The Supply Calendar

This is the part of SpaceX that is knowable in advance, and it is the part both closed campaigns on this name left unreconciled.

The offering priced on June 11 and listed June 12. SEC filing: “We are offering 555,555,555 shares of our Class A common stock” SEC filing: “The initial public offering price is $135.00 per share” That is a gross raise of about $75.0 billion, and a free float of roughly 4.2% of the 13.08 billion shares outstanding after the offering. Everything else was restricted — and the prospectus then wrote a release schedule instead of a cliff. Every date below is counted from the prospectus date of June 11.

The filing's own words, tranche by tranche: SEC filing: “Up to 911.5 million shares of Class A common stock” SEC filing: “Up to 455.8 million additional shares of Class A common stock” SEC filing: “Up to 319.0 million shares of Class A common stock, representing 7% of the shares subject to the 180-day lock-up period” SEC filing: “Up to 328.4 million shares of Class A common stock, representing 7% of the shares subject to the 180-day lock-up period” SEC filing: “Up to 1.3 billion shares of Class A common stock, representing 28% of the shares subject to the 180-day lock-up period” SEC filing: “up to 797.6 million shares if the Additional Release Shares were not released on the First Earnings Release Date” SEC filing: “Up to 6.4 billion shares of Class A common stock, representing all shares held by Mr. Musk”

Three readings matter here.

The conditional tranche lapsed, which makes December bigger. The extra 455.8 million shares released only if the closing price had been at least 30% above the offer price — $175.50 — on five of the ten trading days through the first earnings release. SEC filing: “is at least 30% greater than the public offering price set forth on the cover page of this prospectus for at least five of the ten consecutive trading days ending on, and including, the First Earnings Release Date” None of the ten closes from July 22 through August 4 came near it: every one of them was below the $135.00 offer price, and the highest was the August 4 close of $125.33. The condition was not met, so those shares stayed in the pool and December 8 carries the larger of its two numbers — 797.6 million rather than 328.4 million.

The two biggest releases of the whole schedule are still ahead, and one of them lands on an earnings date. The third-quarter release is 28% of the 180-day pool in a single step, and it is triggered by the report rather than by the calendar — the desk's event calendar carries the print for November 3, which would put the release around November 5. Adding the remaining tranches, roughly 3.1 billion shares become eligible between September 24 and December 8. Against 77.6 million shares of average daily volume that is about 40 days of turnover arriving over eleven weeks — and about 16.8 days of it in the single November step. Eligible is not the same as sold; the schedule sets the ceiling, not the flow.

After December 8 the overhang changes character rather than ending. SEC filing: “an aggregate of 7.8 billion shares (including 100% of the shares owned by Mr. Musk), representing approximately 60% of our shares outstanding” The extended pool — roughly 1.76 billion shares, judging from its own 20% steps — does not begin releasing until the fourth-quarter report, and the founder's stake is explicitly exempt from every early-release provision until June 12, 2027. A first-half-2027 calendar of 351.9 million, 176.0 million, 351.9 million, 176.0 million and then 6.4 billion is a second, slower wave rather than a cliff, but the June 2027 date is the single largest share-supply event on any calendar the desk tracks.

Two other supply channels sit outside the lock-up table. The company issued shares for the Cursor merger on August 14: SEC filing: “an aggregate of 389,289,254 shares of the Company’s Class A common stock” SEC filing: “based on an implied equity value of Cursor of $60.0 billion” — about 3.0% of the shares outstanding, issued unregistered in reliance on the private-placement exemption, which the filing states plainly and about which it says nothing further. It does not say whether those shares carry a lock-up, and no resale registration statement for them appears in the company's filing index through September 17; as unregistered restricted stock they would need either a registration or a Rule 144 exemption to reach the market. Separately, the company registered the resale of shares issued under its historical equity-compensation plans on Form S-8 on August 4, which the prospectus says makes those shares immediately eligible for resale subject to affiliate limits and any lock-up still binding them.

Index membership is the one supply-side fact this dive cannot source properly. A capture dated September 2 lists SPCX in the Invesco QQQ at a 1.21% weight, which if accurate means passive funds are a standing bid against the release schedule. That is an attributed social post, not an exchange notice: the inclusion date and official weight remain unreconciled, and no primary source for either is on file.

