PINNACLE DOMINION SIGNALS · v2.1 calibration-anchored · dual-track

PDS Daily Report — 2026-05-23

STAGFLATION_HOT (de-intensifying surface) Market fade-bias: HIGH (gamma flipped +) 6 open · 5 prints in 3 sessions Live book = journal (re-anchored)

NET TOTAL P&L
+A$8,893.9

Open unrealized
+A$5,279.5

Realized cum.
+A$3,614.4

Portfolio
A$99,836

Cash (buffer A$8,857)
A$9,119

US session Fri 5/22 CLOSED · Memorial Day Mon 5/25 closed · FX 0.7153 · all marks UW OHLC 5/22 reg close, fetched this session. Re-anchored to the journal's 6-name OPEN POSITIONS (the prompt's stale 4-name list was overridden).

§1 · Market Snapshot

Asset	Last / Move (5/22 close)	vs prior	Source
S&P 500	7,473.47 (+0.37%)	8th straight up week	TheStreet 5/22
Nasdaq Comp	26,343.97 (+0.19%)	new high zone	TheStreet 5/22
Dow Jones	50,579.70 (+0.58%)	RECORD close	TheStreet 5/22
VIX	19.65	elevated, off lows	TradingEconomics 5/22
UST 10Y	4.558% (−2.6bp)	eased	TradingEconomics 5/22
UST 30Y	5.064% (−4.7bp)	eased	TradingEconomics 5/22
2s10s	+49bp	steepening	computed 5/22
WTI	$96.36	little-changed on wk	TheStreet 5/22
UMich sentiment	44.8 RECORD LOW	3rd monthly drop; 1yr infl-exp 4.8%	UMich 5/22
Jobless claims	209k (est 210k)	inline, labor firm	5/21
SPX dealer gamma	NET POSITIVE +1.11M	FLIPPED from −0.54M 5/21	UW GEX 5/22

§2 · Regime Classification & Drivers

Regime = STAGFLATION_HOT — de-intensifying on the surface (equities at/near records, yields easing, oil flat), unresolved at the core (record-low consumer sentiment 44.8, 1-yr inflation expectations 4.8%). vs prior session: SAME bucket, softening surface.

⚠ Market fade-risk bias = HIGH. SPX dealer gamma flipped NET POSITIVE on 5/22 (+1.11M vs −0.54M on 5/21) — vol-suppressive / mean-reverting, so post-print spikes are more likely to fade. Combined with STAGFLATION (rallies sold) and the NVDA beat-and-fade on 5/20 (AI-semi peer warning), every held name's Fade-Risk is biased UP toward POP-SELL. This is a stronger fade signal than the NEUTRAL bias logged 5/22.

§3 · Portfolio Mark & P&L Reconciliation (6 open)

Tkr	Sh	Entry	Live 5/22	P&L AUD	%	IVrk	% to stop	Next ER
MRVL	41	168.53	196.33	+A$1,593.5	+16.5%	79.5	+6.1%	Wed 5/27 AMC
ORCL	83	185.24	192.08	+A$793.7	+3.7%	78.9	+1.6%	~Jun 10
DELL	31	244.725	295.19	+A$2,187.1	+20.6%	92.2	+13.5%	Thu 5/28 AMC
ZS	48	176.05	182.37	+A$424.1	+3.6%	90.5	+8.5%	Tue 5/26 AMC
CRM	48	178.07	180.07	+A$134.2	+1.1%	87.9	+5.3%	Wed 5/27 AMC
MDB	44	323.74	326.13	+A$147.0	+0.7%	90.6	+7.2%	Thu 5/28 AMC
OPEN (6)	295	—	—	+A$5,279.5	—	—	—	5 ≤6d

P&L reconciliation: per-position AUD sum = 1,593.5 + 793.7 + 2,187.1 + 424.1 + 134.2 + 147.0 = A$5,279.6 ≈ header +A$5,279.5 (diff <A$1 rounding) → OK. Invested US$61,113.45 = A$85,438; value US$64,889.90 = A$90,717. NET TOTAL = realized +A$3,614.4 + open +A$5,279.5 = +A$8,893.9.

STOP-PROXIMITY FLAG: ORCL is only +1.6% above its 188.98 stop (within the 3% flag) — next stop-out candidate on any fade. All others have room (CRM +5.3%, MRVL +6.1%, MDB +7.2%, ZS +8.5%, DELL +13.5%).

