Swing Trade
What to swing trade this week: every day at 15:10 HKT this page re-reads the last US session — biggest pump & dump per subcategory, why it happened, what the second-order effect is, and how fast to act.
Sector heat map
The playbook
#1 Deleveraging cascade — basket-wide forced selling; resilient names cracking, big caps more volatile than small — a whale fight. Wait for dust to settle.
#2 Overheat cooling — hot category taking profit; volume still high, sentiment intact, no bad news — knife-catch candidate.
#3 Fading — volume drying up, sentiment flipping, attention gone — careful, knives here cut.
#4 Earnings / econ prints — post-report repricing — always ask what the second-order effect is before entering.
#5 Policy — bans/tariffs/subsidies — first move often overshoots; ask what policy CAN'T fully do.
#6 No-reason dump — no news, no filings, reddit confused — investigate; conviction longs live here.
Daily brief — 2026-10-01
Sept 30 US close: Nasdaq rose to finish Q3 and Wall Street closed in on a winning September after PCE inflation landed below expectations, with Treasury yields still elevated. DATA NOTE: this run executed at ~19:30 HKT (07:30 ET) rather than 15:10 HKT, so every US row (chg/vel/rvol) is an Oct 1 PREMARKET snapshot against the Sept 30 close - volumes are tiny and rvol/z are not valid. The daily-bars archive (massive_stock_data.db) is also stale at 2026-09-28, so r5d/r20d baselines are ~2 sessions old. Read US moves as indications and confirm at the open.
Act-fast list
Velocity 0–10 = how fast to act: move size vs the name's own volatility (z), volume confirmation (RVOL), raw magnitude. The faster and bigger the dump, the faster the work.
JBLspot Jabil
Optical
6.0 KNIFE-CATCH
-10.0% on the day · z 4.18 · RVOL 0.0 · 5d +5.7% · 20d -1.9%
Why: PREMARKET print (run was late, ~07:30 ET) - volume/rvol/z are not comparable to a full session. Down ~10% on the fiscal Q4 report: Q4 beat and a record FY27 revenue guide of ~$44.5B (AI data-centre demand 'accelerating'), but the stock sold off on profit-taking after running +5.7% over 5d into the print.
Cause: #4 Earnings reaction (class 4) - a beat-and-raise sold. The model is intact (guide is up, AI mix rising); what broke is the positioning.
Second-order: Peers FLEX/SANM/CLS (EMS/optical-adjacent) trade in sympathy; Serenity flagged JBL (bullish) on 9/30, so the dip is where tracked longs may add.
Play: Beat+raise sold is classic overheat-cooling; wait for the open to see if -10% holds on real volume before sizing - do not chase a premarket print. [1] [2]
ROG Rogers Corp
Semi
6.0 WATCH
+10.7% on the day · z 4.71 · RVOL 0.0 · 5d +6.2% · 20d +6.2%
Why: PREMARKET print (run was late, ~07:30 ET) - volume/rvol/z are not comparable to a full session. +10.7% premarket on near-zero volume; searches found no dated catalyst. Semi/EV materials name 9.5% under its high with the Asia semi-equipment tape (Advantest +9.8%, Lasertec +11.1%) ripping the same night.
Cause: #6 No visible company news (class 6, flagged). Most likely sector read-through from the Japan/US semi-cap bid; unconfirmed.
Second-order: Semi-equipment/test names (FORM +9.6%, ACMR +7.0%) move together - ROG may just be riding the basket.
Play: Flagged for review: no catalyst found and the print is premarket thin. Needs the open + volume to confirm.
AIB AIB (saas micro-cap)
SaaS
4.9 AVOID
-22.2% on the day · z 2.94 · RVOL 0.15 · 5d +15.8% · 20d +17.4%
Why: PREMARKET print (run was late, ~07:30 ET) - volume/rvol/z are not comparable to a full session. -22% premarket on ~164k shares, ~$100M cap, 79% off the high. No catalyst found (search results were the unrelated Irish bank AIB).
Cause: #6 No visible reason (class 6, flagged). Micro-cap, +15.8% over 5d before this: looks like a give-back of a spike.
Second-order: None identified; thin liquidity means slippage dominates.
Play: Illiquid micro-cap dump with no identified cause - deciding AVOID fast is the action. Flagged for manual review.
DMRC Digimarc
SaaS
6.0 WATCH
+17.1% on the day · z 4.84 · RVOL 0.0 · 5d -1.8% · 20d -30.4%
Why: PREMARKET print (run was late, ~07:30 ET) - volume/rvol/z are not comparable to a full session. +17% on a 3-share premarket print - effectively no price discovery. Last known news: gift-card rollout to ~600 locations slipped to Jan 2027; stock -30% over 20d, 67% off high.
Cause: #6 No visible catalyst (class 6, flagged); print too thin to read.
Second-order: Tends to squeeze on tiny volume; watch the open.
Play: Ignore the premarket gap until it trades real volume. [1]
ZETA Zeta Global
SaaS
4.7 WATCH
+9.0% on the day · z 2.94 · RVOL 0.0 · 5d -2.6% · 20d +2.3%
Why: PREMARKET print (run was late, ~07:30 ET) - volume/rvol/z are not comparable to a full session. +9.0% premarket, 3.8% off the 52w high, after -2.6% over 5d. Search found only older AI-momentum/earnings pieces, no 10/1 catalyst.
Cause: #6 No visible same-day catalyst (class 6, flagged). Reads as saas/AI-marketing momentum breaking out toward highs.
Second-order: Other saas leaders (FROG, AI, VERX, ATEN +5-7% premarket) move with it - a basket bid, not a one-name story.
Play: Breakout toward highs is not a knife; wait for a hold on the open rather than buying the gap.
FORMspot FormFactor
Memory
3.9 WATCH
+9.6% on the day · z 2.01 · RVOL 0.0 · 5d +9.6% · 20d +12.8%
Why: PREMARKET print (run was late, ~07:30 ET) - volume/rvol/z are not comparable to a full session. +9.6% premarket, 6.8% off high, +9.6% over 5d. Consistent with the semi-test/equipment rally in Japan (Advantest +9.8%, Lasertec +11.1%, Tokyo Electron +6.5%) - FormFactor is the US probe-card/test sibling.
Cause: #2 Basket-wide semi-cap bid (read-through from Asia), no company-specific news found.
