Not investment advice. This is a personal learning log. The authors are not SEBI-registered investment advisers or research analysts. Nothing here is a recommendation to buy or sell any security. Intraday levels are published only after the session has closed, as a record of what was tested. The portfolio is simulated, and past results don't predict future ones. Do your own research.

NSE next-day intraday candidates

A living playbook for building an evening list of intraday candidates for the next NSE session. A trader sizes and executes the trades; the picker never places orders. Update this file after every session: log the brief and its outcome, and date any rule change under Lessons.

Ground rules

Never invent numbers. Cite a source for every figure: Yahoo Finance India, NSE or Moneycontrol for prices and volume; Screener.in for fundamentals; named news outlets for catalysts.
Levels come from the last close, never from live prices. Say how fresh the data is.
Post fewer names rather than forcing a pick. An empty list is a valid result.
Write numbers-first in plain English, using ₹ and NSE tickers.
Aim for modest, repeatable gains. No gambling.

When

Each trading evening at about 7:20 PM IST, for the next session. Check the NSE holiday calendar first (for example, 2 Oct is Gandhi Jayanti and the market is closed).

Step 1: tradable universe (owner-approved 30 Sep 2026)

Base pool: NSE F&O stocks plus the Nifty Midcap 150.
Liquidity floor: 20-day median daily turnover of at least ₹100 crore.
Price: ₹100–₹3,000, so at least 5 shares fit the sizing.
Exclude: ASM/GSM surveillance, T2T or circuit-bound names, stocks in the F&O ban list, names with results or a corporate action the next session, and any live governance scandal or fraud news.
Fundamentals (ROE, ROCE, debt from Screener.in) are a tie-breaker and risk note only, not a gate.

Step 2: price fit and sizing

The trader risks at most ₹150 per trade. Sizing (trader's rule, 30 Sep 2026): shares = ₹150 ÷ (entry − SL limit), where the SL limit is the stop minus ₹4. Example: a 1% stop on a ₹2,000 stock gives 150 ÷ 24 = 6 shares. Notional stays at or under ₹15,000, so prefer stocks priced at about ₹3,000 or less, where at least 5 shares fit. Quote share counts on this basis.

Step 3: end-of-day technicals, weighted to recent momentum

Price above a rising 20-day and 50-day moving average (DMA). A flat DMA means reject.
Stronger than the Nifty over the past 1–4 weeks, measured as the 20-day return gap in percentage points.
A breakout, or higher highs and higher lows. Avoid fresh 52-week lows or stocks below their 50- or 200-DMA.
A close in the top part of the day's range on volume above the 20-day average (about 1.3x or more is good). Don't relist a name that closed at its day low after a target hit.
RSI(14) strong but not stretched, about 55–70.
Weak market: while the Nifty is below a falling 50-DMA, list at most 2 names, and only those clearly beating the index (about +5 points of relative strength or more). Skip everything if the index gaps sharply lower.

Step 3b: noise against the stop

Use or30_med (the median first-30-minute high-to-low range over the last 20 sessions), or30_vs_stop and atr14_pct from eod_tech.py. - or30_vs_stop above 1.0 means the normal opening swing is bigger than the stop. Drop the name, or flag it as noisy. - Treat 0.9–1.1 as borderline and say so. - atr14_pct above 2.5 is a backup flag, and the only test when 15-minute data is missing. - A wider stop needs a wider target. For a noisy name offered with a wider stop, the worst-case risk is the wider stop plus the trader's ₹4 SL-trigger buffer. Size it at ₹150 divided by that risk. The target must be at least 1.5× that risk away and sit at a realistic level. If no such level exists, drop the name. Whether to take the wider stop is the owner's call.

Step 4: levels

Limit entry near the last close. Stop about 1% below entry.
Target 1.5–2R (R = entry minus stop), set at a realistic level such as a recent high or nearby resistance.
The target stays fixed after a gap (the owner's rule), so avoid targets that only just clear 1.5R.
Minimum 1.5R after any gap shift. State the open price above which the trade is skipped: target minus 1.5R.

Trader's execution rules (so skip lines match)

Gap rule, judged on the official open:
An open less than 1% above entry moves the entry and stop up by the gap, and the target stays fixed.
An open 1% or more above entry, at or below the stop, or leaving less than 1.5R to the target, is a skip.
After a 1R gain (updated 30 Sep 2026): the stop trails at the higher of entry + ₹1.50 or 1% below the day's high, and only moves up. The target is a fixed limit sell at the picker's level.
An unfilled entry is cancelled at 12:00. Positions are squared off by 3:05 PM, before Zerodha's 3:12 PM cutoff. Nothing is held overnight.
Confirmation entry (owner-approved 1 Oct 2026, from Mon 5 Oct): while the Nifty is below a falling 50-DMA, there's no buy at the open. At about 9:30, buy only if the price is at or above the entry and the low since 9:15 has held above the weak-hold level, with a limit no higher than the 1.5R skip price. Otherwise skip. Mark such briefs "confirmation entry" and always give a weak-hold level.
Open-mode weak skip (owner-approved, from Mon 5 Oct): on normal open-entry days the trader skips at the open if the last price is at or below weak-hold + 0.25R. So set the weak-hold level at least about 0.5R below entry, or ordinary opening wobble will trigger a skip.
Confirmation days need room: the 09:30 buy limit is capped at the 1.5R skip price, so a pick with exactly 1.5R reward can only fill at its entry and rarely confirms. On confirmation-entry days prefer reward of about 1.8R or more. Mark the brief header "confirmation entry" (the trader's runner uses entry_mode "confirmation") and give weak-hold and the 1.5R skip price for every name.
The owner approves each trade the evening before.

