Technology — 10-Day Rolling Flow

Sector Breadth Snapshot — 05/01 Close

The headline split shows 69/103 bullish (67.0%) at the SECTOR QUICK SUMMARY level — comparable to the 04/22 cold-start (80/103 = 77.7%). But the composition is bifurcated underneath: the bear column now has the Layer 1 NVDA flip (-0.56% on $1.73B volume) and the entire semi equipment subgroup (KLAC, AMAT, LRCX, ASML), while the bull side has the storage breakout (SNDK +8.25%, STX +7.91%) and the software vertical re-bid (ORCL +6.47%, NOW +3.23%, TEAM +29.58% post-print). Rule 9 fragility (3-of-4 flags + 2 amplifications) caps any single-name conviction at Tier 2 regardless of how clean the constituent flow looks.

NVDA Layer 1 bear signal is the key tell of the session. The chunk's 15-day flow history shows 6 of 10 days flagged divergent (label vs price), and the 04/30 SELL print at -4.6% on $3.02B volume is the largest single-session distribution in the window. The 05/01 +$758M label on -0.56% price is a textbook Rule 5/10 spread compression artifact in a SLOW tape — the label says buying but the price-volume signature says distribution continued. This is the #1 mega-cap divergence to anchor the synthesis around.

Top 15 Tech Tickers by 05/01 Net Darkpool Flow (Absolute)

Read: AAPL ($1.35B) and MSFT ($1.33B) anchor mega-cap accumulation, validating the regime_snapshot.md "MAGS BIFURCATED" call — both extended above $700+B market cap on AAPL post-print bid. NVDA's $758M label print on -0.56% price = Rule 5/10 distribution-disguised-as-accumulation (see Rule 5/10 Watch section). GOOGL +$561M ACCU (EME) confirms continued institutional conviction in the cloud capex thesis. GOOG -$482M is the GOOG/GOOGL spread split — treat the combined vote as net +$78M with the GOOGL bid winning. ORCL +$270M with 6.47% price = clean accumulation post-print. STX/SNDK/TEAM are the day's single-name outliers off earnings catalysts (TEAM +29.58% on print, SNDK +8.25% storage breakout, STX +7.91% storage rotation).

Tape/label discipline reminder: Most 05/01 L1 verdicts in the chunk's per-ticker history register as UNKNOWN because the recon script ran before the final-price reconciliation pass. The L1 column in this table reflects the SECTOR QUICK SUMMARY verdict (price-validated), not the raw label aggregation. The Net $M column is the raw label aggregation. When the two disagree, the L1 verdict wins per Rule 5 (Price Action Is The Signal) — e.g., MU +4.84% with -$226M label = BUY confirmed by price, label artifact discounted.

Rolling Scorecard — 8-Day Window (04/22 → 05/01)

Each cell shows the price-validated Layer 1 verdict for that session: BUY (price up + flow up), SELL (price down + flow down), LEAN (mixed-signal soft conviction), NEUT (no directional read or chunk reported UNKNOWN). The 05/01 column uses the SECTOR QUICK SUMMARY verdict; prior columns use the per-ticker 15-day flow history L1 column.

Window arc: Most mega-caps run a continuous BUY streak through 04/22-04/30, then bifurcate on 05/01 along earnings/positioning lines. NVDA, META, NFLX (the bear flips), and ORCL/SNOW (the BUY flips out of prior-week distribution sequences) are the rotation tells. The semis (ASML, AMAT, KLAC, LRCX, ADI, MRVL, QCOM) carry SELL prints across multiple sessions in the window — consistent with the regime_snapshot SMH = NEUTRAL classification (TSM/MU bull offset by AMD/TXN bear). Storage (SNDK, STX) and software (ORCL, NOW, TEAM, CRWD, INTU) flipped to BUY late in the window on earnings catalysts. The cybersecurity vertical (CRWD, ZS, NET, PANW, S, FTNT) is unanimously BUY for the 05/01 close.

Rolling Net Directional Flow — 8-Day Per-Day $M (Label-Net)

Each cell shows the daily aggregate label net from the chunk's 15-day flow history (raw AtAsk minus AtBid before Layer 1 reclassification). Green = positive net, red = negative net. Per Rule 5/10, label net can disagree with the L1 verdict in fast tape conditions — cross-check against the Rolling Scorecard above for the price-validated read.

Read the label-net table conservatively. Large negative prints on UP-price days (and vice versa) are spread compression artifacts in fast tape, not genuine institutional intent. NVDA's row is the canonical example: the 6/10 divergence flag in the source chunk means roughly half of the daily label-net values are directionally wrong relative to the price-validated verdict in the Rolling Scorecard. Use this table for magnitude/scale context (whose flow is moving in $B vs $M), not for direction calls.