Second-order: ACMR +7.0% and ROG +10.7% same basket; Advantest/Lasertec lead, FORM follows.
Play: Already extended (+9.6% over 5d); better entries on a pullback, and the Asia strength confirms the theme.
Catch-up signals
Rule: a basket pumped (or ≥2 leaders +5%) while these members lagged — potential catch-up longs. Always check WHY a laggard lagged before entering. perp = Bitget stock perp · spot = Bitget rTOKEN onchain spot.
Optical — catch-up long candidates
basket -1.5% · 79% down · 5d +4.1% · 20d +2.6%
Leaders: MXLspot -2.6% · AXTIperpspot -1.0% · SANM -3.5%
Lagging: COHRperpspot -1.5% 5d -6.8% · 20d +2.9% · AAOIperpspot -1.4% 5d -3.6% · 20d -17.0% · NOKspot -2.1% 5d -2.7% · 20d +2.6%
basket pumped, these lagged — potential catch-up long; verify each laggard isn't lagging for a reason | VETTING: late premarket run - laggards NOT individually searched; treat each as unvetted (include-and-flag). 'No visible reason' is unconfirmed.
Robot — catch-up long candidates
basket -0.0% · 50% down · 5d +1.0% · 20d -2.8%
Leaders: OUSTperpspot +0.8% · ALNT -0.5% · INVZ -3.7%
Lagging: CCXI -5.8% 5d -3.6% · 20d -14.9% · BOTperpspot -0.6% 5d -3.1% · 20d -1.5% · BYDperp -0.8% 5d -2.4% · 20d -13.8%
basket pumped, these lagged — potential catch-up long; verify each laggard isn't lagging for a reason | VETTING: late premarket run - laggards NOT individually searched; treat each as unvetted (include-and-flag). 'No visible reason' is unconfirmed.
Semi — catch-up long candidates
basket -0.0% · 52% down · 5d +5.6% · 20d +2.4%
Leaders: ACMRspot +7.0% · ROG +10.7%
Lagging: ERIC -1.8% 5d -6.3% · 20d -6.1% · BMI -1.9% 5d -2.5% · 20d -1.1% · DAKT -0.1% 5d -2.2% · 20d -12.7%
basket pumped, these lagged — potential catch-up long; verify each laggard isn't lagging for a reason | VETTING: late premarket run - laggards NOT individually searched; treat each as unvetted (include-and-flag). 'No visible reason' is unconfirmed.
Quantum — catch-up long candidates
basket -0.2% · 80% down · 5d +3.5% · 20d +0.2%
Leaders: ARQQperpspot -0.2% · IONQperpspot -0.1%
Lagging: XNDUperp -4.0% 5d -32.9% · 20d -52.0% · BAH -0.4% 5d -5.3% · 20d -7.5% · QNT -1.4% 5d -4.0% · 20d -11.6%
basket pumped, these lagged — potential catch-up long; verify each laggard isn't lagging for a reason | VETTING: late premarket run - laggards NOT individually searched; treat each as unvetted (include-and-flag). 'No visible reason' is unconfirmed. XNDU is -33% over 5d / -52% over 20d / 89% off high: a falling knife, not a catch-up.
SaaS — catch-up long candidates
basket +0.7% · 33% down · 5d -1.0% · 20d -5.2%
Leaders: ATEN +5.0% · VERX +5.7% · AI +6.3% · FROG +7.4%
Lagging: KPLT -3.7% 5d -47.0% · 20d -15.0% · KDK -9.9% 5d -25.9% · 20d -37.8% · PPHC -0.6% 5d -15.6% · 20d +0.8%
basket pumped, these lagged — potential catch-up long; verify each laggard isn't lagging for a reason | VETTING: late premarket run - laggards NOT individually searched; treat each as unvetted (include-and-flag). 'No visible reason' is unconfirmed.
Space — catch-up long candidates
basket -0.3% · 62% down · 5d +2.7% · 20d -7.0%
Leaders: SATLW -19.7% · SATLperp -1.2% · FLYperp +1.3% · VIAV +0.2%
Lagging: VOYG -1.6% 5d -12.6% · 20d -19.3% · RCI -0.2% 5d -9.0% · 20d -12.7% · YORW +0.4% 5d -5.2% · 20d -8.4%
basket pumped, these lagged — potential catch-up long; verify each laggard isn't lagging for a reason | VETTING: late premarket run - laggards NOT individually searched; treat each as unvetted (include-and-flag). 'No visible reason' is unconfirmed.
Week movers
Biggest 5-day moves — the day gets the focus, the week is the context. Steady grinds don't show up in 1-day velocity.
| Ticker | Category | 5d | Day | 20d | off 52w high |
|---|
| SATLW | Space | +48.0% | -19.7% | -21.2% | -90.4% |
| KDKRW | Robot | -34.0% | -11.4% | -50.4% | -88.7% |
| XNDUperp | Quantum | -32.9% | -4.0% | -52.0% | -89.2% |
| INTWperp | Semi | +26.0% | +7.1% | +65.4% | -40.6% |
| KDK | SaaS | -25.9% | -9.9% | -37.8% | -83.1% |
| CIRC | SaaS | -25.1% | +11.8% | -47.5% | -89.7% |
| METUperp | Mag caps | +25.1% | -3.6% | +82.8% | -29.7% |
| BWETperpspot | Wars | -24.6% | +8.6% | +53.0% | -7.9% |
| RMBSspot | Memory | +20.1% | +3.5% | +15.8% | -38.7% |
| VERI | SaaS | +19.1% | -9.8% | +35.6% | -87.4% |
Asia session — Bitget perp underlyings
The 23 HK/CN/KR/JP names behind Bitget stock perps. Their cash sessions close before this 15:10 HKT brief, so this block is here every day — including Monday, when there is no completed US session. Underlying as of 2026-10-01. Hong Kong closed China A closed Korea closed Japan closed
| Region | Basket | Top pump | Top dump | n |
|---|
| Hong Kong | +0.2% 46% down · 5d -1.7% · 20d -2.8% | ZHIPUperp +4.3% 1.7 | CSOPSS2LHKDperp -2.8% 1.1 | 13 |
| China A | -2.2% 100% down · 5d -11.8% · 20d -7.9% | — — — | GIGADEVICEperp -3.7% 2.0 | 4 |
| Korea | +2.8% 22% down · 5d +0.5% · 20d +4.1% | HANMIperp +4.7% 2.3 | NAVERperp -1.3% 1.2 | 9 |
| Japan | +5.5% 0% down · 5d +9.0% · 20d +6.7% | LASERTECperp +11.1% 5.5 | — — — | 7 |
Japan led by a wide margin (median +5.6%, 0% down, +9.0% over 5d): Lasertec +11.1%, Advantest +9.8%, Tokyo Electron +6.5%, Kioxia +5.6%, SoftBank +4.2%, Nikkei +3.3%. Korea firm (median +2.8%, 22% down; Hanmi +4.7%, Doosan Robotics +4.7%, SK Hynix +3.2%). China and Hong Kong did not trade (Golden Week / HK holiday), so CN (-2.2%, 100% down, -11.8% over 5d) and HK (+0.2%) show the 9/30 close. The Asia semi-equipment surge CONFIRMS the US premarket read (FORM +9.6%, ACMR +7.0%, ROG +10.7%) for semis/test, while CN memory (GigaDevice -3.7%, -11% over 5d) still CONTRADICTS the memory bid - it is not yet pan-Asia. DATA FLAG: Tokyo Electron shows r20d -77.7%, likely a stock-split artifact in the OHLCV history - ignore its 20d.
