A chart that stops moving is not a candle. It is the exchange refusing to print a price beyond a rule. If you treat that freeze as "strong close", "no sellers", or "breakout held", you are practising a story the microstructure never told.
Indian cash and derivatives tapes run inside two different circuit logics. Mix them up and your replay log credits setups that only "worked" because trading was not allowed.
The problem: the last print looks like a decision
On a finished daily bar, a name that hit its upper band looks like a full-bodied close. On a five-minute chart, you get a run of identical highs, then a gap when the band is flexed or the next session opens. Live, nobody could lift that offer. In replay, a market order "fills" at the frozen print because the simulator has no queue.
That is not a small distortion. It is the difference between practising an entry and practising a cartoon of an entry.
The same confusion hits at index level. A nationwide halt is not "Nifty found support." It is a coordinated pause. The first bar after the reopen is an auction residue, not the next bar of the same auction you were in at 11:00.
The misconception: circuits are rare drama, so ignore them
Traders remember the days the whole market stopped. Those days are famous, so they feel like exceptions. Stock-level bands are not exceptions. NSE publishes price-band hitters as ordinary market data. Surveillance lists can tighten how far a name is allowed to travel. Names under extra watch, and many thinner cash names, hit those walls far more often than index futures do.
The other misconception is the reverse: treating every flat high as "it must have been circuit." Sometimes the tape is just thin. You still should not practise a breakout into a wall you have not identified. Liquidity and a freeze can look similar on OHLC; they are not the same lesson.
Better framing: two machines, two chart artefacts
Market-wide index circuit breakers sit on Nifty 50 and Sensex. NSE's circuit-breakers page (retrieved 17 September 2026) states the system applies at 10%, 15% and 20% either way, triggered by whichever of Sensex or Nifty 50 is breached first, and that a trigger brings a coordinated halt in all equity and equity derivative markets nationwide. Limits are computed daily from the previous close. After a halt, the market is designed to reopen with a pre-open call auction, except where the published table says the session is done for the day.
NSE currently lists halt duration against clock time — for example, a 10% breach before 13:00 IST as a 45-minute halt plus a 15-minute pre-open; a 20% breach as remainder of the day. That table is the exchange's presentation of SEBI circular CIR/MRD/DP/25/2013 dated 3 September 2013, as implemented in NSE circular 85/2013 dated 11 October 2013. The original market-wide framework dates to SEBI circular SMDRPD/Policy/Cir-37/2001 (28 June 2001), in force from 2 July 2001. Check the live NSE circuit-breakers page before you memorise minutes. Exchanges have revised this framework before.
What the chart teaches, if you let it: a blank stretch, then a new opening print. That new print is not "continuation of the last five-minute bar." It is a restart. Treat it closer to an open than to the bar that died at the halt.
Stock-level price bands are a different machine. A security may only trade inside an exchange-set percentage band around a reference price. Hit the band and the name can stop, wait, or move only after the band is revised. NSE keeps a Price Bands & Surveillance Actions page because the permitted width is not the same for every name, and it is not fixed forever. Additional surveillance can narrow the room a stock has to move. Do not quote a 2% / 5% / 10% / 20% menu as if it were a permanent law of your practice name. Look up that name's current band on the exchange.
On a historical chart the artefact is simpler than the circular: price could not print through a wall. Your stop "beyond the spike" may sit in a region that never traded.
Workflow: the freeze tag before the setup tag
Run this on randomised NSE replay with dates hidden. One liquid large-cap or index chart per session is enough. Do not hunt famous crash days.
Before you press play
- Write whether today's sandbox is cash stock or index. Circuit logic differs.
- Write one setup rule. Freeze it.
During candle-by-candle replay
When you see any of these, stop tagging the setup and tag the tape:
- several bars with the same high or low and collapsing range
- a sudden empty stretch in clock time
- a reopen gap that does not look like an overnight gap
- a "breakout" that never trades through the level, only kisses it and dies
Log fields:
tape: normal | stock-band | index-halt | unsuredecision: enter | skip | flattenfill-fantasy: yes | no— would a live order have queued at a frozen print?setup-valid: yes | no— was the pattern even allowed to complete?
If tape is not normal, the setup column is commentary. Score process on whether you skipped, not on whether the next day opened green.
After twenty tagged freezes (not twenty trades)
Write one paragraph: did your rules keep firing into walls? Did you call a halt a reversal? That paragraph is the lesson. A green replay fill after a band hit is not.
Illiquid names hit stock bands more often; practising them as if they were Nifty-weight cash is how people learn courage they cannot fill. Keep intraday product constraints in the same notebook: a freeze near 15:20 is a different problem from a freeze at 10:10.
Limits of this practice
You cannot reconstruct the order book from OHLC. A historical freeze tells you trading was constrained. It does not tell you how long the queue was, whether the band was flexed intra-day, or whether your hypothetical size would have been a meaningful fraction of what actually traded.
Index-halt days are poor default data for an opening-range or pullback study. They are excellent data for one skill: noticing that the session you thought you were in has ended. Stock-band days are poor data for breakout statistics. They are useful for learning that Indian cash names do not always have permission to complete the pattern your rule described.
Replay Trader's historical NSE charts, candle-by-candle reveal, and randomised starts will still show the freeze. They will not warn you in a banner that "this was a circuit." That is the point. Your job is to notice structure that a finished-chart caption would have labelled for you.
A ₹999 thirty-day block is a finite place to collect freeze tags, not a promise that you will "know circuits." Ten practice decisions a day × 30 days ≈ 300 practice decisions — including the skips that should have been skips when the tape was not allowed to move.
Practise tagging frozen prints before you credit the setup. Join the waitlist.
Related reading
- Why Liquidity Matters When You Practise on NSE Names
- Equity Intraday Constraints in India: Practice Implications
- Gap Trading Practice on NSE Opens
- Why Small-Cap Charts Mislead in Practice
Educational note
Educational practice content only. Not investment advice. No guaranteed returns. Practice ≠ live profitability.