Why This Moment Is Different

The stock's first three months were a float story, not an earnings story. A 4.2% float met a $1.77 trillion valuation at the offer price, ran about 67% above that price inside three weeks, and then sold off for six straight weeks into the first two release dates. The first earnings report, on August 4, arrived on the day after the low and beat: the stock closed +9.43% that session at $125.33 and then fell 13.61% on 2.97× volume the next day. That pair of sessions is the cleanest evidence available that this tape is being driven by share supply rather than by the P&L.

What is different now is that the calendar is public and the stock has stopped making lower lows into it. The shelf at $147.62 formed after the September 9 and September 10 releases, not before them — the market absorbed 378 million newly eligible shares without breaking the rebuild. That is the most encouraging structural fact in this dive, and it is also the reason the November step matters: it is about three and a half times larger than what has just been absorbed.

What the Company Is and How It Makes Money

The prospectus describes a vertically integrated company across three businesses. Launch: SpaceX says it has launched more than 80% of the world's mass to orbit each year since 2023, with an over 99% mission success rate on Falcon. Connectivity: roughly 9,600 Starlink satellites in low-Earth orbit serving consumer, enterprise and government customers across 164 countries and territories as of March 31, 2026, plus a direct-to-mobile constellation covering about 30 countries. AI: xAI, founded in 2023 and acquired by SpaceX in early 2026, brings the Grok model, the Colossus data-center build and the X platform, which the filing frames as the distribution and data engine for the AI stack.

The stated growth path is explicit about its dependency. The company's first filed risk factor is that any failure or delay in developing Starship at scale, or in achieving the required launch cadence and reusability, would delay the next-generation satellites, the satellite-to-mobile service and orbital AI compute. Regulatory risk comes next: FAA launch and reentry licences, then FCC and international spectrum authorizations, then the evolving legal regime around AI products and the X platform. Then the one that matters most for a reader looking at the capital-spending line: the strategy depends on deploying products, platforms and infrastructure at unprecedented scale, which the company itself labels a significant execution, cost and timing risk.

The mix is where the disclosure is thinnest. The quarterly XBRL facts carry one consolidated top line; the desk has no filed segment split for launch, Starlink and AI compute.

The widely circulated figures for the June quarter — AI compute up 247% to $2.6 billion, Starlink subscribers doubling to 12.0 million, backlog of $47.5 billion — come from the desk's social capture of the call, with the basis unstated, not from a filing. They are used here as context and nothing is built on them.

What the Street and the Public Conversation Say

The fresh news pull on the evening of September 17 returned quote pages, a stock-screener earnings page and the August merger item. Nothing in it explains the move. The September 16 gain of +5.15% and Thursday's +2.6% therefore carry no cause on file, and the honest answer is cause unknown.

One thing the settled tape does say: Thursday was a lane-wide day, not a SpaceX day. Across the space list, RKLB closed +6.47%, FLY +7.10%, RDW +7.04%, LUNR +6.56%, PL +6.05% and ASTS +5.80%. SPCX was among the smallest gainers in its own group.

Three leads circulating on quote-page aggregators do not survive contact with the desk's receipts, and are killed here by name. A William Blair Outperform rating after a new compute deal: no receipt on file — the only appearance of that firm in any SpaceX document here is as one of the IPO's underwriters on the prospectus cover. A reported ~$10 billion Technip Energies contract: no receipt on file, nothing in any capture, filing or news pull. Insider purchases: no receipt on file. The only ownership filing in the last two weeks is a September 15 Form 4 from director Antonio Gracias covering 42,790,223 shares under transaction code J, reported as of September 11. Code J is the catch-all "other" category — it is neither an open-market purchase nor an open-market sale, and the desk's capture carries no dollar value for it. Nothing about it supports an insider-buying story.

The positioning color is worth one line, attributed. The highest-ranked item in the social sweep is a July piece headlined that SPCX is Wall Street's most shorted new stock, and a CNBC item from the week of the IPO argued that passive investors who avoided bitcoin were being forced to own a stock three times as volatile. The filed short-interest series is more useful than either: 111.3 million shares at the June 30 settlement, 207.8 million at July 31, 184.4 million at August 14 and 173.6 million at August 31, at 2.36 days to cover. Short interest peaked into the July decline and has come down through the rebuild. The percentage of float is withheld by the desk's own data because the float receipt is missing or stale — with a float this small and this fast-moving, a float-based short percentage would be the wrong number to print.

