener is a market-stats utility with fundamental and technical filters. PatternsRadar is a technical screener with a query language and a configurable replay on every scan.

## Optional

- [The screener](https://dev.patternsradar.com/screener.md): the editor itself — chips and code over a results grid
- [Chartink scan converter](https://dev.patternsradar.com/tools/chartink-converter.md): paste a Chartink `scan_clause`, get the same scan in Sift, and see what it matches today
- [Pricing](https://dev.patternsradar.com/pricing.md): plans and limits
- [About](https://dev.patternsradar.com/about.md): who builds this, where the data comes from, how it is verified
- The instrument directory at https://dev.patternsradar.com/stocks.md: every listed instrument, browsable by sector (https://dev.patternsradar.com/stocks/sector/information-technology.md and 21 others) or by first letter (https://dev.patternsradar.com/stocks/a.md … https://dev.patternsradar.com/stocks/0-9.md), which is the complete index — roughly a third of the instruments carry no sector. Machine-readable as `GET https://api-dev.patternsradar.com/v1/instruments?letter=|sector=|industry=`, no key needed
- Symbol pages at https://dev.patternsradar.com/s/SYMBOL.md (e.g. https://dev.patternsradar.com/s/RELIANCE.md): per NSE instrument — price history, delivery percentage and technicals, plus quarterly results, shareholding and the filing calendar where they are on file. The full symbol list is `GET https://api-dev.patternsradar.com/v1/symbols` (no key needed); every instrument page is listed in https://dev.patternsradar.com/sitemaps/instruments
- [sitemap.md](https://dev.patternsradar.com/sitemap.md): every markdown document on the site, by section
- [llms-full.txt](https://dev.patternsradar.com/llms-full.txt): this file expanded — every scan's query and explanation, the category primers, and the MCP tool surface inline
- [sitemap.xml](https://dev.patternsradar.com/sitemap.xml): the sitemap index of every HTML page, in four sections — site, content, community, instruments — plus a fifth, markdown, listing every page's markdown twin
- [ai-catalog.json](https://dev.patternsradar.com/.well-known/ai-catalog.json): Agentic Resource Discovery — the MCP server (with its [server card](https://dev.patternsradar.com/.well-known/mcp-server-card.json)), the REST API and these documents, typed for an agent