IBM IBM Common - Trade Plan — 2026-05-21

TL;DR

NO-TRADE. Size zero. Conviction 3. The research-manager ruled NO-TRADE with the bear winning the debate on three structural points the bull did not rebut: the Jul-17 spread expires five days before the Jul-22 Q2 print, the Q1 reaction function showed a -22% drawdown on a +5.4% EPS beat (consulting CC growth is the gating variable, not EPS), and IBM's +0.413 conditional HYG correlation amplifies — does not diversify — the existing four-leg US-tech basket (MSFT/AAPL/AMZN/GOOG, ~15.1% NAV). We are flat IBM. Spot $249.38. Re-engagement is a post-Jul-22 question, gated by the triggers below. Dry powder preserved.

Source verdict

Verbatim from data/reports/IBM/2026-05-21/research-verdict.md:

Decision. NO-TRADE — wait for the 2026-07-22 Q2 print to resolve the consulting binary, OR for the Bear's Trigger-A macro confirmation (HYG > 80.4, TLT > 88, XLK-vs-SPY 1m RS < +2%) to materialise, OR for concentration relief via close/size-down of one of the four existing US-tech longs (MSFT, AAPL, AMZN, GOOG). This is not a directional short — the volume signature, ADX/DI confirmation, and FY revision breadth (ERB +0.43) make the short side a fight-the-tape position with no edge. The right side today is FLAT.

Conviction. 3.

Side won: bear. First call: true. Reversal: false.

Direction & instrument

FLAT. No instrument. No equity, no long calls, no put-write, no spread, no pair. The bull's pivot to a Jul-17 $245/$270 debit call spread was the only structured expression considered in the debate and it has an unrebutted structural defect: Jul-17 expires 2026-07-17, the Q2 print is 2026-07-22. The spread fully consumes its time value before the catalyst it is implicitly betting on resolves. Rolling to Aug-15 or Sep-19 crosses the print but pays through the nose for the IV (IVP-77) that prices the binary — putting 90-100% of premium at risk on a consulting miss, which is the binary the structure was supposed to hedge.

There is no clean trade expression that converts the bull's fundamental case into risk-controlled exposure at this spot, on this horizon, at this concentration. The honest answer is zero.

Entry

• Method: none
• Price: n/a
• Validity: n/a

What this trader is explicitly not doing on the entry side:

• Not buying spot at $249.38 (chase risk: %B=1.25, RSI 63.8 after a 26-point five-session surge, measured wedge target hit intraday).
• Not setting a $241 limit "in case it pulls back" — the verdict is NO-TRADE, not "weak limit-LONG." A resting limit is a position.
• Not initiating the Jul-17 $245/$270 call spread the bull proposed — structural defect, expires before the catalyst.
• Not initiating a longer-dated spread (Aug-15 / Sep-19) to cross the print — IVP-77 makes this 90-100%-of-premium-at-risk on a binary the system has no edge calling.
• Not selling puts to "get paid to wait" — selling vol into IVP-77 on a binary-gated name is the wrong side of the IV cone.

Stop

• Price: n/a (no position to protect)
• Type: n/a
• Justification: Size zero needs no stop. The relevant exits are the re-engagement triggers and the invalidation conditions below, not a price stop.

Targets

n/a — no position. Reference levels (for journal continuity and the next /decide call) are captured under "Reference levels" below.

Horizon

0 trading days for the trade. Next review 2026-07-22 (post Q2 earnings), with earlier review triggered only by the named observables below. Standard 21-day default does not apply: a NO-TRADE plan does not have a holding horizon, it has a re-evaluation window.

Size

0.0% NAV. $0. 0 shares.

Position-sizing skill not invoked — input requires a non-zero direction. The size is the verdict.

For the journal's sizing context (so the next call can see the slot that was available):

• Notional slot considered: 2.0-2.5% NAV (capped by US-tech basket concentration; the GOOG decision on 2026-05-20 set this ceiling for any fifth correlated US-tech leg).
• Slot status: returned to dry powder. Available for redeployment to a non-HYG-correlated name.
• The conviction-3 NO-TRADE here is "wait, not abstain." Conviction 3 is not license to deploy a fractional exploratory position. The research-manager's instruction to the trader is explicit: "Do not interpret conviction 3 as license to sketch a small exploratory position."

Critical assumptions

These mirror the verdict's assumptions. If any breaks before 2026-07-22, the trade plan re-opens.