| Base | Name | Exch | Session | Vel | 5d | 20d | off 52w high |
|---|
| LASERTECperp | Lasertec Corporation | TSE | +11.1% | 5.5 | +11.7% | +35.9% | -20.5% |
| ADVANTESTperp | Advantest Corporation | TSE | +9.8% | 5.4 | +10.6% | +11.6% | -0.4% |
| TOKYOELperp | Tokyo Electron Limited | TSE | +6.5% | 3.2 | +11.3% | -77.7% | -84.0% |
| KIOXIAperp | KIOXIA HOLDINGS CORPORATION | TSE | +5.5% | 1.8 | +2.7% | -61.8% | -82.4% |
| HANMIperp | Hanmi Semiconductor | KRX | +4.7% | 2.3 | +10.3% | +24.9% | -34.4% |
| DOOSBOTperp | Doosan Robotics Inc. | KRX | +4.7% | 3.2 | +1.4% | +4.1% | -56.9% |
| ZHIPUperp | Knowledge Atlas Tech Joint | HKEX | +4.3% | 1.7 | +0.4% | -42.6% | -73.4% |
| SOFTBANKperp | SoftBank Group | TSE | +4.2% | 1.6 | +9.0% | +28.9% | -22.4% |
| GIGADEVICEperp | GigaDevice Semiconductor Inc. | SSE | -3.7% | 2.0 | -11.2% | -9.9% | -57.9% |
| MINIMAXperp | MiniMax Group Inc | HKEX | +3.6% | 1.3 | -10.3% | -27.0% | -79.8% |
| JP225perp | Nikkei 225 | INDEX | +3.3% | 2.9 | +3.9% | +4.0% | -4.7% |
| SKHYNIXperp | SK hynix Inc. | KRX | +3.2% | 1.5 | -1.6% | +8.3% | -37.2% |
| SAMSUNGEMperp | Samsung Electro-Mechanics | KRX | +3.2% | 1.9 | +3.6% | +9.2% | -31.2% |
| QNTSTOCKperp | QuantumCTek Co., Ltd. | SSE | -3.1% | 1.7 | -15.3% | -13.2% | -56.4% |
| SAMSUNGperp | Samsung Electronics Co., Ltd. | KRX | +2.8% | 1.5 | -3.3% | +5.8% | -23.9% |
| CSOPSS2LHKDperp | CSOP Samsung Electronics Daily (2x) Leveraged (HKD) | HKEX | -2.8% | 1.1 | -5.0% | +15.3% | -67.0% |
| LENOVOHKDperp | Lenovo Group (HKD) | HKEX | -2.6% | 1.1 | -4.1% | +12.9% | -8.1% |
| KUAISHOUperp | Kuaishou Technology | HKEX | +2.3% | 1.7 | +1.4% | -8.0% | -32.0% |
| SUMIELECperp | Sumitomo Electric Industries, Ltd. | TSE | +2.2% | 1.6 | +7.2% | +6.7% | -32.1% |
| MEITUANperp | Meituan | HKEX | +1.8% | 1.6 | -0.8% | -8.9% | -32.5% |
| CXMTperp | CXMT | SSE | -1.4% | 0.7 | -6.5% | -3.0% | -11.3% |
| NAVERperp | NAVER | KRX | -1.3% | 1.2 | -1.8% | -11.1% | -33.4% |
| POPMARTperp | Pop Mart International Group Limited | HKEX | -1.3% | 1.7 | +1.0% | -1.0% | -47.1% |
| HYUNDAIperp | Hyundai Motor Company | KRX | +1.2% | 0.9 | -1.3% | -12.9% | -53.5% |
| SMICperp | Semiconductor Manufacturing International Corporation | HKEX | -0.7% | 0.7 | -4.9% | -11.1% | -33.4% |
| DOOSENERperp | Doosan Enerbility Co., Ltd. | KRX | +0.6% | 0.3 | +0.5% | +0.9% | -39.8% |
| ZHONGJIperp | Zhongji Innolight | SZSE | -0.6% | 0.2 | -12.4% | -5.9% | -41.5% |
| LGELECTRONICSperp | LG Electronics | KRX | -0.5% | 0.3 | +1.2% | +2.4% | -46.1% |
| HSIperp | Hang Seng Index | INDEX | +0.4% | 0.7 | -0.6% | -2.8% | -12.0% |
| CSOPSK2LHKDperp | CSOP SK hynix Daily (2x) Leveraged (HKD) | HKEX | +0.2% | 0.1 | -5.7% | +16.5% | -78.6% |
| TENCENTperp | Tencent Holdings Limited | HKEX | -0.2% | 0.4 | -1.7% | -1.6% | -36.4% |
| XIAOMIperp | Xiaomi Corporation | HKEX | +0.2% | 0.1 | -5.0% | -10.1% | -57.5% |
| NETEASEperp | NetEase, Inc. | HKEX | -0.1% | 0.1 | +3.2% | +3.1% | -8.7% |
Perp vs underlying — gap watch
Every Asia cash market is shut right now, so every price in the perp column is an opinion with no cash market to check it against — that is exactly when the basis is information. Underlying is USD-normalised. Rows at |basis| ≥ 3% are highlighted. Stale hours = time since the underlying's last settled daily bar.