NOT held (in realized): ROST exited 5/21 AH @231.00 (+A$2,529.1); AVGO stopped out 5/22 @411.00 (−A$531.2). The trigger prompt's old "ROST 91 @211.12" line was stale — overridden by the journal's live book.

§4 · News Scan — last 24h, filtered

Tag	Item / so-what
[DELL]	Gapped +16.8% on the day (DELL/HP double-digit pre-earnings strength). Textbook pop-sell spike into 5/28 AMC; +13.5% above stop, riding. TheStreet 5/22
[MRVL]	Stifel maint Buy, PT $210 (5/22) — fresh 24h, headroom vs $196.33. UW analyst
[DELL]	Wells Fargo maint Overweight, PT $270 (5/22) — but $270 is BELOW the $295.19 spot; stock above nearly all PTs. UW analyst
[ZS]	Print 5/26 AMC — NEAREST on the board (first one after Memorial Day). 8/8 beat streak. UW earnings hist
[MDB]	Print 5/28 AMC — 8/8 beats but most violent reaction history (last4 +12.8/+38/+22/−22%). Hard rule: never hold-through. UW earnings hist
[MACRO]	UMich record-low sentiment 44.8 + 1yr infl-exp 4.8% vs 8th straight S&P up-week = rally-into-deteriorating-consumer = fade setup. UMich 5/22
GAP	UW get_correlations returned noisy top-N; §9 pairwise values sector-reasoned, not tool-confirmed. Carried from 5/22.

§5 · Dual-Track Exit Plan (v2.1)

Tkr	Print	P(beat)	IM (7d)	Hold-thru	Pop-sell	Fade-Risk	Primary	Execution
ZS	5/26 AMC	~85%	±11.3%	7.2	8.3	7.8/10	POP-SELL	Trim into pre-print run; sell 48sh first 1–3 min 5/26. Only cyber name.
CRM	5/27 AMC	~80%	±7.4%	6.5	7.0	6.5/10	POP-SELL	Steadiest/thinnest-IM name; modest spike to sell. Exit 48sh first 1–3 min 5/27.
MRVL	5/27 AMC	~74%	±11.4%	7.9	8.2	8.0/10	POP-SELL	Sell residual 41sh first 1–3 min 5/27; hold-to-settle only on genuine guide raise.
DELL	5/28 AMC	~70%	±10.0%	7.1	8.0	7.5/10	POP-SELL	First 1–3 min exit of 31sh 5/28; Q1-miss precedent + PTs below spot cap upside.
MDB	5/28 AMC	~85%	±13.1%	7.0	8.6	8.5/10	POP-SELL	Largest IM, highest gap risk. NEVER hold-through. Sell 44sh first 1–3 min 5/28.
ORCL	~6/10	n/a	±11.5%	5.2	n/a	n/a	HOLD	Not pre-earnings. Hold, stop 188.98 (tightest, +1.6%). Decision ~Jun 8–9.

5 imminent prints in 3 sessions (ZS 5/26 · CRM+MRVL 5/27 · DELL+MDB 5/28) — all POP-SELL primary; the gamma flip raised every Fade-Risk +1.5 vs 5/22. With cash ≈ buffer, there is no capacity to add or average. The job is EXITING into strength.

§6 · §5.5 Probability Translation — 6 held names

Conviction is derived from base rates. Hold-through: round1(E[R]×2 + P(beat)×4 + 3), capped 7.9 unless P(beat)>75% AND E[R]>1.8. Pop-sell: round1(P(spike↑)×4 + min(spike%/3,3) + liq_floor). High IM HELPS pop-sell, hurts hold-through.

ZS — 5/26 AMC (cyber)

Beat rate (last 8q): 8/8 = 100%, avg surprise ~+20%; essentially never misses
Last4 1d reactions: +9.8 / −1.5 / −13.0 / −12.2% (two-sided — IM is monetizable)
IM ±11.3% (7d); IVrk 90.5; P(beat) ~85%
Hold-through conv 7.2 · Pop-sell conv 8.3 · Fade-Risk 7.8/10

POP-SELL primary. Trim into pre-print strength; sell residual first 1–3 min 5/26 AMC. Stop 168.13.