The most substantive third-party read on the quarter came from SemiAnalysis, which wrote after the August 4 call that Musk put 6-8GW of incremental data centers on the 2027 plan, said it believed the number, and sketched its own path to a $300 billion annualized run rate. That is an attributed forecast from an outside research shop, not company guidance the desk has verified, and it is the single most aggressive claim in this dive's evidence pile.

The recurring Tesla-merger speculation has no source this desk can cite, so it is not treated as a claim here. What the filings do establish is the structural fact underneath the speculation — the founder controls approximately 82.4% of the voting power through a ten-vote Class B, SpaceX is a controlled company under Nasdaq rules and relies on the associated governance exemptions, and minority holders have no vote that could stop a related-party combination. That is the risk, and it is documented; the merger rumor is not.

Grading It Against the Written Rules

This section grades the name against the desk's written playbook rather than making a recommendation. Here is the test-by-test result on the settled close.

Lane test — passes. Trend-hold trades durable secular-uptrend names, explicitly not deep-value or falling-knife bargains. SpaceX is a real-revenue operating company in two lanes the desk already tracks. It is not a value trap and not a pre-revenue lottery ticket. It is tradeable: the registry marks it a US Nasdaq common share with normal trade access.

Archetype test — fails all five, and that is the finding. Secular compounder requires strong-up, a golden cross and repeated new highs with positive one-year alpha: there is no 200-day average, no cross, and no one-year figure at all. Cyclical-at-peak requires a parabolic three-month move: this one is down 16.3% over three months. Quality dip requires the golden cross intact above a rising 200-day: neither line exists. Broken downtrend requires a death cross below a falling 200-day: same problem, and the trend reads neutral rather than strong-down. Meme/squeeze requires a 3-10× relative-volume spike: volume is 1.08× average. The playbook's own footnote names the bucket this belongs in — the sixth, special-situation and event-driven, which covers post-IPO lock-up names and says plainly that they trade off the event and the date, not the chart's technical rules. That is the correct classification for SPCX today, and it is why this dive is a calendar rather than a chart.

Sequence test — between pop and flag, with no breakout. The written path is bottom → pop → flag → breakout. The bottom was August 3. The pop was the post-report rally. What has formed since is a range with a volume shelf at $147.62 and the most-traded price at $152.65, which is a flag at best. The breakout has not happened. Trend-hold's first entry rule is unambiguous: no pre-confirmation buys, wait for the breakout to confirm, buying before it is gambling.

Box test — base-buy, alert only. The applicable box is base-buy, whose entry rule reads: a coiled base is research, set the range-high alert, there is no trade yet, and a confirmed base breakout permits a starter position. The alert level is $156.87, the highest price of the trailing 23 sessions and the only range boundary the current base has produced. The box's own invalidation is a failure back inside the base, which here is a close below the $147.62 shelf. The desk's advisory check returns waiting under this box with a compounder archetype declared, and research-only with no archetype declared — in both cases, no actionable entry.

The rule conflict, stated plainly. A scheduled single-name accumulation plan in SPCX is not covered by any box in the written playbook. The scheduled-buy box is written for broad, durable quality bought on schedule, where the point is adherence rather than tape timing; it is a program box, not a single-name box, and it maps to a mechanical paper persona rather than to the discretionary doctrine. Trend-hold's first entry rule forbids buying ahead of confirmation. Base-buy allows an alert and nothing else. Meanwhile the supply calendar hands a rules-follower five dated windows — September 24, October 9, October 26, roughly November 5 and December 8 — on which a name in the event-driven bucket is supposed to be watched rather than accumulated. The rules do not conflict with each other here; they all say the same thing, which is watch the level and the dates. What conflicts is the pull of a $2 trillion story against a written rule that says wait.

What re-classifies it. A 200-day average becomes computable around April 2027, at which point the compounder and quality-dip tests can actually run. Before that, only two things change the read: a confirmed close above $156.87 with the December supply behind it, which starts the breakout leg the doctrine requires, or a loss of $136.14, which ends the rebuild.