Real |basis|>=3%: ZHONGJI +11.4% and GIGADEVICE +8.5% - but NOT a normal gap read. Oct 1 starts China's Golden Week, so mainland A-shares are shut until ~Oct 8 and the underlyings are stale 35.5h (last bar 9/30). The perps barely moved (ZHONGJI -0.5%, GIGADEVICE -0.2% on 24h), so they are drifting, not gapping; a +8-11% reopen gap would need a week of buying with no price discovery - treat as a stale-underlying artifact unless the perp starts to trend. The HKD-quoted twins (CSOPSK2LHKD +707% and siblings) are the known quote-currency artifact (HKD perp vs USD-normalised underlying, ratio ~7.8): true basis ~0, still flagged for the owner of swing_movers.basis_watch. HK also holds an Oct 1 holiday, so HK rows are the 9/30 close. JP/KR are live and tight: Advantest -0.3%, Lasertec -0.3%, Tokyo Electron -0.6%, Kioxia -0.4%, Hanmi -1.0% - perps priced the +4% to +10% cash rally (24h perp moves match), no gap edge. Funding: only HANMI is non-trivial at -0.10% (shorts paying, perp +4.5%). No name moved >3% on a 24h perp basis with cash shut all along.
| Base | Name | Perp | Underlying $ | Basis | Perp 24h | Funding | Stale | Cash |
|---|
| CSOPSK2LHKDperp | CSOP SK hynix Daily (2x) Leveraged (HKD) | 41.265 | 5.1104 | +707.47% | +3.26% | 0.0% | 35.5h | shut |
| CSOPSS2LHKDperp | CSOP Samsung Electronics Daily (2x) Leveraged (HKD) | 80.14 | 10.0271 | +699.23% | +2.59% | 0.0% | 35.5h | shut |
| LENOVOHKDperp | Lenovo Group (HKD) | 34.557 | 4.4019 | +685.05% | -0.08% | 0.0% | 35.5h | shut |
| ZHONGJIperp | Zhongji Innolight | 134.4 | 120.6861 | +11.36% | -0.47% | 0.0% | 35.5h | shut |
| GIGADEVICEperp | GigaDevice Semiconductor Inc. | 57.32 | 52.8311 | +8.50% | -0.17% | 0.0% | 35.5h | shut |
| CXMTperp | CXMT | 8.411 | 8.1792 | +2.83% | +0.57% | 0.0% | 35.5h | shut |
| SAMSUNGperp | Samsung Electronics Co., Ltd. | 199.85 | 202.3935 | -1.26% | +1.15% | 0.0% | 11.5h | shut |
| JP225perp | Nikkei 225 | 68270.0 | 68956.7188 | -1.00% | +2.10% | 0.0% | 11.5h | shut |
| HANMIperp | Hanmi Semiconductor | 195.0 | 196.8937 | -0.96% | +4.46% | -0.1039% | 11.5h | shut |
| SAMSUNGEMperp | Samsung Electro-Mechanics | 1135.3 | 1144.6967 | -0.82% | +1.57% | 0.0% | 11.5h | shut |
| SKHYNIXperp | SK hynix Inc. | 1333.89 | 1344.1569 | -0.76% | +1.39% | 0.0% | 11.5h | shut |
| TOKYOELperp | Tokyo Electron Limited | 79.06 | 79.5307 | -0.59% | +7.18% | 0.0% | 11.5h | shut |
| NAVERperp | NAVER | 140.86 | 140.2088 | +0.46% | -1.57% | 0.0% | 11.5h | shut |
| KIOXIAperp | KIOXIA HOLDINGS CORPORATION | 120.25 | 120.75 | -0.41% | +3.97% | 0.0% | 11.5h | shut |
| MEITUANperp | Meituan | 9.169 | 9.1376 | +0.34% | +0.50% | 0.0% | 35.5h | shut |
All categories — biggest pump & dump
| Category | Basket | Top pump | Top dump | Read |
|---|
| Drone | -1.6% 90% down · 20d -14.6% | RCAT +0.1% 0.1 | AIRO -2.6% 1.4 | |
| Optical | -1.5% 79% down · 20d +2.6% | LPTHspot +2.9% 1.0 | JBLspot -10.0% 6.0 | JBL -10% premarket on a guide-beat sell-off drags the basket (79% down); MXL/AXTI/SANM ran +13-16% over 5d. |
| Mag caps | +1.0% 18% down · 20d +6.6% | AAPUperp +2.1% 1.2 | METUperp -3.6% 1.3 | |
| Defense | -0.7% 68% down · 20d -10.9% | BAER +9.6% 5.1 | AIR -5.0% 2.9 | |
| SaaS | +0.7% 33% down · 20d -5.2% | DMRC +17.1% 6.0 | AIB -22.2% 4.9 | Broad premarket bid (ZETA, FROG, AI, VERX, ATEN +5-9%) alongside micro-cap carnage (AIB -22%, KDK -10%, VERI -10%). |
| Prime | -0.7% 100% down · 20d -12.4% | — — — | NOCperpspot -4.2% 4.2 | |
| Neocloud | -0.6% 64% down · 20d +1.5% | BIRD +11.5% 4.3 | AGPUperp -9.4% 3.5 | |
| Space | -0.3% 62% down · 20d -7.0% | PHI +5.0% 3.3 | SATLW -19.7% 3.1 | SATLW warrant -19.7% after +48% 5d - warrants whip; not a read on the basket. |
| Components | +0.2% 36% down · 20d -13.1% | MPTI +2.9% 4.1 | IPX -8.1% 3.3 | |
| Quantum | -0.2% 80% down · 20d +0.2% | QBTSperpspot +0.8% 0.4 | XNDUperp -4.0% 1.5 | 80% of the basket down premarket despite +3.5% 5d median; XNDU is a knife (-33% 5d). |
| China Ai | -0.2% 67% down · 20d -9.7% | KWEBperpspot +0.6% 0.5 | KSTRperp -2.7% 1.8 | |
| Wars | +0.2% 45% down · 20d -4.3% | BWETperpspot +8.6% 2.7 | BATL -2.8% 1.5 | |
| Ai Model | +0.1% 50% down · 20d -17.1% | DXYZperp +0.7% 0.4 | VCXspot -0.4% 0.2 | |
| Memory | +0.1% 45% down · 20d +11.2% | FORMspot +9.6% 3.9 | RAMperp -3.1% 1.1 | |
| Semi | -0.0% 52% down · 20d +2.4% | ROG +10.7% 6.0 | CBRSperpspot -8.9% 3.8 | Semi-cap/test leading (ROG +10.7%, ACMR +7.0%); confirmed by Japan equipment names overnight. |
| Robot | -0.0% 50% down · 20d -2.8% | HSAIspot +2.8% 1.2 | KIDZW -19.0% 3.1 | |
| Rare Earth | +0.0% 47% down · 20d -12.3% | MPperpspot +4.1% 2.3 | SQM -1.8% 1.1 | |
| Compute | — % down · 20d — | OSSperp +0.9% 0.4 | — — — | |
| Directed Energy | — % down · 20d — | — — — | LASRspot -0.2% 0.1 | |
| Manufacturing | — % down · 20d — | VELO +0.8% 0.3 | — — — | |
🐋 = large caps dumping harder than small caps (inverted volatility order — possible forced selling; don't stand between whales).
