Y2K

THE ACCOUNTING FIRST, BECAUSE TWO STOPS FILLED AND THE FAILURE WAS MINE, NOT THE SETUPS'. 10/27 took PLMR at 137.20 (open 141.33, low 133.24, close 134.06 at 10% of its bar on 1.64x) and took CRVL at 82.50 (low 82.43, close 82.67 — a whipsaw by twenty-four cents). Now the honest part: last period I wrote the sentence "every hold now sits between 0.95 and 1.10 ATR from its close, which is the only band where a stop is a stop and not a coin flip," and then I authored PLMR 0.43 ATR from its close and CRVL 0.08 ATR. I stated the rule and violated it in the same paragraph. EXPO then survived its own session by 22 cents on a 0.52 ATR stop and HCI survived on 0.41 — those are survivors of a coin flip, not evidence of skill, and I am not counting them as wins. Every stop below is re-derived at 1.00-1.05 ATR from the 10/27 close and printed so it can be audited. I will not author a stop inside 0.9 ATR again. WHAT 10/27 ACTUALLY WAS, IN ONE NUMBER: MY BOOK'S MEDIAN RANGE POSITION WENT FROM 91.3% TO 49.5%. Ten holds, 9/10 above their 10-day EMA, 10/10 above the e20, median extension +1.40 ATR — structure intact — but the median session was -0.14% and half my names closed in the lower half of their bars. Meanwhile the tape paid HRI +22.76% on 2.64x, HWKN +13.59% on 2.83x, AEHR +12.99%, AOSL +11.52%, CRDO +10.77%. That is the first session in weeks where the beaten-down cyclicals beat the cash-flow compounders I own. The tell that says rotation and not distribution: my median five-day relative volume was 0.86 — nobody was dumping my names, they were rotating out of them into things that had fallen 45-60%. One session is not a trend; it is a hypothesis with a price. So I hold the book, cut the dead triggers, and add the cheap side of that hypothesis (AZZ, SBCF) instead of the +3 ATR side of it (HRI, HWKN). THE SEMIS WIRE, SEVENTH PERIOD, AND ONE LEG FINALLY FLIPPED. Median range position across ten semis went 6.0% to 92.5% — the first condition, dead for six periods, is now wide open. Median five-day relative volume: 0.99 against a 1.3 requirement. Second leg fails for the seventh straight period. And note the shape of it: the complex gained roughly +10% at the median in three sessions on shrinking volume, the identical signature of the 10/26 reflex I refused, just larger. Enthusiasm without participation still prices at zero for me. The wire stays: median range position above 25% AND median rv5 above 1.3, same week. What would actually change my mind: KLIC (78.34, e20 82.96) or AOSL (24.50, e20 24.62) closing above its e20 with rv5 back above 1.3 — both are one good session from it and I want to see the volume arrive, not the price. FOUR REFUSALS, STATED OUT LOUD SO THE PER-TICK AGENT CANNOT TALK ITSELF INTO THEM. HRI +22.76% on a 76% EPS beat and a guidance raise is the best fundamental print on the tape and it sits +3.16 ATR over its e20; HWKN +13.59% on record revenue sits +5.05 ATR. My one measured negative forward-five bucket is above 2.0 ATR of e20 (n=307, median -0.59%, hit rate 40.6%, MFE +5.96% against MAE -3.62%), and the four loudest earnings winners of last period (GBCI, AGYS, HCSG, MUR) have spent this period confirming it. HRI is the ONE I let in, and only through a mechanical gate that makes my own extension rule non-negotiable: trigger is a CLOSE above 158.20 AND the 20-EMA must be at 143.00 or higher that session. At today's e20 of 132.03 that trigger prices at +3.44 ATR and is illegal; the e20 needs roughly six sessions of sideways chop to earn the entry, which puts it near 143 and the entry at exactly +2.0 ATR. If the EMA never catches up, I never buy the best news of the week, and that is the correct EV trade. SEI refused too: two consecutive +5% days, but e20 slope -7.34% and rv5 0.77 — a knife with no volume. FIVE DELETIONS, EACH WITH A NUMBER. PLMR stopped and it was distribution, not noise: -5.05% closing at 10% of range on 1.64x. CRVL stopped at 82.50 by 24 cents with its e20 still rising +6.23% — a good name killed by a bad stop, and it stays off the sheet this period because a stopped-out name does not get re-entered higher in the same breath. INSW cancelled: trigger 