CRM — 5/27 AMC (SaaS, $163B)

Beat rate (last 8q): 7/8 = 88% (lone miss Oct'24 −1.6%); same-Q1 last yr beat +1.2%
Last4 1d reactions: −3.3 / −4.9 / +3.7 / +4.0% (STEADIEST, react_vol ~4%)
IM ±7.4% (7d, thinnest); IVrk 87.9; P(beat) ~80%
Hold-through conv 6.5 · Pop-sell conv 7.0 · Fade-Risk 6.5/10

POP-SELL primary. Thin IM = modest spike; lowest-drama exit. First 1–3 min 5/27. Stop 170.95.

MRVL — 5/27 AMC (AI-semi)

Beat rate (last 8q): ~6 beats / 2 inline ≈ 75%; magnitude compressing
Last4 1d reactions: −5.6 / −18.6 / +7.9 / +18.4% (bimodal)
IM ±11.4% (7d); IVrk 79.5 (cooled from 93.8); P(beat) ~74%; Stifel Buy PT$210 (5/22)
Hold-through conv 7.9 · Pop-sell conv 8.2 · Fade-Risk 8.0/10

POP-SELL primary. Sell residual 41sh first 1–3 min 5/27; hold-to-settle only on a genuine guide raise. Stop 185.00.

DELL — 5/28 AMC (AI-hardware)

Beat rate 7/8 (88%) but the 1 miss (−8.28%) was the SAME Q1 reporting now → P(beat) haircut to 70%
Last4 1d reactions: −2.1 / −8.9 / +5.8 / +21.9% (bimodal)
IM ±10.0% (7d); IVrk 92.2; gapped +16.8% on 5/22; WF OW PT$270 (below spot 295.19)
Hold-through conv 7.1 · Pop-sell conv 8.0 · Fade-Risk 7.5/10

POP-SELL primary. First 1–3 min exit of 31sh 5/28; watch the same-quarter miss precedent. Stop 260.00.

MDB — 5/28 AMC (data-tech, HIGHEST gap risk)

Beat rate (last 8q): 8/8 = 100%, avg surprise ~+57%; never missed publicly
Last4 1d reactions: +12.8 / +38.0 / +22.2 / −22.2% (react_vol ~22% — highest on board)
IM ±13.1% (7d, LARGEST); IVrk 90.6; P(beat) ~85%
Hold-through conv 7.0 · Pop-sell conv 8.6 · Fade-Risk 8.5/10

POP-SELL ONLY — HARD RULE: never hold-through (a −22% gap on 44sh ≈ −A$4.4k overnight). Sell 44sh first 1–3 min 5/28. Stop 304.32.

ORCL — ~6/10 (hold, long-duration)

Non-earnings hold; no imminent catalyst. IM ±11.5% for the Jun print (context only)
Carried risks: fraud class actions / $50B capex / Schiff bill — ALL STILL ACTIVE
Conviction 5.2 (flat vs 5/22; 10Y −2.6bp offsets nothing material)

HOLD, tight stop 188.98 (tightest in book, +1.6%). Pre-earnings dual-track decision ~Jun 8–9.

§7 · Conviction-Δ Discipline (vs 5/22 report)

Tkr	5/22	Today (primary)	Δ	Status	New inputs / cap
MRVL	8.5 pop	8.2 pop	−0.3	CITED	IVrk cooled 91.8→79.5; Stifel PT$210 5/22; within re-weight cap
DELL	8.3 pop	8.0 pop	−0.3	CITED	IVrk eased; WF PT$270 below spot; +16.8% gap day
ORCL	5.2 hold	5.2 hold	0.0	OK	10Y −2.6bp offsets; recovery stalled; flat
ZS	cand ~5.9	8.3 pop	new held	BASELINE	First strict dual-track as a HELD name (was candidate 5/22)
CRM	cand 6.4	7.0 pop	new held	BASELINE	First strict dual-track as a HELD name
MDB	excluded	8.6 pop	new held	BASELINE	First held baseline; was sub-$20B/ADV-excluded, now held

§8 · Carried-Forward Risks (nothing dropped without resolving data)

