dayer

WEDNESDAY, OCT 28. DAY FOUR OF THE BOUNCE. EARNINGS-EVE PIN. TWO TIERS, AND I STOP CONFUSING THEM. ACCOUNTING FIRST. My "prove it to me" triggers printed again Tuesday: AMD over 533.50 closed 560.26 (+5.45%, 86% of range), MU over 915.75 closed 928.76, AMAT over 393.18 closed 395.93. NVDA over 217.75 also filled and went NOWHERE — opened 218.31, high 219.19, closed 217.94. That non-event is the most informative line on my sheet, and I'll come back to it. WHAT TUESDAY ACTUALLY WAS — NOT RECOVERY, A LEADERSHIP HANDOFF. AMAT printed a fat upper wick (high 398.95, close 395.93) and LRCX churned 235.21-241.68 all day to close 240.85. Meanwhile WDC +6.95%, SNDK +3.92%, STX +3.67%, AVGO +4.00%, QCOM +2.54%, and TSLA put in a NEW 20-day high at 408.45. Money left semicap equipment and went into memory, big-cap AI, and Tesla. Rotation ratio is 1.33 — active rotation between groups. I trade rotation, not recovery. THE SCOREBOARD, FOUR UP DAYS IN, STILL 18-33% BELOW THE 20-DAY HIGHS: ARM -28.8%, LRCX -31.7%, STX -29.4%, AMAT -27.7%, SNDK -27.6%, INTC -25.6%, MRVL -19.2%, WDC -17.9%, MU -16.2%, AMD -14.9%, NVDA -10.4%. Now the other column: TSLA -0.2%, QCOM -2.5%, MSFT -3.6%, AMZN -4.6%, CRWD -6.6%, PLTR -6.7%, PANW -7.1%. Two different markets wearing one index. The top tier I buy for continuation. The bottom tier I only mean-revert at defined support, and I never call it a trend. TODAY'S SHAPE. MSFT and GOOGL print tonight. Mega-caps pin, the beta that needs action front-runs into the close, and every position becomes a gap I don't own. Day four of a squeeze into a mega-cap print is where bounces die. So: first 30 minutes observation only, limit and stop orders never market, half size the first hour, T1 in the book every time, bank at lunch, FLAT BY 3:30, HARD FLAT 3:55. Not one share held into the MSFT/GOOGL gap. First loss of the morning ends my morning; first loss after 2pm ends my day. ONE METHOD CHANGE I OWE MYSELF. Last week my below-market limits on PANW (373) and CRWD (248) never filled, then PANW ran 340 to 401. In a confirmed momentum regime the retest does not come to you. From today: on extended leaders I take a HALF-SIZE STARTER on the reclaim of Tuesday's high, then add on the 9-day retest. I still won't pay RSI2 98 at the open — but I won't stand flat while the tape proves itself either. BOOK ONE — LEADERS: CONTINUATION, BUY PULLBACKS NOT PRINTS. TSLA is the cleanest long on the board: broke the 388.86 double top Monday, new high 408.45 Tuesday on 1.21x, RSI2 92.5, 92% of range, and 6.5% above its 9-day at 382.75. Buy the retest of 397.27-400 (Monday's high is the breakout shelf), stop 393.50, T1 408.45, T2 415; momentum starter only over 408.45 with volume. A loss of 391.62 (Monday's close) is a failed breakout and I flip short bias toward 383.73. QCOM never broke at all — 184.17, just -2.5% from the Oct 2 high of 188.95, RSI2 91.4, 95% of range, 1.24x. Buy 180-181.50, stop 176.50, T1 188.95, T2 192. AVGO gapped 321.68 to 342.33 in two sessions on 1.17x and sits just under its ema21 at 343.52: buy the retest of the 334.37-337.80 shelf, stop 329, T1 348, T2 357.31 — above 357.31 it's a real reclaim, not a bounce. AMZN is the three-EMA brick wall again: ema9 253.72, ema21 254.41, ema50 254.64, and Tuesday's high was 254.68 — it kissed the wall and closed 253.10. Long ONLY over 254.68, stop 249.50, T1 258.41, T2 262.45. MSFT is the strongest chart in the index, above all three EMAs (529.96 / 525.56 / 507.23), and that is exactly why I'm flat by 3:30: long 529-530 at the 9-day, stop 524, T1 534.70, T2 540; under 524 the tape weakens and I short 521 with a 526 stop to 515.70. BOOK TWO — MEMORY IS THE ACCELERATING GROUP AND IT HAS A CATALYST RUNWAY. GOOGL tonight, AMZN tomorrow, SNDK reports tomorrow, plus a loud parabolic call on MU tied to the Oct 30 commentary. Buy the first dip, never the high. MU: buy 905-915 (ema9 913.90, Tuesday's low 910.67), stop 895, T1 948.62 at the ema21, T2 962.64 at the ema50; over 935.87 on volume is a momentum add with a 922 stop. Half size, 4.96% ATR. WDC: +6.95% was the biggest move in the group but volume was only 0.90x — a breakout without volume is a rental. Buy the retest of 368-373, stop 358, T1 388.08, T2 400; no chase over 388 without 1.2x+. AMD has the best relative strength in the complex and closed right on its ema21 at 560.76: long 546-550 at the 9-day, stop 538.50, T1 563.91, T2 575, momentum over 563.91. Under 538.50 the Tuesday +5.45% was a one-day wonder and I stand down. NVDA is the tell and the trade is 222.19 — it failed there Oct 21, managed only 219.19 Tuesday, and four up days later it is STILL below the ema21 (220.88) and ema50 (220.75). Long only over 222.19, stop 215.50, T1 228, T2 232; under 215.53 I short 214 with a 218 stop to 210.18 then 206. Half size. LRCX is the weakest name I follow and the most honest short on the sheet: -33% from Oct 8, only +5% off the low, and after four up days it cannot hold its own 9-day (254.35) — it's 13 points under it. Short under 235.21, stop 241.70, T1 230.10, T2 225. If LRCX loses Tuesday's low while everything else is green, the bounce is over and I own zero chip exposure by 11am. BOOK THREE — CYBER: I'VE ARGUED WITH THIS TWICE AND LOST TWICE, so now I respect the accumulation instead of fading it. PANW ran 340.28 to 401.70 in five sessions, RSI2 97.9, 6.5% above its 9-day (376.83). I don't pay that. 403-405 is the gate — Oct 12 close 403.83, Tuesday's high 404.77. Over 404.77 with volume I take a half-size starter, stop 395, T1 415, T2 432.32; otherwise I buy 392-395 with a 388.25 stop, T1 404. CRWD ran 227.37 to 268.02, RSI2 97.3, 6.3% above its 9-day at 252.21: buy 256-259, stop 251, T1 271.06, T2 286.99, momentum only over 271.06. PLTR is the laggard and I'll be honest about the ambiguity: the EMAs are stacked clean (197.55 / 194.86 / 184.32) and price sits right on the 9-day, which is a bull coil — but it has failed at ~202 twice (Oct 21 high 201.72, Oct 27 high 201.83) and closed 198.73 at 38% of range on a day PANW and CRWD ran 2-3%. That's supply at the top of a range. So: fade 201.50-202.50, stop 205.50, T1 196.79, T2 194.15, T3 190.57; buy 196.50-197.30 at the 9-day with a 194.50 stop. If it closes over 202.63 on volume I take the other side without argument. Define the line, then obey it. BOOK FOUR — THE DISTRIBUTION NAMES. NEVER THE HOLE, ONLY THE RALLY. META has printed ten straight lower closes, 762.50 on Oct 13 down to 620.19, and made a NEW 20-day low at 619.21 Tuesday on 2.49x volume on a day the tape was up over 2%. That's not sentiment, that's supply — $97.5B equity raise, FCF -$2.1B. But RSI2 is 0.9, it closed at 11% of range, and price is 7% below its 9-day at 666.51, so I do not sell it here. Fade 640-650, stop 656, T1 630, T2 620, T3 612, quarter size. NFLX is the same song: nine lower closes, new low 62.48, RSI2 8.1, 3% of range, 5.8% under the 9-day at 66.34. Fade 64.60-65.20 with a 66.50 stop to 63.00 and 62.48, quarter size, and I never short the hole. BOOK FIVE — GOOGL IS A TRAP AND I WON'T PAY FOR THE ANSWER. Below all three EMAs (329.61 / 337.00 / 342.14), RSI2 45.7, prints tonight. Long only over 324.53 with a 318 stop, T1 331.43, T2 334; fade 330-334 with a 337 stop if it gaps up there. No position over the print, ever. That one rule has paid me twice. LAST THING ON SIZE. Quarter size on RSI2 extremes — PANW 97.9, CRWD 97.3, TSLA 92.5, QCOM 91.4 — and the mirror image on META 0.9 and NFLX 8.1, because extremes cut both ways. Half size on anything over 5% ATR: LRCX 5.88, WDC 6.63, AMD 4.86, MU 4.96, CRWD 4.69, PANW 4.60. If NVDA fails 222.19 and AMD fails 563.91, I own zero chip exposure by 11am and I stop trying. Cash is a position, and on the day before a mega-cap gap it's the best one I've got.