The Conviction Chain

1. The lock-up schedule is the dominant price driver through year-end. Evidence: seventeen dated release steps in the prospectus; the June-to-August decline ran straight into the first two; the shelf formed after the September releases were absorbed. Confirms it: price weakness clustering around the September 24, October and November dates. Breaks it: the stock trending through those dates on ordinary volume, which would say the float has found real holders.
2. The business is compounding fast enough to matter. Evidence: revenue +91.9% year over year in the June quarter, gross margin 55.3%, operating loss down from −$1.94 billion to −$143 million in one quarter. Confirms it: a third-quarter report that holds the margin while capital spending stays elevated. Breaks it: a revenue deceleration or a reversal in the operating line.
3. The balance sheet removes financing risk for years, not quarters. Evidence: $93.52 billion of cash plus $6.49 billion of short-term investments at June 30, against $38.29 billion of long-term debt. First-half free cash flow was −$25.01 billion. Confirms it: capital spending funded from cash without new equity. Breaks it: a follow-on offering, which would add supply to a calendar that already has plenty.
4. The tape cannot be graded yet. Evidence: 68 bars, no 200-day, no one-year alpha, no archetype fit, and the desk's own screen refusing to classify it. This is the weakest link in the chain and it is why conviction is low: everything above is a filings judgment, and the chart cannot corroborate or contradict any of it for another six months.

Setup

All levels use the settled September 17 close.

• Entry: none. The base-buy rule permits an alert, not a position. Alert on a confirmed close above the range high, $156.87.
• Second decision level: $180.34, the value area high. Above it the post-IPO range has resolved.
• Setup invalidation: a close back below the support shelf at $147.62 after any breakout.
• Base invalidation: a loss of the anchored VWAP from the August 3 low, $136.14. Below that the six-week rebuild has failed.
• Dated risk windows: September 24, October 9, October 26, roughly November 5 and December 8.
• Current action: watch. A name whose tradeable supply more than doubles over eleven weeks does not need to be anticipated.

Bull Case

1. Revenue nearly doubled year over year at a 55.3% gross margin, with the operating line within $143 million of breakeven after a −$1.94 billion first quarter. Operating leverage is visible in one sequential step.
2. The balance sheet is a fortress. $100.0 billion of cash and short-term investments against $38.29 billion of long-term debt, and a 0.30 debt-to-equity ratio. The build is self-funded for years at the current pace.
3. The September releases were absorbed. 378 million shares became eligible on September 9 and 10 and the stock is higher, not lower, a week later.
4. Three businesses, one integration story. Launch economics fund connectivity, connectivity funds AI compute, and the AI compute has a distribution platform attached. Few companies at this scale own all three layers.
5. Passive demand is structural if the index membership is real. A capture dated September 2 puts SPCX in the QQQ at a 1.21% weight — attributed, not reconciled, but if accurate it is a standing bid that does not care about the release calendar.

Bear Case

1. Roughly 3.1 billion shares become eligible between September 24 and December 8 — about 40 days of average turnover, with 16.8 days of it in one November step.
2. The build consumes cash far faster than the business generates it. Capital spending of $28.48 billion in the first half, against $3.47 billion of operating cash flow, left free cash flow at −$25.01 billion. The cash pile is enormous and so is the draw on it.
3. The valuation prices the story, not the quarter. About 65 times the June quarter's revenue annualized, on a company that lost $541 million in that quarter and $4.82 billion in the first half.
4. The chart is unusable. No 200-day, no cross, no one-year record. Every technical rule the desk owns is either untestable or returns wait.
5. Governance is absolute. Approximately 82.4% of the voting power sits with the founder under a ten-vote share class, and the company relies on Nasdaq's controlled-company exemptions. Minority holders have no vote that matters.
6. Segment disclosure is thin. Consolidated revenue only in the XBRL facts. The AI-compute growth rate everyone quotes is a social capture, not a filed number.

Catalysts

• The third-quarter report, calendared for November 3. It is both the earnings event and the trigger for the largest single release on the schedule.
• September 24, October 9 and October 26 releases. Each is 328.4 million shares; the tape's behavior around them is the cleanest available test of whether supply still sets the price.
• December 8. The 180-day pool empties. After that, the next scheduled supply is the fourth-quarter report in 2027.
• A resale registration for the Cursor shares, or the absence of one. It would convert 389 million restricted shares into tradeable ones.
• Any filed segment disclosure that splits launch, Starlink and AI-compute revenue. It would replace the most-quoted numbers in this name with audited ones.