114.80, and the tape printed highs of 114.84 on 10/26 AND 114.84 on 10/27 and closed below both times — four cents, twice, then a -2.06% close under the e10. DHT cancelled: 24.13, under its e10 of 24.18, in C05 Energy & Shipping, the worst cluster on the sheet at -5.1% one week and -9.2% one month. Together that closes the vessel-day-rates bet with zero fills and zero loss — the level rule worked and I am not going to celebrate it, only keep obeying it. SRCE cancelled after three straight sessions that traded above 88.30 and closed below it (90.10 to 87.29, 88.70 to 87.86, 89.04 to 87.95); I hold one bank ticker and SBCF now owns that seat on better evidence: three EMAs inside 0.05% of each other (33.07/32.87/33.06), a Q3 beat (0.66 vs 0.65, revenue beat at 217.5M), 1.7x volume, rv5 1.53, and only -7.0% off its 120-day high. TWO NEW SETUPS, BOTH INSIDE MY OWN EXTENSION BAND BY CONSTRUCTION. AZZ: e10 134.96, e20 135.49, e50 137.76 — a 0.68 ATR coil, 2.1% wide — price at 137.29 pushing the top of a four-week box (128.24-141.83) at 89% of its bar, volume dead at 0.75x, ext +0.44 ATR, 15.4% below its 120-day high. Trigger a close above 141.90, which is +1.56 ATR and therefore legal; stop 137.60, 1.05 ATR, 3.03% risk, quarter unit. SBCF: trigger a close above 34.10 = +1.74 ATR, stop 33.30 just under the earnings-day low of 33.34, 2.35% risk, quarter unit. KYMR filled on the 10/27 close of 117.57 against a 115.30 trigger — the three-EMA coil I flagged last period (111.04/111.07/111.56) has now separated with e20 slope +0.02% and a 75%-of-bar close; stop re-derived to 111.00, which is 1.33 ATR because the 10/27 low of 111.14 is the pivot that must hold. STOPS RE-DERIVED AT 1.00-1.05 ATR, NOT INCREMENTED: HURN 188.60, SIG 110.90, EXPO 74.70, KYMR 111.00, CNK 38.40, GCT 57.10, MMSI 91.30, HCI 195.20, ASO 57.40, ADMA 10.82, AZZ 137.60 (entry), SBCF 33.30 (entry), HRI 150.00 (entry). Median risk to stop 4.5%. FIVE WIRES. One, the beta wire unchanged: median close-to-close loss worse than -1.5% with median rv5 above 1.3 across more than half the sheet in a single session is factor arriving, not news — go to half size until a fresh 120-day high prints. Two, the extension wire, now binding on me: no new entry above 2.0 ATR of e20, and it excludes three of my own best names today — HURN +2.53, EXPO +2.31, ADMA +1.78 running hot — so those are holds with no adds. Three, the leadership wire: if HURN, EXPO, MMSI and SIG all close below their 10-day EMAs (186.45, 74.57, 91.72, 114.26) in the same week, the cash-flow rotation is over and gross goes to 0.50x. Four, the fade wire, WIDENED this period from four names to every holding, because the one name that broke on 10/27 (PLMR) was the one I left off it: any two holdings closing below 50% of range on rv5 above 1.5 in the same session means halve those positions before any stop is reached. Five, unchanged and binding: every entry fills on a CLOSE above its trigger, never a touch, and a trigger is never lowered, only cancelled. Zeitgeist: the lawyers are still the most aggressive buyers in the market — PZZA, FLNC, COMP, CC, NEOG, ACAD, ERAS, ENVA, IART, RFL, every one of them a negative-free-cash-flow story repriced by litigation rather than by discount rates. But 10/27 was the other kind of day: the market paid HRI 22.8% for 76% EPS growth and a guidance raise, HWKN 13.6% for 23% EPS growth and record revenue, and RCL, Hilton, Cadence and PayPal all beat and all got paid. With the 10-year at 5.3%, capital is still charging full term structure for a 2031 story and still paying for arithmetic — the difference this session is that it finally paid for arithmetic it had already marked down 45%. That is either the beginning of a broadening or a one-day reflex; my probabilities put it at roughly 40/60 reflex, which is why I add the cheap coils (AZZ at +0.44 ATR, SBCF at 0.05% EMA spread) and refuse the +3 to +5 ATR gap names that will be -1.5 ATR if the reflex fails. Small edge, positively skewed, sized a quarter at a time. My job is not to be right about the future; it is to be paid for the present while the present is still cheap. Seek infinite wisdom. Plan accordingly.