Tier Assignments — Post 05/01 Close (Rule 9 Caps Tier 2 Max)

Per the regime_snapshot fragility dashboard, 3 of 4 flags hold (CCR elevated on mega-cap bull names, 200DMA stretch ~520+ pts, multi-index quarterly stretch with 5 of 6 above QTD upper) plus 2 amplifications (rotation winner FLIP across yesterday's Tier 1 names; bank distribution at scale -$1.34B aggregate). Tier 1 conviction is therefore CAPPED at Tier 2 across the board. The names below earn Tier classification on flow + structure but cannot be sized as full conviction until fragility flags clear or convergence re-expands to +5-7 net.

Tier 2 (Cap from Rule 9, Otherwise Tier 1 Quality)

AAPL ($280.14, +3.24%): +$1.35B label net, ACCU (MOD) ladder, post-print bid extension. Largest single-day flow in the sector and the cleanest mega-cap setup — clean fundamentals + post-print absorption is the AAPL/GOOGL template the regime_snapshot calls out as the reaction archetype. Tier 1 quality but Rule 9 cap holds.

MSFT ($414.44, +1.63%): +$1.33B label net, ACCU (STR) ladder. Largest active accumulation strength rating in the chunk and a sustained 8-of-8 BUY scorecard. Tier 1 quality but Rule 9 cap.

GOOGL ($385.69, +0.23%): +$561M ACCU (EME), continuing post-earnings institutional re-rate. Cloud capex thesis intact. The +$561M with +0.23% price = 80%+ Ask conviction, not a label artifact. GOOG -$482M is the spread split — combined GOOG/GOOGL treats the bid as net positive.

AMZN ($268.26, +1.21%): +$23M label net is small in absolute terms but ACCU (EME) ladder + 8-of-8 BUY scorecard means structural participation. Trade is the post-print absorption, not the day's flow magnitude.

TSM ($397.67, +0.41%): +$578M with DIST (MOD) ladder = Rule 5/10 case where the price-validated verdict (BULLISH) trumps the labeled DISTRIBUTION ladder. Confirms the regime_snapshot SMH = NEUTRAL but TSM = bull split.

ORCL ($171.83, +6.47%): +$270M with 6.47% price move on FAST tape. DIST (MOD) ladder is the prior-week distribution sequence being unwound by this print — the prior 1B/6S scorecard pattern flipped to BUY today. Software vertical re-bid leader.

Tier 3 (Hold but Watch)

AVGO ($421.28, +0.92%): -$62M label net but BULLISH SECTOR QUICK SUMMARY verdict and the broader semis re-rate continues. Holding, not adding.

AMD ($360.54, +1.34%): ACCU (MOD) ladder holds but regime_snapshot specifically flags AMD as a SMH bear constituent. Holding only as long as price stays above the 04/24 breakout level.

SNDK ($1187.00, +8.25%) / STX ($726.93, +7.91%): Storage breakout pair on the 05/01 print. SNDK ACCU (EME). Late-cycle rotation candidates; allow the second day of follow-through before adding.

CRWD ($455.64, +1.34%) / PANW ($181.08, +0.92%) / FTNT ($86.29, +1.42%) / ZS ($139.81, +6.99%) / NET ($217.50, +6.11%): Cybersecurity vertical universally BUY across the 8-day window. ZS and NET are the +6% movers worth flagging individually. The vertical's relative strength inside a fragile MAGS holds even when mega-caps wobble.

TEAM ($88.88, +29.58%): Post-print spike. Treat the first day as event-driven; second-day reaction is the directional read.

Tier-Bear / FADE (Rule 9 Bear-Side Cap Also Applies)

NVDA ($198.45, -0.56%): SECTOR QUICK SUMMARY = BEARISH. The 6/10 divergence flag in the chunk's 15-day flow history is the highest in the sector. The 04/30 SELL print at -4.6% on $3.02B volume + the 05/01 +$758M label on -0.56% price = sustained Layer 1 distribution disguised by spread compression artifacts. This is the single most important mega-cap divergence to anchor the synthesis around. Hedge candidate, not long.

META ($608.74, -0.52%): +$113M label on negative price = Rule 5/10 label artifact in SLOW tape. ACCU (MOD) ladder is stale relative to today's price action. Mild bear, not full FADE.

NFLX ($92.06, -1.41%): ACCU (MOD) ladder reads stale — price action negative the entire window with multiple SELL prints. The streaming subsector decays continue.

ASML / AMAT / KLAC / LRCX / ADI / MRVL / QCOM: Semi-equipment + foundry-exposed names with multiple SELL prints across the window. The NEUTRAL SMH classification is built on the bull constituents (TSM, MU, LRCX in some sessions) offsetting these. Position-wise, these are short-the-rip names, not long.

