To The Moon

$AMD: Shipments have started. Gross margin is the real test. My sell put pick.

The two days

• Thursday: −17.26% to $404.40 in the AI chip selloff. Intraday low $402.10.

• Friday: +11.04% to $449.04 on 45.7M shares, about double the ~22M average. No AMD-specific news; the whole chip sector bounced (SOX +4.2%).

• AMD is still 30.9% below its record of $649.42.

The topic asks what would confirm this is more than an oversold bounce. My answer goes through all three candidates.

1. MI450 shipments: already happening

On the Nov 3 call, management said Q3 was "the first quarter that we shipped MI455 and Helios racks for revenue." Volume shipments went to customers including Anthropic, Meta, Microsoft, OpenAI and Oracle. Management also said: "We have a big step up in Q4. We'll continue to ramp throughout 2027." So shipments have started. What still needs proving is how fast the ramp goes.

2. Data center growth: already strong

• Data center revenue $8.5B, up 97%. It's now 64% of total sales.

• Data center operating margin 34%, against 25% a year ago

• Management expects data center to be more than 80% of revenue in 2027

3. Gross margin: the real test

• Q3 non-GAAP gross margin was 56%, up 190 basis points

• Q4 guidance is ~55%, slightly below the 55.3% estimate

• AI rack systems carry lower margins than the company average, so margin comes under pressure as Helios scales

If Q4 revenue lands near the $17B guidance (about +27% over Q3) and margin holds around 55%, the ramp is profitable and the bounce has a base. If margin slips below that, Friday was just short-covering.

The quarter itself

• Revenue $13.4B, up 45%, above the top of AMD's own guidance

• EPS $2.09, up 74%, against ~$1.95 expected

• Q4 guidance ~$17B ±3%, against a $16.2B estimate

• The CFO says operating costs should grow at less than half the rate of revenue in 2027

The risks

• Gaming fell 40%, and Intel warned that PC weakness could run into 2027

• After the midterms there are state ballot fights over building data centres. UBS estimates about 30% of expected 2028 US capacity additions could be delayed.

• Valuation: ~98× trailing earnings and ~30× 2027 earnings. A stock priced like this swings both ways.

Checking it against my rules

✅ Trend: $449.04 vs 200-day $415.98, +7.9%. Last week my trigger was a close back above the 200-day. Friday did it.

✅ Earnings: next report around early February 2027 (estimated), after expiry

✅ IV: 30-day IV 75, against a 1-year range of 42–89, so IV rank is about 70

⚠️ IV/HV: about 1.15, just under my 1.2 bar. Close, but not a clean pass.

Strike math (Dec 18 expiry, 39 days)

• 1 standard deviation down ≈ $346

• 200-day × 0.98 ≈ $407.7

• $345 has delta 0.10, too far out to be worth selling, so I move up to $370 (delta 0.16)

The pick

• Sell 1× AMD Dec 18 $370 put

• Model premium ≈ $9.90 at 70% IV

• Break-even ≈ $360, about 17.6% below Friday's close

• $370 is 8% below Thursday's panic low of $402.10

• About 77% model odds of expiring worthless

• Cash needed: $37,000 per contract. Tier 2, so 2% max.

• Take profit: GTC buy-back at ~$4.95 (50%)

• Stop: ~$24.80 (2.5× the credit), one roll at most

The verdict: the trigger has been met, so I'm selling the fear at $370. What I'm watching is gross margin. Revenue tells you demand is there; margin tells you whether it's profitable.

Personal framework, education only, not advice. The premium is a Black-Scholes estimate, not a live quote, so check the chain in Tiger before trading. Options carry risk, including assignment.

#AMD #MI450 #SellPut #Options #WheelStrategy