Output (2–3 names, fewer if nothing qualifies)

A one-line market context (Nifty close and bias), plus the next session's calendar.
For each name (the trader asked for these fields on 30 Sep 2026):
Ticker, and whether it's the primary or a secondary pick.
A quality or risk note, including fundamentals where relevant.
Setup: close, high, low, DMAs, RSI, volume against its average, and relative strength against the Nifty.
Noise: or30_med, or30_vs_stop and atr14_pct.
Entry, stop, target, reward-to-risk, share count per Step 2, the 1% skip price (entry × 1.01), the 1.5R skip price (target − 1.5R), and a weak-hold level that must hold in the first 15–30 minutes.
The reason, with sources.
Short reasons for rejecting the obvious names.
Which name to take if there's only one slot.

Script

scripts/eod_tech.py (Python stdlib, Yahoo chart API) takes SYM ... [--date YYYY-MM-DD] [--out file]. It outputs: - O/H/L/C/V, dma20/dma50 with rising flags, and above_20/50. - rsi14, vol_x, close_pos, ret20 and rs20_pp. - Entry, stop, targets, and the share count for a ₹150 risk (check it against the Step 2 formula with the ₹4 SL buffer). - price_fit and stale. - atr14/atr14_pct, or30_med/or30_med_pct/or30_n, or30_vs_stop and noise_flag. The thresholds are constants at the top of the file.

Before 15:30 it drops today's unfinished bar. stale: true means Yahoo is missing the latest daily bar, so verify against NSE.

Daily brief log (newest last; add a row every session)

Session	Candidates (entry / stop / target)	Outcome	Lesson
Mon 28 Sep 2026	ZYDUSLIFE 1,200 / 1,188 / 1,218–1,224 (taken); APOLLOHOSP 8,880 / 8,790 / 9,015–9,060	Skipped. Opened at 1,210.70, at or above the skip line.	A stock closing near its high on strong volume often gaps up, so leave room between the entry and the skip line.
Tue 29 Sep 2026	ZYDUSLIFE 1,212 / 1,200 / 1,230; DIVISLAB 9,530 / 9,435 / 9,670–9,720 (failed price fit)	Target hit. Opened at 1,217.6, so the entry and stop shifted up and the target stayed. Bought 12 at 1,217.50 and sold at 1,230 within 2 minutes. About ₹134 net.	Keeping the target fixed after a gap was right, because the high was 1,232.
Wed 30 Sep 2026	EMCURE 1,959 / 1,939 / 1,995 (6 shares, skip above 1,965); LALPATHLAB 1,950 / 1,931 / 1,985–1,995 (watch)	Stopped out. Opened at 1,960, bought 6 and stopped at 1,940 at 9:42. −₹120 gross, about −₹132 net. The first 30 minutes ranged 29.2 points.	EMCURE's median first-30-minute range was 26.1 points against a 20-point stop (1.31×). Added Step 3b and the weak-market filter.
Thu 1 Oct 2026	DRREDDY 1,233 / 1,221 / 1,255 (9 shares, skip above 1,237); no secondary	Stopped out. Official open 1,233 (no shift). Bought 9 at 1,225.90 at 09:15:08, stop filled 1,220.80 at 09:16:22. −₹45.90 gross, about −₹58 net (est.), −0.43R. Nifty −0.41% at the open.	Second straight stop-out in the first minutes, despite clean noise (or30_vs_stop 0.84). In a weak market the open itself sold off. Flag names whose open is weak relative to the entry; consider a confirmation entry (owner's call).

Running tally after 3 real trades: about −₹56 net (estimated). Judge the method on the trader's R scorecard after about 20 trades.

Lessons and rule changes (dated)

25 Sep 2026: use strong liquid names, momentum technicals and modest targets. No live data.
28 Sep 2026: added the price fit (share price about ₹3,000 or less). Skip lines follow the trader's gap rule.
29 Sep 2026: the target stays fixed after a gap, so set realistic targets with room. Don't relist a name that closed at its day low after a target hit.
30 Sep 2026: added the noise check (Step 3b), the weak-market filter, the 1.5R minimum after a gap, and wider stop means wider target. The owner widened the universe beyond fundamentally screened stocks (Step 1).
30 Sep 2026: the trader changed sizing to ₹150 ÷ (entry − (stop − ₹4)) and the post-1R trail to the higher of entry + ₹1.50 or 1% below the day's high. Each pick now lists the 1% skip price, the 1.5R skip price, a weak-hold level and or30_vs_stop.
30 Sep 2026: the owner approved the new universe rules in Step 1 (F&O + Midcap 150, ₹100 cr turnover, ₹100–3,000 price, exclusions), effective from the Sun 4 Oct list.
1 Oct 2026: the trader asked that each brief flag names where the open is weak relative to the entry (e.g. state the weak-hold level and note if the index is set to open lower). Two opening stop-outs in a weak market; the owner approved a confirmation entry while the market is weak (see Trader's execution rules).

Improving this skill

After every session, add a log row with the brief, the actual open, the fill, the exit, the P&L and one lesson. Every week or two, review the gap-skip rate, target-versus-stop hit rate, R per trade, and which setups worked. Change a rule only when several sessions support it, and date it under Lessons.