1. Jul-22 Q2 is in fact binary for the trade horizon. Q2 EPS breadth 5 up / 9 down (-0.24), Q1 -22% drawdown on +5.4% EPS beat, consulting +1% CC as gating variable.
2. HYG stays in the 78.0-80.4 band. Below 78.0 the basket-wide drawdown is the modal outcome and NO-TRADE strengthens (and one of the four existing longs likely takes a stop, opening the concentration gate). Above 80.4 the dominant macro factor flips constructive and the verdict is reconsidered.
3. The US-tech basket (~15.1% NAV) stays intact. If any of MSFT, AAPL, AMZN, GOOG closes, gets stopped, or is sized down meaningfully, the concentration gate opens.
4. First-call discipline holds. Zero closed directional outcomes; 4-for-4 NO-TRADE discipline on binary-gated holds (DTE.DE, EOAN.DE, BOSS.DE, BAYN). Until at least three directional positions close, binary-avoidance is the only edge this system has empirically earned.
5. The breakout holds structurally. IBM does not close below $233 before Jul-22.

Thesis-invalidating events (would change the NO-TRADE call before Jul-22)

These are the re-engagement triggers. Any one of them materialises → re-open the debate, do not enter unilaterally.

Macro de-risk trigger (cleanest)

• HYG > 80.4 on a daily close (currently ~79.2). Flips the dominant macro factor. This is the Bear's own Trigger A1 and is the cheapest single piece of incremental information.

Tape confirmation trigger

• OBV 20-day slope flips positive over 5+ consecutive sessions. Currently -55,853 units/day. The technical-analyst's stated conditional. If OBV sustains a positive slope through end of next week, the distribution pattern is broken and the breakout becomes structural rather than a one-day event.

Price-action trigger

• IBM pulls back to $240-243 HVN, holds it on a closing basis, prints a daily close back above $245. This is the Bear's Trigger B5 — the cleanest single technical setup for post-print re-engagement, but if it appears pre-print combined with one of the other triggers, it is also a valid earlier re-entry signal.

Concentration relief trigger

• One of MSFT / AAPL / AMZN / GOOG closes, stops out, or is sized down meaningfully. This opens the Bear's concentration gate (Trigger C). IBM becomes a different portfolio question because HYG correlation amplification stops being the binding constraint.

Earnings-resolution trigger (the modal path)

• Post-2026-07-22: consulting prints ≥ +3% CC growth (or the equivalent guide-up at the BofA / Goldman TMT mid-cycle conferences in June). The consulting binary partially resolves; the bull's thesis transitions from "right idea, wrong time" to "right idea, market-priced setup." Re-engage at whatever price the market sets, sized inside the concentration cap.

Hard bearish flip triggers (would make NO-TRADE stronger or invite a short)

• HYG < 78.0 daily close. Bull's own Round-2 kill-switch. Basket drawdown becomes the modal scenario. Bring forward the next review immediately.
• IBM closes below $233. Bull's defended stop level. The breakout is dead, the bouncing_in_downtrend regime resumes, and pre-print re-engagement is moot.
• IBM closes below $221.73 (swing low 2026-04-23). The bounce has failed entirely; the question becomes whether the short side is worth re-debating.

Reference levels (for journal and the next /decide call)

What I am explicitly not doing

• Not chasing $249. %B 1.25, RSI 63.8 from 37.9 five sessions ago, measured wedge target hit intraday. Mechanical exhaustion signals on top of a binary-gated catalyst calendar.
• Not initiating the Jul-17 $245/$270 debit call spread. Expires five days before the Q2 print. The structural defect that decided the debate.
• Not rolling that spread to Aug-15 or Sep-19. IVP-77 makes the premium prohibitive; you would be paying full vol-of-binary to cross the binary the structure was supposed to hedge.
• Not selling puts. Wrong side of the IV cone on a binary-gated name; you collect a small premium in exchange for the exact tail the bear's case sketches.
• Not sizing up if the bull arguments resurface without a regime confirmation. The bull's thesis (margin expansion, $1B CHIPS quantum grant, breadth +0.43 on FY, breakout volume) is real. None of it changes the binary inside the trade window. Wanting to express the thesis is not the same as having a place to express it.
• Not initiating any pair / hedged-LONG / hedged-SHORT structure. No paired short candidate clears the concentration and HYG-beta filter; no put-protected long fixes the IVP-77 cost. The HEDGED flavours of the verdict do not unlock a defensible expression here.

Next review

• Hard date: 2026-07-22 (Q2 earnings — consulting CC growth print). This is the primary review window. Re-run /decide on IBM the morning after the print with the consulting number known.
• Earlier triggers (run /decide on IBM the same day):
  • HYG daily close > 80.4 (re-engagement)
  • HYG daily close < 78.0 (bear's downside trigger — reassess the entire basket, not just IBM)
  • IBM daily close < $233 (breakout failed — reassess pre-print)
  • One of MSFT / AAPL / AMZN / GOOG closes / stops / is sized down (concentration gate opens — IBM becomes a different question)
  • Confluent acquisition disclosure with $1B+ high-growth ARR pre-print (Bear's own LONG-flip trigger; partially detaches trade from consulting binary)
  • DoC formalises the CHIPS Act quantum grant pre-print (removes one of five probabilities in the bear's compound ledger)

Until then: flat IBM. Dry powder preserved. The 4-for-4 binary-avoidance discipline is the only edge this system has empirically earned. Honour it.