Today, August 6, 2026, the first lockup on SpaceX shares expires. On paper, roughly 911.5 million shares held by employees, founders and early backers become eligible to trade. At the current price that is about $123 billion of stock. For scale: SpaceX’s entire public float — the shares the market has actually had to work with since the June 12 listing — stands at roughly 638.9 million shares, the Class A stock placed in the offering after underwriters exercised their over-allotment option in full. The paper coming off restriction today therefore equals about 140 percent of everything that has traded so far. If all of it reached the open market, the float would rise from 638.9 million to roughly 1.55 billion shares — more than doubling.
Even then it would remain a sliver of the company. SpaceX had roughly 13.18 billion Class A and Class B shares outstanding at the end of June, and only about four to five percent of that was freely tradable at listing. That narrowness is precisely why a single release can be so enormous relative to the float — and why market capitalization is a completely misleading yardstick here.
There is no useful precedent. A typical large-cap lockup expiry frees roughly $1.8 billion of stock. Today is about 68 times that. It is also 64 percent larger than the offering itself, which raised around $75 billion. Read those numbers alone and you brace for a collapse.
Which is where it gets interesting, because the supply shock arriving today is substantially smaller than the headline — and the reason is not luck. It is a contract clause that works in exactly the opposite direction from what most investors assume.
What actually unlocks today — and what does not
SpaceX skipped the standard structure. The convention is a cliff: nothing for 180 days, then everything at once. Instead the release was cut into stages. On August 6, insiders may sell no more than 20 percent of their restricted holdings. Further tranches of roughly seven percent each follow around August 21 and again around September 10. The main 180-day lockup does not expire until December 8, 2026.
So the 911.5 million shares in every headline become roughly 180 million shares that may actually be sold today. That is still enormous — close to 29 percent of the entire existing float, which is more stock in a single session than most initial public offerings place in total — but it is not 911 million, and the gap between those two numbers is the gap between a crisis and a demanding trading session.
What can be sold and what will be sold are different things again, and nobody knows the second number in advance. What is knowable: insiders overwhelmingly received their shares far below the $135 offering price, many of them years ago at a small fraction of it. The incentive to sell therefore survives at roughly $108, a price that represents a loss for anyone who bought the IPO. That asymmetry is worth internalizing, because it explains behavior that otherwise looks irrational: an early investor’s gain and a new shareholder’s loss are measured against the same quote.
The clause that did not trigger
The structure includes a second and far larger block: 455.8 million shares worth roughly $62 billion. That block was conditional. It would have been released only if the stock had traded at or above $175.50 — around 30 percent above the offering price — on five out of ten consecutive trading days. A related early-release provision could have freed an additional ten percent before the first earnings report.
Neither condition was met. The stock last closed at $108.27, roughly 38 percent below the trigger. The $62 billion block stays locked.
This mechanism is worth committing to memory precisely because it is counterintuitive. Investors dread lockup expiries because they imagine supply crashing onto a weak market. With a price-contingent schedule the logic inverts: paper is delivered into strength and withheld in weakness. The structure is countercyclical by design. It releases more stock when the market can absorb it and closes the tap when it cannot.
Do not turn that into good news, though. The big block failing to trigger is not an achievement — it is the direct consequence of a stock sitting more than 50 percent below its $225.64 high from its first week of trading and over 30 percent below the offer price. The clause is reporting weakness. It merely softens the mechanical consequences of it.
Why a date everyone knew about still moves prices
This is where the academic record earns its place, because lockup expiries are one of the best documented embarrassments in capital markets theory. The date is printed in the prospectus. It has been public since day one, free, for months. Under any textbook version of a semi-strong efficient market, nothing should happen, because no new information arrives.
Something happens anyway. Laura Casares Field and Gordon Hanka examined 1,948 lockup agreements for The Expiration of IPO Share Lockups in the Journal of Finance (volume 56, 2001, pages 471 to 500). They found a permanent 40 percent increase in average trading volume and a statistically robust three-day abnormal return of minus 1.5 percent. Later work found abnormal returns near minus 2.55 percent over a window running from two days before to two days after expiry for lockups of 180 days or less; studies of Nordic listings land near minus 1.1 percent.
The theoretical damage is usually missed. Standard finance assumes the demand curve for an individual stock is close to horizontal: a share is a claim on future cash flows with near-perfect substitutes, so extra supply should not move the price as long as nothing about the business changes. Lockup expiries falsify that. Nothing about the business changes, and the price falls measurably anyway. Supply is a price factor in its own right, independent of information.
Note also what Field and Hanka did not find: a complete explanation. They tested several hypotheses and none of them fully accounted for the effect. Anyone claiming today to know exactly what a lockup expiry does to a share price is asserting more than the evidence supports. One detail of their work does map uncomfortably well onto the present case: both the abnormal return and the volume jump were considerably larger when the company was venture-backed, and venture capitalists sold more aggressively than executives and other holders. SpaceX is the most venture-backed listing in market history.
Facebook already ran this experiment
The closest American precedent is instructive mainly because it refuses to deliver a clean rule. Facebook listed in May 2012 at $38 and also used a staggered structure — five separate releases at roughly 90, 150, 180 and 365 days.
The first, on August 16, 2012, freed about 271 million shares held by investors including Microsoft, Accel Partners, Tiger Global Management, Goldman Sachs and Peter Thiel. The stock fell to its lowest level to that point, $19.69 intraday, and closed down 6.23 percent at $19.87 — barely half the offering price. Textbook outcome.
Then came November 14, 2012, the largest release of them all: close to 800 million shares, nearly three times August’s tranche. The stock closed 12.5 percent higher, on roughly 3.1 times normal volume. Two smaller releases followed, about 156 million shares in December and roughly 47 million on the one-year anniversary.
