Phase Space AI

Trade Construction

Broadcom [AVGO]

Broadcom Inc. [AVGO] — Trade Construction

Spot $380.46 (2026-07-28) · 12-month target $603 (+58.5%) · bear $293 (−23.0%)

The memo issues no position verdict. This document sizes and constructs the expression if the book elects to take it, and states what would make it uninvestable.


Liquidity Criteria — chain pulled live, not assumed

criteria.md: "Any proposed options structure requires the actual chain pulled first — open interest and quoted size for the specific strikes and expiry." Pulled from Alpaca 2026-07-28/29.

16-Jan-2027 calls (172 days), IV 53–55%

Strike Open interest Bid Ask Bid size Ask size Delta IV
380 1,803 49.29 52.92 22 11 0.555 0.540
400 7,202 43.66 45.00 1 116 0.502 0.547
420 3,431 35.45 38.22 32 64 0.447 0.538
450 4,935 27.51 29.82 46 62 0.373 0.537
500 9,437 17.77 19.05 1 102 0.269 0.533
550 2,131 11.33 12.37 28 58 0.189 0.533
580 937 8.60 10.04 40 37 0.155 0.537

16-Oct-2026 calls (80 days), IV 55–59%

Strike Open interest Bid Ask Delta
400 3,856 25.54 27.43 0.433
410 16,694 22.53 23.33 0.394
440 27,100 14.49 16.32 0.294
470 4,028 9.02 10.81 0.210

Liquidity Criteria: PASS. Every strike from $350 to $580 in January-2027 carries four-figure open interest with two-sided quoted size. Bid/ask on the $400 strike is $1.34 wide on a $44 premium — 3.0%. This is the opposite of the HCA case (maximum 18 contracts across an entire chain). The equity itself trades a $1.81tn market cap; size is not a constraint at any book-relevant scale.


Vehicle selection

Three candidates, judged against what the thesis actually is.

1. Long equity. The thesis is a contracted revenue ramp visible over four to eight quarters, not a single dated event. Time works for the holder: each quarterly filing re-discloses committed RPO, so the thesis is re-tested every ninety days at no cost. Equity captures that without paying decay.

2. Jan-2027 $420 calls (delta 0.447, ask $38.22). Captures the $603 target with 2.2x notional leverage per dollar of premium. Breakeven $458.22 — 20.4% above spot, and below the median-multiple target of $458. Costs 10.0% of notional for 172 days. This is the aggressive expression of the same view.

3. Jan-2027 $400/$500 call spread. Long $400 at $45.00, short $500 at $17.77 → net debit $27.23, maximum value $100, maximum payoff 3.67x. Breakeven $427.23 (+12.3%). Caps out at $500, which is below the base target of $603 — so the spread monetises the base case fully and forfeits the bull case. Open interest 7,202 and 9,437 on the two legs; both fillable.

Recommended: long equity as the core, with the Jan-2027 $400/$500 call spread as the incremental expression if the book wants convexity without paying 10% of notional in premium. Rationale: the mechanism is contracted rather than event-driven, so the primary risk is duration of the ramp, not timing of a catalyst — and a defined-risk spread whose cap sits below the base target is the honest way to buy convexity on a thesis you expect to grind rather than gap.

Reject the naked $420 call as the core: at 53% IV a 172-day option needs +20.4% just to break even, and the mechanism does not promise a gap.


Sizing

Inverse-volatility, per the framework's interim protection on the Downside Criteria.

AVGO AMD MRVL
252-day realised vol 47.2% 70.0% 76.9%
Jan-2027 ATM implied vol ~54% ~74% ~94%
Inverse-vol relative weight (realised) 1.00 0.67 0.61

Broadcom is the lowest-volatility name in the cluster and simultaneously the only one that passes the Valuation Criteria. Inverse-vol sizing therefore points the same way the analysis does, which is unusual and worth noting: normally the protection fights the conviction.

Concentration constraint, and it is the binding one: AVGO, AMD and MRVL are the same trade. All three are fabless AI silicon designers selling into the same handful of hyperscaler capex budgets; MRVL's own 10-K names AMD and Broadcom as direct competitors. They must be sized as one cluster exposure, not three positions. Correlation of the underlying revenue driver is close to 1 even where the equities diverge. portfolio_book.json limits apply at the cluster level and are the book's call, not the memo's.

Entry: momentum is the weakest of the three (12-1 = +23.7% versus AMD +199.6% and MRVL +249.7%). Per criteria.md momentum governs when, never whether. Here it argues for building rather than chasing — the fundamentally strongest name in the cluster is also the one the market has re-rated least, which is a reason to scale in rather than to wait.


Exits and invalidation

Trigger Action Rationale
Committed RPO declines sequentially Exit The mechanism is the RPO. If it falls, the PASS WITH ARGUMENT collapses to FAIL — the argument was the RPO.
12-month RPO conversion % falls below 25% Halve Conversion has been 35% → 33% → 30%; a further step down means the book is lengthening, not converting.
Largest-distributor share exceeds 50% of revenue, or AR concentration exceeds 55% Halve The credit exposure starts to dominate the contracted-revenue argument.
Non-AI semiconductor revenue turns negative YoY Reduce Removes the annuity that justifies any part of the multiple outside AI.
Infrastructure software revenue declines YoY Reduce At +1% (Q1 FY26) and +9% (Q2) there is no cushion.
Price above $713 (own-history 90th-percentile EV/Sales on NTM) Trim Above that the multiple, not the business, is doing the work.
Price below $293 without an RPO deterioration Add That is the p25-multiple level on unchanged fundamentals.

Stop discipline: there is no price-based stop. The invalidation is a disclosure, and it arrives four times a year on a known date. That is a better stop than a percentage, and it is the reason the equity core is preferred to a dated option.


Ledger entry

To be appended to trade_recommendations.jsonl if the book takes the position. Recorded here so the recommendation is falsifiable either way: