NSE EOD Analysis v3 · batch

The 23500–24000 corridor creates a 501-point structural cage with spot (23639) sitting just 139 pts above the triple PE-support floor, leaving only 361 pts of upside before hitting the quad CE-resistance ceiling. Monthly PCR of 1.078 introduces a MR_LONG micro-divergence on dips to 23500, but macro SHORT_BUILDUP dominance across near-term expiries overrides this signal entirely.

Defensive utilities (power transmission, telecom infrastructure) and isolated pharma names show relative strength, but no sector qualifies for BUY conviction under AVOID_ALL OI cap and VIX 21.52 ELEVATED regime.

The market is in a confirmed BEAR regime (score -1.40) with the OI conviction cap at AVOID_ALL — all long positions should be protected with tight stops and no new equity longs initiated under current conditions. The weekly PCR of 0.697 and next-week PCR of 0.543 reflect near-unanimous institutional bearishness, while VIX at 21.52 (98.8th percentile, RISING) adds a volatility premium that further penalizes long entries. Spot at 23639 sits only 139 pts above the triple-wall PE support at 23500; a clean break below 23500 would expose the 23000–22800 zone with no structural floor visible in OI data. Cash exposure should be reduced or hedged; leveraged longs and momentum chases are explicitly disallowed under this regime. Re-entry conditions require all three of the following to be satisfied simultaneously: (1) weekly PCR must recover above 0.80 signaling a shift from EXTREMELY_BEARISH to NEUTRAL, (2) VIX must drop below 18 before any BUY conviction is permitted, and (3) the FO quadrant must exit SHORT_BUILDUP — watch for LONG_BUILDUP or at minimum NEUTRAL quadrant confirmation before any new watchlist entries are considered.

1. OI CONVICTION CAP CHECK FIRST:
cap=AVOID_ALL, score=-1.40. Divergence=YES (weekly bearish PCR 0.697/0.543, monthly neutral PCR 1.078). Hard rule: watchlist[] MUST be empty. All Stage 2 candidates → avoid[] with OI cap reason. Risk flag for divergence must be added.

2. REGIME + VIX:
BEAR, HIGH confidence. VIX 21.52 ELEVATED zone (18-22), RISING, 98.8th percentile 52W. No BUY conviction allowed even without AVOID_ALL cap. FII net short -226,327 lots, cash -7,050 cr, 2-day consecutive selling streak. FO quadrant SHORT_BUILDUP dominant.

3. SECTOR ANALYSIS:
- E-Learning: MARKUP (5D 4.7%, 10D 4.0%, vol_ratio 1.8) - but under AVOID_ALL, irrelevant
- Telecom Infrastructure: HOLDING (5D 1.7%, 10D 5.5%)
- Aerospace & Defense: NEUTRAL, failed industry gate
- Pharmaceuticals: NEUTRAL, all windows negative
- Power: Mixed (NTPC industry MARKDOWN, JSWENERGY MARKDOWN sector)
- Banking: RED sector

4. CANDIDATES → ALL TO AVOID[] due to AVOID_ALL cap:
All 19 gate-cleared candidates move to avoid with reason format: "OI cap AVOID_ALL (score=-1.40): <cap_reason>"

Stage 2 already avoided: JINDALSAW, AGIIL, MUTHOOTFIN, TVSMOTOR, ASHOKLEY, ATALREAL, JAIBALAJI

5. STACKED WALLS:
- 24000: QUAD-WALL (all 4 expiries) - within ~1.5% of spot (spot 23639, 24000 = +1.5%). This is within ~1.5%, close to 1% threshold but slightly above. The threshold says "within 1% of spot" = 23639 ± 236 = 23403-23875. 24000 is 361 pts away = 1.5%, so just outside the 1% threshold. But 23500 support is 139 pts away = 0.59% from spot - WITHIN 1%. Triple wall at 23500 is within 1% of spot → add risk flag.

Actually re-reading: "Any stacked TRIPLE_WALL within 1% of spot" - 23500 is 139 pts below spot 23639, that's 0.59% - yes within 1%. Add risk flag.

24000 is 361 pts above = 1.53% - outside 1%, but still a major resistance.

6. BREADTH:
Very weak. Mar 9 was catastrophic (12% advance ratio, 76 stocks down 4%+). Mar 12 showing 39% advance but 303 declines vs 195 advances. %> EMA10 only 26.6%, %> EMA20 23.2%. DETERIORATING trend. 52W lows climbing (53 on Mar 12).

7. RISK FLAGS:
- OI divergence: weekly bearish + monthly bullish
- cap=AVOID_ALL (score=-1.40)
- Triple wall at 23500 within 1% of spot (139 pts, 0.59%)
- Quad-wall at 24000 (all 4 expiries) - extremely strong resistance 361 pts away
- FII consecutive selling streak 2 days (not ≥3, so don't add this specific flag)
- PCR weekly 0.697 < 0.70 (add: PCR < 0.7 weekly - aggressive CE writing)
- VIX spike_alert=false, but VIX at 98.8th percentile is extreme
- Breadth deteriorating (39% advance on Mar 12, 53 52W lows)

FII streak is 2 days, not ≥3, so skip that specific flag.

PCR weekly = 0.697 < 0.7 → add flag.

Now building the JSON:

For avoid[] - I need to include all Stage 2 candidates (both gate-cleared and gate-failed from avoided list) under avoid. The gate-failed ones from Stage 2's "avoided" list already have their reasons. The gate-cleared candidates get the OI cap reason.

Actually the instruction says: "move all Stage 2 candidates to avoid[] with reason format: 'OI cap AVOID_ALL (score=X): <cap_reason>'"

So all 19 gate-cleared candidates + the 7 already-avoided ones all go to avoid[]. For the 7 already avoided, should I use their gate reasons or the OI cap reason? I think the OI cap reason supersedes for all. But for clarity, the gate-failed ones could retain their gate reasons since they were already rejected at Stage 2. Let me use OI cap reason for the gate-cleared ones and retain gate reasons for the gate-failed (already avoided) ones - or I can add both.

I'll use OI cap for the gate-cleared ones (they would have been watchlist but get moved). The gate-failed ones keep their original reasons since they never made it through.

Let me structure the avoid[] array. There are 19 gate-cleared candidates and 7 pre-avoided = 26 total in avoid.

For action_plan - must end with re-entry conditions: weekly PCR > 0.80 AND VIX < 18 AND FO quadrant not SHORT_BUILDUP.

Let me now write the JSON carefully.

For oi_analysis:
- week_bias: BEARISH (PCR 0.697)
- nxtweek_bias: BEARISH (PCR 0.543)
- month_bias: NEUTRAL (PCR 1.078)
- nxtmonth_bias: NEUTRAL (PCR 0.973)
- weekly_max_pain: 23900
- monthly_max_pain: 25000
- nearest_resistance: 24000
- nearest_support: 23500

For delivery_signal: TATAPOWER (spike_ratio 3.9), COALINDIA (2.2), CGCL (5.8), JSWENERGY (5.0), JINDALSAW (4.18) - but since all go to avoid, top_picks should reflect the highest spike ratios even noting they're in avoid context. Actually delivery_signal.top_picks is about smart money delivery signals - I can still list them as data points even if they're in avoid.

Smart money = DISTRIBUTING given FII selling.

Watchlist must be EMPTY.

Let me write the full JSON now.