Same company, same year, same mechanism, opposite outcomes — and the bigger event produced the better day. The lesson is not that unlocks do not matter. It is that the market prices them in advance, imperfectly and unevenly, so the observable move on the day reflects the difference between the supply that arrives and the supply that was already feared. By November, Facebook’s selling had been dreaded for six months and the fear was larger than the flow. That is exactly the variable no model captures, and it is why position sizing beats prediction here.
The quarter underneath the supply story
To separate supply from information you need the information. SpaceX reported second-quarter results on August 4, two trading days before today’s expiry. Revenue rose 92 percent to $7.814 billion, beating consensus of roughly $6.8 billion by nearly a billion dollars. The connectivity segment that houses Starlink grew 66 percent to $4.3 billion against expectations of $3.83 billion. Starlink subscribers reached 12 million, double a year earlier. Connectivity was also the only segment to post an operating profit, at $1.66 billion for the quarter. Artificial-intelligence revenue rose 247 percent.
The stock fell anyway, dropping as much as eight percent after hours. The reason sat in the capital spending line, not the revenue line: capital expenditure hit $18.4 billion in a single quarter, against $10.1 billion in the prior three months. Doubling capital spending while exactly one segment earns an operating profit is a statement about how long this company intends to consume capital. Elon Musk countered with a $1 trillion revenue ambition and plans for robots on the moon. The market chose the capex line.
One further number deserves mention without being overweighted: average revenue per Starlink subscriber fell 22 percent year over year as the company pushed into international markets and introduced cheaper plans. Subscribers are doubling faster than revenue per subscriber holds. For a growth story that is not automatically bad — it moves the question to what scale the network needs before it carries itself.
The calendar is the risk, not the day
Anyone staring at today’s tape is watching the wrong time axis. After August 6 come August 21, September 10, and then December 8, when the main lockup lapses. Supply arrives as a calendar, not as an event. That is what makes it awkward for the share price: a one-off shock gets digested, whereas paper delivered across four months acts like persistent headwind, because the same buyer hesitation rebuilds ahead of every date.
Valuation permits both outcomes, which is why published targets sit so far apart. SpaceX carries a market capitalization near $1.4 trillion and trades at roughly 49 times expected revenue. Morgan Stanley, through analyst Adam Jonas, rates the shares overweight with a $300 target — about 166 percent above current levels — arguing that fundamentals are largely unchanged and that pre-expiry weakness offers an attractive entry. HSBC initiated at hold with a $115 target. The distance from $115 to $300 is a factor of 2.6. That spread is itself the most honest statement available: at 49 times revenue, almost everything rests on assumptions nobody can yet evidence.
What this means for a US investor in practice
Three specifics matter on this side of the Atlantic. First, the selling is visible if you look. Affiliates selling restricted stock file under Rule 144, and Form 144 notices are public; many insiders also sell through pre-arranged 10b5-1 plans adopted months in advance, which means a headline sale may carry no view on the stock at all. Reading the filing tells you whether you are looking at a decision or a standing instruction.
Second, a large share of insider supply never touches the open market. It moves as negotiated block trades to institutions. Ownership changes hands, the tape barely registers it, and the effect shows up in volume rather than price — which fits neatly with the permanent volume increase Field and Hanka measured.
Third, taxes shape behavior on both sides of the trade. An insider selling shares held for more than a year pays long-term capital gains rates; a shareholder who buys today and sells inside twelve months is taxed at ordinary income rates that run up to 37 percent. If you intend to trade a supply event rather than own a business for a decade, the position belongs in a tax-advantaged account, because the strategy is short-horizon by construction. There is also a mechanical offset worth knowing: index providers weight by free float, so a lockup expiry can raise a company’s index weight at the next scheduled review, generating passive demand from the very same event that created the supply — just with a lag.
What argues against all of this
The strongest objection is methodological. The minus 1.5 percent in Field and Hanka is an average across nearly two thousand cases, mostly from the 1990s, mostly at companies a fraction of today’s size. SpaceX is an outlier on every axis: market capitalization, ratio of locked stock to float, and sheer attention. Averages do not describe outliers. The literature establishes that the effect exists; it does not forecast its magnitude in this specific case.
Second, Facebook in November 2012 is a live demonstration that the largest unlock can coincide with the best day. Aswath Damodaran has argued at length that the liquidity story around lockup expirations is weaker and messier than the folklore suggests. Treating the calendar as a short signal has a real record of losing money.
Third, an expiry is not a buy signal either. Several studies suggest the effect does not cleanly reverse and is partly permanent. Buying today on the theory that the event is behind us confuses one date with a release schedule that runs into December.
Fourth, the objection against my own framing: the $62 billion block failing to trigger is mechanically helpful but substantively a confirmation of weakness. A clause that releases more paper into strength is one you would rather have seen fire.
The four questions that are already in the prospectus
The transferable part of today has nothing to do with rockets. It is this: on any violent price move, separate information from supply. Four questions do the work, and all four can be answered before you buy, because the answers are printed in the offering document.
One: who is permitted to sell, and from when? Two: how much actually comes free on each date — the tranche is what counts, not the total; today it is 20 percent, not 100. Three: how large is that quantity relative to the float rather than to market capitalization? Market capitalization makes every release look trivial; float shows what the market genuinely has to absorb. Four: what further dates follow, and are any of them conditional?
For SpaceX the answers read: insiders and early backers; 20 percent of 911.5 million shares today, which is close to 29 percent of the float, while the full release would equal about 140 percent of it; and August 21, September 10 and December 8 still to come, with a $62 billion block tied to a price condition that was not met. None of this is privileged information. It is a public document that almost nobody reads — which is precisely why a date known for months still moves the price.
Try TradingView Free for 30 Days
Plus get a $15 discount on your first subscription through this link.