52-week deals in tracked universes
13F consensus / Serenity / Citrini / Burry names at or near 52-week lows — or trading against a tracked short.
| Ticker | Setup | Day | 52-week | Tracked by | Note |
|---|
| BYDperp | near 52w low | -0.8% | 0.0% above 52w low · -27.0% off high | serenity·long | |
| APPperpspot | near 52w low | -5.0% | 0.0% above 52w low · -61.0% off high | 13f·held | |
| UAMYspot | near 52w low | -0.2% | 0.2% above 52w low · -78.8% off high | serenity·long | |
| EFX | near 52w low | -2.5% | 0.2% above 52w low · -46.3% off high | citrini·long | |
| PEPperpspot | near 52w low | -1.5% | 0.2% above 52w low · -23.9% off high | 13f·held | |
| FLUT | near 52w low | +0.1% | 0.3% above 52w low · -73.8% off high | burry·long | |
| DKNGperp | near 52w low | -3.0% | 0.3% above 52w low · -55.6% off high | 13f·held | |
| ZTS | near 52w low | -0.7% | 0.4% above 52w low · -52.2% off high | 13f·held, burry·long | |
| KTOSperp | near 52w low | -0.8% | 0.5% above 52w low · -68.1% off high | citrini·long, serenity·long | |
| KRMN | near 52w low | -1.3% | 0.7% above 52w low · -72.6% off high | citrini·long | |
| FISV | near 52w low | -0.4% | 0.8% above 52w low · -65.4% off high | burry·long | |
| LULU | near 52w low | -0.7% | 0.9% above 52w low · -57.4% off high | burry·long, serenity·long | |
| CHTR | near 52w low | +0.2% | 0.9% above 52w low · -61.2% off high | 13f·held, citrini·short | |
| OPENspot | near 52w low | +0.0% | 0.9% above 52w low · -73.6% off high | 13f·held | |
| XLUperp | near 52w low | -0.7% | 1.1% above 52w low · -16.2% off high | serenity·long | |
| FICO | near 52w low | -4.1% | 1.1% above 52w low · -70.3% off high | 13f·held | |
| LHX | near 52w low | +0.1% | 1.2% above 52w low · -36.8% off high | citrini·long | |
| TMUSspot | near 52w low | +0.1% | 1.2% above 52w low · -30.9% off high | citrini·long | |
| RKT | near 52w low | -1.5% | 1.2% above 52w low · -52.8% off high | 13f·held | |
| XPEV | near 52w low | +0.9% | 1.3% above 52w low · -66.1% off high | serenity·long | |
| AMTM | near 52w low | -0.2% | 1.3% above 52w low · -52.5% off high | 13f·held | |
| NOCperpspot | near 52w low | -4.2% | 1.4% above 52w low · -37.0% off high | serenity·long | |
| SPGIspot | near 52w low | +0.5% | 1.5% above 52w low · -33.1% off high | 13f·held | |
| YUM | near 52w low | -1.0% | 1.5% above 52w low · -19.5% off high | 13f·held | |
| EOSE | near 52w low | -1.4% | 1.8% above 52w low · -84.9% off high | serenity·long | |
| CPRT | near 52w low | +0.5% | 2.0% above 52w low · -40.6% off high | 13f·held | |
| CMCSAperp | near 52w low | +0.7% | 2.3% above 52w low · -32.1% off high | 13f·held, citrini·short | |
| FUN | near 52w low | -1.4% | 2.5% above 52w low · -60.9% off high | 13f·held | |
| TTD | near 52w low | +1.0% | 2.7% above 52w low · -78.4% off high | serenity·long | |
| NKLR | near 52w low | +1.8% | 3.4% above 52w low · -84.5% off high | serenity·long | |
| ORLY | near 52w low | -0.8% | 3.4% above 52w low · -21.4% off high | 13f·held | |
| SGI | near 52w low | -0.2% | 3.4% above 52w low · -36.5% off high | 13f·held | |
| BSX | near 52w low | -0.5% | 3.4% above 52w low · -58.7% off high | 13f·held | |
| HIG | near 52w low | -2.8% | 3.5% above 52w low · -15.9% off high | citrini·short | |
| CPNGperpspot | near 52w low | +0.1% | 3.5% above 52w low · -57.9% off high | 13f·held | |
| JPST | near 52w low | -0.0% | 3.7% above 52w low · 0.3% off high | 13f·held | |
| ETOR | near 52w low | +0.1% | 3.9% above 52w low · -41.9% off high | serenity·long | |
| HONA | near 52w low | -0.3% | 3.9% above 52w low · -47.9% off high | citrini·long | |
| TREE | near 52w low | +0.3% | 4.0% above 52w low · -65.7% off high | citrini·short | |
| DHI | near 52w low | -2.0% | 4.3% above 52w low · -21.5% off high | 13f·held | |
Smart money today
Serenity (today's tweets)
- AAOIperpspot Serenity posted (High conviction), last_date 2026-09-30.
- NVDAperpspot Serenity posted (High conviction), last_date 2026-09-30.