Risks

• Starship development or launch cadence slips, which the company itself names as the risk that would delay next-generation satellites, direct-to-mobile service and orbital compute together.
• Regulatory: FAA launch and reentry licences, FCC and international spectrum authorizations, and the evolving legal regime around AI products and the X platform.
• Execution at unprecedented scale — the company's own language — across three capital-intensive businesses at once.
• A follow-on equity offering into a calendar that already schedules billions of shares.
• Launch failures and the physical risks of operating in orbit, including constellation proliferation and debris.
• Concentration in one founder who controls the vote, sets the strategy across three companies, and whose stake is locked until June 2027.

Financials

Financials source: the SEC 10-Q filed August 4 for the quarter ended June 30, 2026, the SEC XBRL company facts covering revenue, operating income, net income, EPS, cash, operating cash flow and capital expenditure, and the Massive income-statement and balance-sheet series. Percentages, the free-cash-flow line and the share-tranche sums were computed from those filed values. The January–March 2026 column is not a filed quarter — SpaceX's first 10-Q as a public company covers the June quarter, so those figures are the first half less the June quarter, which is how the Massive series reports them.

Operating series

The June quarter's operating margin was −1.8%, against −41.4% in the March quarter. Two quarters is not a trend, and the first quarter includes whatever the early-2026 xAI acquisition brought with it, so read the improvement as a fact about one step rather than a rate of change.

Balance sheet and cash flow, at or through June 30, 2026

Cash was $24.75 billion at December 31, 2025. The IPO raised approximately $75.0 billion gross in June. The runway question therefore answers itself on filed numbers: at the first half's rate of cash consumption, the cash balance alone covers roughly two years, before any further financing and before the $6.49 billion of short-term investments. No capture on this desk carries the third-party claim that SpaceX has less than a year of runway, and the filed numbers do not support it.

Two provider caveats belong on the record. The Massive statement series flags the March-quarter share count as suspect and non-positive and withholds it, and reports the March-quarter total-liabilities line as negative and suspect; both are held rather than used, and the balance-sheet figures above are the June period-end values. Basic shares in the June quarter were 5,864 million on the filed statement, which is an average for the period and not the 13.08 billion shares outstanding after the offering — the two are different measures and should not be mixed.

Positioning

Short interest as a percentage of float is deliberately absent: the float receipt behind it is missing or stale, and a float-based percentage on a name whose float changes on eleven scheduled dates would be the wrong number.

Research Log

Cross-references

• Both prior SpaceX campaigns closed without reconciling float, lock-up or index dates. This dive closes the lock-up half of that gap from the prospectus itself; index membership remains unreconciled.
• On the settled September 17 tape SPCX was in no monster tier and not on the washout-turn list. It is tracked on the space list and on the orbital-compute buildout list.
• The June 16 entry-zone target for this ticker describes an ETF and a pre-revenue thesis. It refers to a different instrument than this Nasdaq-listed operating company and should not be read alongside this dive.
• The August 4 earnings event and the November 3 scheduled report are the two dates this dive hangs on.

Sources

• SpaceX final IPO prospectus, Form 424(b)(4), dated June 11, 2026 — offering terms, "Shares Eligible for Future Sale", "Lock-Up and Market Standoff Agreements", "Underwriting—Lock-up Agreements", risk factors, business overview
• SpaceX Form 8-K, August 14, 2026 — completion of the Cursor (Anysphere) merger
• SpaceX Form 10-Q for the quarter ended June 30, 2026, filed August 4, 2026
• SpaceX Form 4 filed September 15, 2026 (director Antonio Gracias, code J)
• SemiAnalysis on the first earnings call; Invezz, "the world's most valuable IPO is now Wall Street's most shorted new stock"; CNBC on forced passive ownership; QQQ constituent capture
• Desk receipts: SPCX-37332a4cfc381a4a; SPCX-064f2305a6e22546; income-statements; balance-sheets; short-interest; company; key-financials; submissions; latest; space; 2026-09-17; 2026-09-15-spcx-4-0001213900-26-100186; 2026-09-17-social-search-spcx; 2026-08-12-SemiAnalysis_-elon-shocked-the-world-again-on-spacex-s-first-e-31506158; 2026-09-02-TheETFTracker-here-is-every-stock-currently-in-the-invesco-qqq-60071710; 2026-08-04-spcx-earnings; 2026-11-03-spcx-earnings