The call — WATCH Tesla into tonight's Q3 print. Do not chase $387 (+3.6%). Buy a guide-miss washout at $345–360 for $425–440, kill on daily close <$325. 1–3 months. Conviction: WATCH.
Why now — SpaceX just beat huge ($14.58B rev, +161%) and Musk confirmed Tesla–SpaceX merger committees. That sympathy lift is masking a weak auto print: Street at ~$27.9B rev, ~$0.44 EPS, down ~10% YoY.
The disagreement — Consensus treats SpaceX AI as Tesla AI. It isn't. Starlink ($2.37B op income) pays the bills; Tesla auto still has to prove margins, robotaxi scale, and cash burn against a 388× trailing P/E.
The level that changes everything — $325 daily close kills the thesis. $415 (+7.3% implied move) is where a spike becomes a fade, not a chase.
Tonight's Tesla Q3 print (after the bell, Wednesday Oct 21) is the market's real test, not last night's SpaceX party. SpaceX (SPCX) closed the loop: revenue $14.58B vs $12.72B expected, adjusted EPS $0.45 vs $0.13, operating income $2.8B and net $4.4B vs a $1.05B loss a year ago, per CNBC's live earnings coverage. Starlink subs hit 13.6M, AI ("Super Intelligence") did $7.56B — nine-fold YoY and bigger than Starlink for the first time — and the CFO guided Q1 2027 to $33–37B vs Street $22.2B. SPCX is $163.14, +10.4% intraday Oct 21 (16:00 UTC pull).
Tesla rode it to $387.11, +3.6% on the day. My opinion: that is borrowed strength. Tesla enters with deliveries 486,532 (beat, but -2.1% YoY and only +1.3% QoQ), energy 13.7 GWh (+9.6% but light), and expectations for flat revenue around $27.9–28.1B with EPS ~$0.40–0.45, down ~10% YoY — see Electrek's preview, tastylive's robotaxi-vs-margin frame, and Tech Insider's three things to watch. Q2 already missed on profit. The Schwab market update this morning said it plainly: SpaceX impressed while Meta and Tesla wait in the wings — and the tape is fragile.
What to do: no chase into the print. Options price ±6–7.5% (sources cluster $38 / 7.3% / 6.9%). At $387 that is $360–$415. Let the number set the entry.
Musk confirmed both boards formed independent special committees to explore a Tesla–SpaceX merger — and excluded himself as conflicted — but stressed it needs 50% of all Tesla shares, a very high hurdle. Details are well covered in Al Jazeera's summary and CNBC's call blog linked above.
My take, marked as opinion: the industrial logic is genuine — shared autonomy stack, Megapacks ($430M bought by SpaceX AI last year), Cybertrucks ($143M), Roadster thrusters co-developed with SpaceX, and now Terafab (Texas AI-chip fab with Tesla + Intel targeting volume in Q3 2028). But none of that helps Q3 auto gross margin or Q4 deliveries. A $3T+ combination would be the largest merger ever and will take quarters of negotiation, shareholder votes, and regulatory review. Trading tonight's auto print on merger euphoria is category error.
Second-order effect most are missing: if the merger advances, Tesla holders effectively underwrite SpaceX's AI capex ($19.6B in Q3 alone, $16.4B to AI; Q2 debt + leases $36.8B, up from $22B). That cuts both ways — access to Starlink cash flow ($2.37B quarterly op income) helps fund robotaxi/Optimus, but Tesla's already-negative free-cash-flow + $25B+ 2026 capex plan gets heavier, not lighter.
At $387, Tesla is ~$1.2–1.5T market cap on $1.08 trailing EPS — a ~388× trailing P/E, ~389× forward, with 3-year EPS growth -33% and ROE 4.6% (financial-handler pull, Oct 21). Gross margin 18.9% is the number to watch tonight; bears will pounce if auto gross ex-credits slips again despite the delivery beat.
The bull case arithmetic: deliveries stabilized (486k), energy storage scaling (13.7 GWh), and Q2's "best Q2 ever" energy print showed a second engine exists. Add Starlink-like recurring revenue if robotaxi moves from 3 Florida cities (July launch) to real paid miles + Cybercab production + Optimus milestones, and a $400+ stock can be defended on 2028–2030 earnings power.