- MUperpspot Serenity posted (High conviction), last_date 2026-09-30.
- GOOGLperpspot Serenity posted (High conviction), last_date 2026-09-30.
- MRVLperpspot Serenity posted (High conviction), last_date 2026-09-30.
- AMZNperpspot Serenity posted (Bullish), last_date 2026-09-30.
- JBLspot Serenity posted (Bullish), last_date 2026-09-30. JBL is -10% premarket on its FQ4 print: the tracked long gets a discount.
- CRDOperpspot Serenity posted (Bullish), last_date 2026-09-30.
- QCOMperpspot Serenity posted (Bullish), last_date 2026-09-30.
- ALABperpspot Serenity posted (Bullish), last_date 2026-09-30.
- CBRSperpspot Serenity posted (Bullish), last_date 2026-09-30.
- FNspot Serenity posted (Watching), last_date 2026-09-30.
- STMspot Serenity posted (Watching), last_date 2026-09-30.
- GNRCperpspot Serenity posted (Watching), last_date 2026-09-30.
- FLEXperpspot Serenity posted (Watching), last_date 2026-09-30.
Burry (this week)
- 9/28 Short Thoughts adds Part V.5 to the 'Heretic's Guide to AI's Stars' series (Part VI on AI financing in progress); 9/26 abridged forensic look at hyperscaler capital cycles (MSFT, AMZN, ORCL, META, GOOG) - bearish AI-capex framing against the semi/optical bid.
Archive
2026-09-30 — Tuesday Sept 29: yields are the story - the 10-year hit 5.29% (a 2007 high) and the 30-year a 2002 high, the Dow slid >100 points and CNN Fear & Greed
- FICO -26.5% 10.0 WAIT — Policy is still unfolding and the gap is 3.8% above the 52w low on record volume - let it base and see FHFA/FICO response before touching; a reflex bounce is possible but this is a
- SSTI +50.1% 9.0 AVOID — Capped upside, single-name deal risk - a done deal is not a swing trade. Only interesting as merger-arb if the spread to the offer is meaningful (verify the offer price).
- BAER -14.6% 10.0 AVOID — Sub-$1 with a $38M cap and no explanation: conviction-long territory only after a real reddit/filing check turns up a cause; not a swing entry from here.
- AIR -7.2% 8.3 WATCH — Dip on financing after a guide raise with sell-side support is watch-worthy; wait for the deal-terms check and a stabilising close above the day's low before sizing.
- HMH -7.8% 7.1 AVOID — Until the resale block clears or the stock proves it can absorb it, knives cut; no edge here versus cleaner energy/war names.
- XPEV -4.8% 6.1 WAIT — At a fresh 52w low with a downgrade the trend is your enemy; wait for a higher low or a reclaim of the breakdown level rather than catching it on a day where every China name is we
2026-09-29 — Monday Sept 28: Nasdaq and S&P 500 fell after Trump rebuffed an Iran peace deal and oil rose; a chip selloff deepened (ARM -9%, QCOM -6%, MRVL -5%) an
- MDBperp -18.5% 9.7 WAIT — Don't catch an -18% gap on a CEO exit before the interim plan and investor-day tone are known; look for a base or a named new CEO.
- BAperpspot -6.9% 9.2 WATCH — Analyst support but no catalyst to reverse the certification timeline; wait for a stabilising close.
- MX +14.8% 8.5 AVOID — Chasing a +15% extension on an unconfirmed catalyst; revisit on a pullback.
- ARMperpspot -8.7% 4.4 KNIFE-CATCH — Class-2 pullback in a strong tape is the setup; scale in small with a stop under the recent range, don't add if semis breadth stays >70% down.
- ALIT -10.4% 5.4 AVOID — Falling knife in a weak basket with no identified catalyst.
- BB +7.2% 5.3 WATCH — Relative-strength name; enter on a pullback rather than after the follow-through day.
2026-09-28 — MONDAY 28 Sep -- no completed US session; the US fields below (deep dives, signals, week movers, deals, basket rows) are Friday 25 Sep carried through
- SHMD +20.7% 8.9 AVOID — Real narrative, weak fundamentals -- a name to watch for a de-rate, not to hold through.
- KLXE +9.5% 8.9 WATCH — Constructive operationally but the rights offering adds float -- let the dilution digest before sizing up.
- KDK -12.6% 8.7 AVOID — Deteriorating momentum on a name with no revenue support -- stay out until launch-delay reports are confirmed or denied.
- KDKRW -21.4% 8.2 AVOID — Deteriorating momentum on a name with no revenue support -- stay out until launch-delay reports are confirmed or denied.
- AESI +13.5% 8.2 ENTER — Contracted, multi-year revenue visibility on a durable theme (AI data-center power) -- the cleanest catalyst of today's list.
- HQI +6.9% 6.8 WATCH — No news to hang a thesis on -- don't chase a no-catalyst pump in an already-overvalued micro-cap.
2026-09-27 — Neither US nor Asia cash markets traded this weekend; Friday's US close (S&P 500 +0.51%, Nasdaq +0.48%, Dow +0.93% on the Akamai-Anthropic cloud deal
2026-09-26 — US stocks closed out a winning week Friday (S&P 500 +0.51%, Nasdaq +0.48%, Dow +0.93%), with megacap tech leading a comeback -- Akamai jumped 15%+ pre
- SHMD +20.7% 8.9 AVOID — Real narrative, weak fundamentals -- a name to watch for a de-rate, not to hold through.
- KLXE +9.5% 8.9 WATCH — Constructive operationally but the rights offering adds float -- let the dilution digest before sizing up.
- KDK -12.6% 8.7 AVOID — Deteriorating momentum on a name with no revenue support -- stay out until launch-delay reports are confirmed or denied.
- KDKRW -21.4% 8.2 AVOID — Deteriorating momentum on a name with no revenue support -- stay out until launch-delay reports are confirmed or denied.
- AESI +13.5% 8.2 ENTER — Contracted, multi-year revenue visibility on a durable theme (AI data-center power) -- the cleanest catalyst of today's list.
- HQI +6.9% 6.8 WATCH — No news to hang a thesis on -- don't chase a no-catalyst pump in an already-overvalued micro-cap.