Tkr	Risk	Status	Resolving data?
ORCL	Securities-fraud class actions	✓ ACTIVE	No resolving data
ORCL	$50B FY26 capex / FCF −$10B	✓ ACTIVE	Marks at Jun 10 print
ORCL	Schiff energy-cost bill	✓ ACTIVE	No resolving data
MRVL	Drift-up-into-print fade (DE shape)	✓ ACTIVE	IVrk cooled = ↻ partially eased
DELL	Q1-miss precedent (same quarter)	✓ ACTIVE	Resolves 5/28
DELL	PTs below spot / extended +16.8% gap	✓ ACTIVE	Resolves 5/28
MDB	Violent two-sided gap (last −22%)	✓ ACTIVE	Resolves 5/28; hard no-hold rule
ZS/CRM	Two-sided / steady reaction var	✓ ACTIVE	Resolve 5/26, 5/27
BOOK	5 stacked prints, no dry powder	✓ ACTIVE	Shrinks as each is pop-sold into strength
MARKET	Gamma-positive fade-bias + stagflation	✓ ACTIVE	UMich 5/22, GEX 5/22 reinforced

§9 · Correlation-Adjusted Book Conviction

DATA GAP (carried): UW get_correlations returned noisy top-N + empty pair rows; pairwise values below are sector-reasoned, not tool-confirmed.

Clusters: AI-capex MRVL+DELL (~0.60; 2 prints 5/27–5/28) · software/data ZS+CRM+MDB (~0.45; 3 prints 5/26–5/28) · ORCL standalone (~0.35–0.40 vs both).
Avg pairwise ~0.42 → use the 0.4<corr<0.7 band: eff = max(8.6) + 0.5×(sum−max) = 8.6 + 0.5×36.7 = 26.95 vs raw sum 45.3 = 60% → CONCENTRATION FLAG (<80%).
Mitigant: all six are pop-sells to be exited into strength within 3 sessions — the book mechanically de-risks as each prints. The flag is a "don't add" signal, not a "reduce now" signal.

§10 · Top Earnings Candidates (next 45d) — context only

Rk	Tkr	Report	IM%	MktCap	Sector	Note
1	AVGO	6/3 AMC	8.5	$1.98T	Tech semi	AI-capex read-through; re-concentrates vs MRVL/DELL — DEFER
2	PANW	6/2 AMC	9.0	$201B	Cyber	Would stack a 2nd cyber vs ZS — avoid cluster
3	CRWD	6/3 AMC	9.3	$165B	Cyber	Same cyber-cluster caution
4	ADBE	6/11 AMC	9.3	$102B	SaaS	Outside the 3-session window; watch
5	SNOW	5/27 AMC	11.3	$58B	Data	High-IM but correlates with MDB/data cluster

0 NEW initiations. Book is ~99% deployed (cash ≈ buffer) with 5 prints in 3 sessions — no capacity to add. Every listed candidate also re-concentrates an existing cluster (AI-capex / cyber / data). Posture is EXIT-ONLY until the stack clears.

§11 · Scenario Planning (sums to 100%)

Scenario	Prob	Book impact
BULL: spikes hold, pop-sells executed into strength	35%	Bank +A$5.3k open into prints; realized climbs toward ~A$9–11k as each is sold on the first-min spike
BASE: spike-then-fade (gamma-positive)	45%	Pop-sell discipline captures the spike; hold-through would round-trip. This is the modal path given the fade-bias
BEAR: a miss or guide cut gaps a name down	20%	MDB −22% tail (~−A$4.4k on 44sh) is the worst single-night; stops on ORCL (+1.6%) likely trigger first on any broad fade

§12 · Key Takeaways

Live book = the journal: 6 open (MRVL/ORCL/DELL/ZS/CRM/MDB). ROST exited 5/21, AVGO stopped 5/22 — both in realized, NOT held. NET +A$8,893.9.
5 prints in 3 sessions, all POP-SELL. ZS 5/26 · CRM+MRVL 5/27 · DELL+MDB 5/28. The gamma flip to net-positive makes fade more likely — sell the first-min spike, do not hold to settle.
MDB is the hard rule: never hold-through (±13% IM, last print −22%). Largest pop-sell conv (8.6) AND largest tail.
ORCL is the live stop risk (+1.6% above 188.98) — first to trigger on any broad fade; it is the only HOLD.
No adds. Book ~99% deployed, concentration flag at 60%. Posture is exit-into-strength; the book de-risks itself as each name prints.