The bear arithmetic, and in my opinion the load-bearing assumption: ~$960B of the cap — roughly two-thirds — is payment for FSD + robotaxi + Optimus that have produced negligible profit to date. If Q3 auto margin disappoints and the call offers no dated robotaxi expansion plan, no Cybercab ramp numbers, and no Optimus unit-economics, that premium compresses fast. Morningstar's long-standing "rich by any standard" critique (234× 2026 earnings on a bull's own number) still rhymes.
Competitive reality: BYD and Chinese rivals press price/mix globally; Waymo presses credibility on autonomy safety/scale; Fluence and Chinese LFP press energy storage pricing. Tesla's edge must show up in cost per mile and cost per kWh, not event demos.
Do not buy $385–390 into a flat-revenue, down-EPS print with a 388× multiple. The prior open thesis (TSLA into Oct 21 alongside Meta) assumed ROI proof — Q3 must deliver it. Netflix last night is the cautionary tale: EPS beat ($0.87 vs $0.82) but rev in line ($12.87B), Q4 guide $13.44B ~0.5% light and 2026 ads cut $3.0B → $2.7B, stock $65.40 -4.8% intraday, sitting right on our EARN-NFLX-1 kill (weekly close <$65). Guide misses are being punished, beats forgiven only if guidance holds.
| Flagship | Detail |
|---|---|
| Ticker / Direction | TSLA — tactical BUY THE WASHOUT (not pre-print) |
| Entry | $345–360 (post-print dip to pre-SpaceX base + options lower bound) |
| Target | $425–440; stretch $470 only on dated robotaxi/Cybercab + merger term sheet |
| Invalidation | Daily close below $325 (breaks July-base structure + confirms margin thesis broken) |
| Timeframe | 1–3 months (through Q4 deliveries + FOMC Oct 27–28 + year-end) |
| Conviction | WATCH (levels set, no trigger — trigger is the print) |
| Sizing (opinion) | Half at $355–360, half on hold >$365 with margin stabilization; no leverage into the gap |
If Tesla spikes on a headline beat without margin proof: fade $415–425 (upper implied-move band), same $325 kill, back toward $380. Never set the invalidation inside the base case — $325 is below both bull and base washes; a $360 stop would fire on normal post-earnings noise.
Base (50%): flat print, soft guide — $350–365 wash, then stabilize. Revenue $27.7–28.1B, EPS $0.40–0.45, auto gross flat, energy in line, vague robotaxi ("expanding to X cities soon") and no merger terms. Stock dips 5–7% then finds $345–355 — the buy zone. This is why we wait.
Bull (25%): margin + roadmap — squeeze to $425–440. Trigger: auto gross ex-credits up QoQ plus one dated commitment (e.g., Cybercab production rate, 5+ city robotaxi list, or Optimus pilot revenue). Merger committee releasing a framework letter would add fuel. Probability capped because Q2 missed and pricing/mix headwinds persist.
Bear (25%): guide cut or cash-burn shock — $325 test, $300 open. Trigger: Q4 auto gross guide down, 2026 capex lifted above $28B with negative FCF deepening, or robotaxi delay language. A daily close <$325 ends the tactical long and re-opens July low $297.
Open-call check: no conflict. TACT-CRWV-1 ($58–65 zone, kill $55) benefits if SpaceX AI validates neocloud pricing; LONG-AMZN-1 ($250–263 → $300) and WATCH-AMD-1 ($538 → $720) ride the same AI-capex proof — a Tesla/AI-cusp disappointment would pressure all three, so size Tesla as the single expression, not an add. BTC fade (TACT-BTC-FADE-1 short $82.5–83.5k) is already broken — BTC $84,506 (+2.4%) with daily close above $84k kill — acknowledge the loss, do not average. NFLX EARN-NFLX-1 ($67–71 → $85, kill weekly <$65) is on life support at $65.40; honor the weekly-close rule.
Trader (1–2 weeks): flat into the bell. Bid $350–355, offer $418–425. Use defined risk (verticals); IV crush will kill naked premium.
Core investor (1–3 months): half-starter only on $345–360 wash with margin stabilization. Add second half on daily close >$390 confirming the $325 floor held. Target $425–440 into Q4 deliveries.
Long-term holder: hold through noise unless daily close <$325. The merger + Terafab + robotaxi story is 2027–2028; tonight only matters if it breaks the balance sheet or the autonomy timeline.
Opinion is mine; facts timestamped below. If you only read the TLDR and this table, you have the thesis and the action.
Contents: A. Data snapshot · B. Model & assumptions · C. Sources