2026-09-25 — Thu Sep 24 was flat on the surface but ugly underneath: S&P and Nasdaq closed near unchanged, the Dow fell a third straight day (−162, −0.31% to 51,35
- GEN -12.1% 10.0 WAIT — Asymmetric on a 'talks ended' headline, but no edge before terms. Revisit on deal terms or a denial. Don't catch it on day one.
- QMCO +13.6% 8.4 AVOID — Chasing a ~7x-off-the-lows microcap at its highs on 5x volume with no catalyst. Revisit only on a flat-volume pullback to the low $20s.
- TSSI +9.8% 7.8 WATCH — Real catch-up room (r20d −1.9%, far off highs). The Oracle DC headline is the risk. Wait for a hold above $10 on normal volume.
- DK -7.4% 6.9 WAIT — Let the hedge selling finish, then reassess vs the conversion premium. Not a thesis break.
- PUBM +6.3% 6.2 WATCH — Initiation pops often fade within a week. Watch whether it holds $18 into next week.
- KPLT -18.7% 5.9 AVOID — The overhang isn't sized or timed. Avoid until volume dries up or holders disclose sales.
- ARMperpspot -7.9% 4.5 WAIT — Knife-catch candidate, but it has only given back ~1/5 of the week. Better entry is nearer the week's breakout (~$280) or after a volume-drying day.
2026-09-24 — Wed Sep 23 was a rates day, not an AI day: S&P 500 −0.75% to 7,706, Nasdaq −1.13% to 26,936, Dow −0.68%. Hot flash PMIs sent the 10-year to 5.135% (hi
- ALKT -19.3% 10.0 WAIT — Deal-premium unwinds take 2–3 sessions to find buyers who hold for fundamentals. Revisit if it holds about $14 on falling volume. The 'GF Value undervalued' headlines are
- VOYG -16.3% 10.0 WAIT — Let the convert price and the hedge selling finish (1–3 sessions). A retest of the low $30s on falling volume after pricing is the earliest look. Don't catch it into the hedge
- INVZ +16.6% 9.0 AVOID — No catalyst, sub-dollar, and reverse-split risk. Deciding fast to skip it counts.
- EGHT +15.1% 9.0 WATCH — Don't chase. If it holds above $2.10 for two sessions and the Citrini-peer bid continues, it becomes a momentum candidate. Otherwise let it fade.
- XNDUperp -7.8% 8.1 WAIT — Unlock supply takes weeks to clear. Wait for volume to fall back toward normal and a higher low before any KNIFE-CATCH.
- FSLY +13.7% 7.4 WATCH — Don't chase a viral-post candle. ENTER on a pullback that holds the investor-day gap (~$26–27). A close back below $26 means the narrative failed.
- GGLLperp -7.7% 7.1 WAIT — Don't trade GGLL (daily-reset decay). On GOOGL, wait for the rate spike to settle. It's a candidate KNIFE-CATCH only if it stabilises above its 50-day while META's M
2026-09-22 — Monday Sep 21 was the best US session since Aug 4: S&P 500 +1.49% to 7,764.70, Nasdaq +2.26% to a record 27,125, Dow +0.71% — a risk-on snapback five
- NUAI +30.6% 9.0 WATCH — Chase-risk: milestone pops on cash-burning developers get sold into financings. Look for a retest of ~$6.5–7 with volume drying up; ENTER only if no raise is announced in the next
- CRMLperp +38.6% 9.0 WAIT — Policy pop with a 3-month-high breakout — don't short it, don't chase it. Reassess after 3 sessions: if it holds >$8 on shrinking volume the Greenland theme has legs;
- ARMperpspot +17.2% 9.0 WAIT — z 4.2 on a $345B name is not an entry. If it holds $300 through Connect and the semi basket's r5d turns positive, it becomes ENTER on the first 3–5% pullback; a close under $2
- MX +16.4% 9.0 AVOID — Micro-cap, financing-driven spike below the deal price on 13x volume — classic one-day pop. Deciding AVOID fast is the trade.
- METAperpspot +11.4% 8.6 WATCH — Hold if long; do not initiate into Connect. A close back under $700 after the keynote is the fade trade; a hold above $740 on the keynote day would be rare and bullish.
- FIVN +14.0% 8.1 WATCH — 52w-high breakout with unknown cause = include-and-flag. Do not short; ENTER only on a retest of the $33–34 breakout level with volume, and only if no 8-K appears explaining a one-
2026-09-21 — MONDAY 21 Sep -- no completed US session; the US fields below (deep dives, signals, week movers, deals, basket rows) are Friday 18 Sep carried through
- KDK -14.4% 10.0 AVOID — Knives that fall on no news with a financing overhang cut deepest. Nothing to buy until a raise is done or a customer number lands.
- DSP +18.6% 9.0 WAIT — Do not chase a cancelled-offering squeeze. The trade is to short strength into the $13-14 zone where the deal was priced, or wait for the re-launch and buy that dip.
- GEMIspot +31.1% 9.0 WATCH — Policy overshoots. Let the first 15-20% giveback happen, then decide -- and trade the bigger venues (COIN/HOOD) for the theme rather than a $750M name that just tripled its volume.
- QCOMperpspot -5.8% 8.3 WAIT — Not a knife-catch yet -- one more red day toward the $165-170 pre-rally base makes it a KNIFE-CATCH with the semi tape on your side. Buying -6% on a +14% month is buying the middle
- PED -10.8% 8.0 WATCH — Overheat-cooling in a $172M name is a KNIFE-CATCH only with tight size; better to wait for a second flush toward the 20-day base (~$11) and confirm oil holds $90.
- KD -6.8% 7.9 AVOID — No catalyst, rising rates, negative FCF: this is exactly the profile the hike week punishes. Revisit only after a quarter that shows FCF turning.
2026-09-20 — SUNDAY 20 Sep -- no cash market traded anywhere; the US fields below are Friday 18 Sep carried through unchanged (S&P +0.17% to 7,650, Nasdaq +0.39%,
- KDK -14.4% 10.0 AVOID — Knives that fall on no news with a financing overhang cut deepest. Nothing to buy until a raise is done or a customer number lands.
- DSP +18.6% 9.0 WAIT — Do not chase a cancelled-offering squeeze. The trade is to short strength into the $13-14 zone where the deal was priced, or wait for the re-launch and buy that dip.
- GEMIspot +31.1% 9.0 WATCH — Policy overshoots. Let the first 15-20% giveback happen, then decide -- and trade the bigger venues (COIN/HOOD) for the theme rather than a $750M name that just tripled its volume.
- QCOMperpspot -5.8% 8.3 WAIT — Not a knife-catch yet -- one more red day toward the $165-170 pre-rally base makes it a KNIFE-CATCH with the semi tape on your side. Buying -6% on a +14% month is buying the middle
- PED -10.8% 8.0 WATCH — Overheat-cooling in a $172M name is a KNIFE-CATCH only with tight size; better to wait for a second flush toward the 20-day base (~$11) and confirm oil holds $90.
- KD -6.8% 7.9 AVOID — No catalyst, rising rates, negative FCF: this is exactly the profile the hike week punishes. Revisit only after a quarter that shows FCF turning.
2026-09-19 — Friday Sep 18 closed mixed to end Fed-hike week: S&P 500 +0.17% to 7,650, Nasdaq +0.39% to 26,523, Dow -0.18% to 51,683. For the week the S&P slipped
- KDK -14.4% 10.0 AVOID — Knives that fall on no news with a financing overhang cut deepest. Nothing to buy until a raise is done or a customer number lands.
- DSP +18.6% 9.0 WAIT — Do not chase a cancelled-offering squeeze. The trade is to short strength into the $13-14 zone where the deal was priced, or wait for the re-launch and buy that dip.
- GEMIspot +31.1% 9.0 WATCH — Policy overshoots. Let the first 15-20% giveback happen, then decide -- and trade the bigger venues (COIN/HOOD) for the theme rather than a $750M name that just tripled its volume.
- QCOMperpspot -5.8% 8.3 WAIT — Not a knife-catch yet -- one more red day toward the $165-170 pre-rally base makes it a KNIFE-CATCH with the semi tape on your side. Buying -6% on a +14% month is buying the middle
- PED -10.8% 8.0 WATCH — Overheat-cooling in a $172M name is a KNIFE-CATCH only with tight size; better to wait for a second flush toward the 20-day base (~$11) and confirm oil holds $90.
- KD -6.8% 7.9 AVOID — No catalyst, rising rates, negative FCF: this is exactly the profile the hike week punishes. Revisit only after a quarter that shows FCF turning.
2026-09-18 — Thursday Sep 17 was the post-Fed relief bounce: after Wednesday's 25bp hike to 3.75-4.00% (the first hike in three years, Chair Warsh framing it as in
- DSP -19.0% 10.0 WAIT — Act-fast means decide fast: not a buy until the reason surfaces. Reassess at $9 or on any 8-K. -1.7% 5d / -10.6% 20d before this — it was already drifting.
- ALMU -14.0% 9.5 AVOID — Small, fading basket, guide cut. Knives in fading baskets cut.
- RGR +8.2% 8.7 WATCH — Merger-arb math, not a swing. Watch for Beretta upsizing or a second filer. Week was -2.3% before this.
- DAIC +163.7% 7.2 AVOID — Untradeable size. Listed only because the velocity screen cannot filter market cap.
- DDD +12.7% 7.0 WATCH — Thin-volume bounce; not chasing. Prefer XMTR/PRLB for the theme.
- IPX +9.1% 6.6 KNIFE-CATCH — Starter size only: 29% off in 20 days, first up-day on 3x volume, basket still red. Stop below the Sep 16 low.
2026-09-17 — The Fed hiked 25bp to 3.75-4.00% — its first hike in three years — and Warsh's presser plus a dot plot with 16 of 18 seeing another hike this year kno
- MRLN -27.5% 10.0 WAIT — Do not catch on day one of a lockup unlock. Watch volume: when a day prints <3x average with price holding, that is the tell for a class-6 entry.
- RCAT -10.9% 9.0 AVOID — Fading + insider sale = no. Revisit only on a contract/order catalyst.
- ONspot -9.0% 7.6 WATCH — Round-trip of the pre-event run. Interested only near the $60 August lows with volume fading; no edge chasing an analyst-day gap-down under a hiking Fed.
- MGY -8.6% 6.9 KNIFE-CATCH — Quality E&P, sector-wide flush, 20d trend still up. Scale in; invalidation is Brent < $100 or a second build next week.
- SABR +11.0% 6.9 WATCH — Refi pop in a sub-$1B-cap levered equity; no follow-through edge unless the tender results (participation) surprise. Not a swing.
- BAperpspot -3.7% 6.6 WAIT — Quality franchise with a self-inflicted FCF cut in a hiking cycle — no rush. Level to watch is the $195 post-Q2 low; entry only on rate-stabilization signals.
2026-09-15 — US Monday (Sep 14): S&P 500 -0.48% to 7,620, Dow -0.29%, Nasdaq -0.56%, but the SOX fell ~5% (SMH -4%) after Anthropic's Amodei urged a slowdown in fr
- ARQQperpspot -18.8% 10.0 AVOID — No-reason dumps in serial diluters are not conviction-long territory. Re-check the 6-K feed; only interesting if volume dries up and no offering appears within 3 sessions.
- EVCM +19.8% 8.8 WATCH — Do not chase +20% in a broken chart. Watch for a higher low above $6.50 over 2-3 sessions; the setup only becomes a trade if software holds its bid through the Fed.
- KGS -11.3% 8.6 WAIT — Whale-fight day inside a power/AI basket — let the Fed and the Johnson AI meeting pass. Interested near $55 if the power names stabilise; the dividend yield does the waiting for yo
- PBF -10.1% 8.5 WAIT — Cracks near record and a -10% day is normally a KNIFE-CATCH, but the Fed on Wednesday and Hormuz headlines make the entry a coin-flip. Wait for the Fed; re-enter only if cracks are
- TENB +16.5% 8.4 WATCH — Do not chase +16%. If CRWD/PANW hold their gaps through the Fed, ENTER TENB on the first pullback toward $32-33; if the gaps fill, the whole rotation was a one-day event.
- TERperpspot -13.3% 8.1 WAIT — Not a knife to catch before the Fed. Watch $310 (the pre-run base); if it holds through Thursday and SOX stabilises, this is the first AI-hardware name to buy back because the down
Not financial advice. Swing trading is risky; velocity scores and verdicts are model+analyst opinions generated from public data. Do your own research.
generated 2026-10-01 19:31 HKT · data: Webull, yfinance, SEC filings, X/Reddit research