ROK / SMCI / WDC / SYM / FISV / SERV / VICR / CAMT / SPOT: Smaller-cap or industrial-tech bears. Full FADE list — multiple SELL prints in the window with no flow conviction reversal.

Rule 5/10 Watch — 8-Day Window Label-vs-Price Divergence

Per Rule 5 (Price Action Is The Signal) and Rule 10 (Labels Lie — Price Doesn't), label flow on FAST or compressed tape is unreliable. The tickers below show the largest label/price disagreement in the 8-day window where the SECTOR QUICK SUMMARY (price-validated) verdict diverges from the raw label-net direction.

Pattern across the window: The most common label/price disagreement is FAST-tape negative-label on positive-price — in 5 of the 6 watch entries above. This is the spread compression signature on a strong directional move where label aggregation under-reads the bid. The single most consequential disagreement is the OPPOSITE pattern: NVDA SLOW-tape positive-label on negative-price, where the label is mislabeling sustained Layer 1 distribution as accumulation. The chunk's per-ticker divergence count (6/10) flagged this explicitly.

Synthesis — Technology Phase 3B Day 14

The 8-day arc traces a clean three-phase structure. Phase 1 (04/22-04/24): the post-04/20-rotation re-crowding into mega-caps with the historic SMH +38.74% 16-day streak as the structural anchor. The cold-start 04/22 report flagged 80/103 bullish on the day after the GOOGL/INTC AMC overnight binary. Phase 2 (04/27-04/29): consolidation with mega-cap mixed prints (NVDA SELL twice, MU/AVGO sustaining BUY) while the rotation winners (financials, energy, defensives) absorbed selling pressure elsewhere. Phase 3 (04/30-05/01): the rotation flip arrives. NVDA's -4.6% on $3.02B volume on 04/30 was the first major distribution confirmation, and the 05/01 SHOOTING STAR at the SPY $722 0DTE call wall (close $720.65) is the regime-snapshot's "provisional first distribution day inside the rotation regime."

What the technology flow says: the sector still prints +$4.0B XLK net on 05/01 (#1 sector by aggregate flow), driven by AAPL +$1.35B, MSFT +$1.33B, GOOGL/AMZN/TSM/ORCL accumulation. The mega-cap re-crowding has not broken — only NVDA has flipped to Layer 1 bear. Storage and software verticals (SNDK, STX, ORCL, NOW, TEAM, CRWD, ZS, NET) are the 05/01 outperformers on earnings catalysts. Cybersecurity is unanimously BUY across all 103 names in that vertical.

What the macro says: the regime_snapshot has dropped convergence from +7 net bull on 04/30 to +1 net bull on 05/01 post-VIX-correction. Three fragility flags hold + two amplifications: rotation winners flipped (TGT/TJX/ANET/NEE/O/BX, JPM/MS/C/WFC all distributed today), and bank distribution at scale (-$1.34B aggregate financials) is the cross-asset bear input. The bond bear is forming (AGG -$428.7M, USFR distribution, TLT at lower daily zone). HYG at $80.06 lower bound with a Silva HYG-vs-SPX divergence building. SPX/SPY/QQQ/NDX/IWM are 5 of 6 above QTD upper EM — multi-index quarterly stretch is at risk of contraction.

The convergence/divergence read: technology flow says the mega-cap re-crowding holds. Macro says the regime is one fragility-flag ignition away from rotation reversion. The two views are reconcilable only if you read the SHOOTING STAR at $722 as a tactical local top inside an intact regime, not the structural top. The regime_snapshot's working bias is "PULLBACK CONSOLIDATION through 5/14 OpEx, then 5/14-5/15 magnet attempt at recovery." Position with that — trim Tier 2 sizing on the 04/30-05/01 distribution candidates (NVDA short hedge, META mild bear, NFLX FADE, semi-equipment bears), keep core mega-cap exposure (AAPL, MSFT, GOOGL/AMZN, TSM/ORCL), avoid adding new Tier 1 conviction until either fragility flags clear or convergence re-expands to +5 net.

The highest-conviction position for the next 5-10 sessions: a Tier 2 long-only book centered on AAPL/MSFT/GOOGL/AMZN with cybersecurity (CRWD, ZS, NET) as the relative-strength vertical hedge inside the sector, plus the storage breakout (SNDK, STX) on 2-day follow-through, paired with NVDA-specific short-side hedges (puts or short-dated put spreads) given the 6/10 divergence rate. If MAGS bifurcation widens further (AAPL/MSFT/GOOGL holding bid while NVDA/META re-distribute), the sector trade migrates from "mega-cap basket long" to "mega-cap pairs trade long-bid/short-distribution." The 5/14 OpEx is the binary — mega-cap basket holds the $720 SPY zone through OpEx = regime intact; loses $720 with conviction = rotation flip thesis confirmed and Tier 2 mega-cap sizing comes off.