The IPO Supply Pipe Opens: SpaceX Lockups Expire as $700 Billion in 2026 Issuance Tests the Market

On August 6, 2026, more than 911 million SpaceX (SPCX) shares unlocked — more than doubling the publicly available float in Elon Musk’s space and AI company and marking the first real-time test of whether the market can absorb the supply wave that has been building since 2026’s record IPO pace began.^[1] The stock rose 6% on the day, a counterintuitive result that tells you something about the gap between theoretical supply overhang and actual selling pressure. But it is early. This was the first in a series of staggered lockup expirations stretching into 2027, and it sits inside a much larger story about the return of US equity supply after two decades of its slow disappearance.

The $700 Billion Pipe

US equity issuance reached approximately $700 billion in 2026 through July, according to Goldman Sachs analysis, combining initial public offerings, secondary offerings, convertible bonds, and SPACs.^[2] Ben Snyder, Goldman’s Chief US Equity Strategist, noted that this represents about 1% of Russell 3000 market capitalization — matching the 2015-2019 average and remaining below 2021’s roughly 1.5% and the dot-com boom peak of 2%.^[2]

By itself, 1% is a manageable number. The question is direction and velocity. The second quarter of 2026 saw the highest level of global secondary equity offerings in five years, as companies took advantage of favorable market conditions to raise additional capital, according to data from Wall Street Horizon.^[3] BigBear.ai (BBAI) filed to sell up to 100 million new shares — 20.9% of its outstanding stock — through Jefferies, which at the July 30 close would raise roughly $283 million.^[4] Twist Bioscience (TWST) upsized a stock sale to $300 million.^[4] The secondary calendar is not thin.

Meanwhile, 225 IPOs have priced on US exchanges through August 8, up 6.64% from the 211 by the same date in 2025.^[5] The Renaissance IPO Index was up 18.8% year-to-date as of August 6, outpacing the S&P 500’s 13.4% gain.^[6] The window is open, and issuers are using it.

The SpaceX Lockup: A Phased Release

SpaceX went public on June 12, 2026, selling roughly 640 million shares at $135 — less than 5% of its total share count.^[1] That tiny float, paired with intense retail demand, sent the stock to a record high on June 16 before it tumbled more than 40% from that peak.^[1] The first earnings report on August 4 revealed higher-than-expected AI capital expenditures, and the stock dropped 7% after hours, closing at an all-time low of $108.27 the next day — 15% below the IPO price.^[1]

The August 6 unlock more than doubled the available float to roughly 12% of total shares.^[1] But SpaceX chose a staggered, tiered lockup calendar rather than the standard 180-day cliff.^[1] More unlocks are scheduled across the coming year. Musk’s Class B shares — worth more than 40% of the company’s value — remain locked until one year after the IPO, meaning June 2027.^[1] Extended lockup periods for other executives begin expiring in 2027.^[1]

The immediate takeaway: the stock absorbed the first unlock without breaking. Retail investors have been net buyers every day since the IPO, and Wednesday and Thursday marked the strongest retail activity since the first three days of trading.^[1] But 88% of the company’s shares are still locked. The supply test is a process, not a single event.

The larger the float grows, the more weight SpaceX carries in the Nasdaq 100. TD Securities estimates that after September’s index rebalance — which will account for the newly available shares — SpaceX’s weighting could rise above 3.5% from its current ~1%, depending on the share price.^[1] A larger index weight means passive funds must hold more of it, which provides some demand-side offset. But it also means the stock’s volatility transmits more forcefully into the index itself.

Biotech Dominates the August Calendar

The week of August 3 saw six IPOs price, and four of them were biotechs — all upsized at pricing.^[6]

Source: Renaissance Capital^[6]

Braveheart Bio was the standout, pricing above its range and finishing the week up 67% — its lead candidate, an oral cardiac myosin inhibitor licensed from China’s Hengrui Pharmaceuticals, is heading into Phase 3 trials.^[6] The biotech cohort’s willingness to upsize signals that underwriters see real institutional appetite for clinical-stage risk, which is a marker of a healthy new-issue market. But the breadth is narrow. Ticketplus, a Latin American ticketing platform, priced at the bottom and finished down 13%.^[6] River City Bank, a commercial lender with ~90% of its loan portfolio in commercial real estate, downsized and priced below range.^[6] The window is open, but it is not equally open for everyone.

The De-Equitisation Reversal: Ninety One’s Warning

The most provocative framing of the supply question comes from Ninety One, the asset manager, in a research paper published in August 2026.^[7] Sahil Mahtani and Dan Morgan argue that two decades of “de-equitisation” — buybacks outpacing new issuance — have been a persistent support for US equity returns, adding an estimated 0.7 percentage points annually between 2015 and 2025.^[8]

That tailwind is ending. The initial AI listing wave — SpaceX, OpenAI, and Anthropic alone could raise $200-250 billion in initial proceeds — is small relative to the $75 trillion US equity market, representing about 0.3%.^[8] But the bigger risk lies further out. As lockup periods expire, free floats typically rise from roughly a quarter of the company to 70% or more within two years.^[8] Applied to a ~$4 trillion cohort of new AI companies, that points to an eventual increase in tradable US equity supply of close to 4% — a scale last seen during the 1990s and early 2000s, concentrated mainly in 2027 and beyond.^[8]

Ninety One’s numbers are direct. Their base-case ten-year US equity return forecast is 2.7% annually, where buyback-driven trends broadly continue. If market composition reverts to its century-average pace of net issuance, that falls close to zero. In a bubble-era scenario resembling 1995-2005, it drops to roughly -2.2% annually.^[8]

Mahtani explicitly cautioned against treating this as a market-top call: “Issuance-based signals are notoriously poor at calling one in real time, and the US market is far too deep to be disrupted by the first round of deals alone.”^[8] The key question, he says, is “how long the resulting increase in equity supply persists, and whether the companies now joining the index can deliver the growth their valuations already assume.”^[8]

The parallel Ninety One draws is to China’s IPO and MSCI index inclusion wave in the 2010s, where expanding equity supply combined with falling valuations erased five to six percentage points of return per year for index investors.^[8] That structural headwind is now fading in emerging markets — just as a comparable dynamic begins assembling in the United States.^[8]

Buybacks vs. Issuance: The Current Balance

Corporate buyback announcements totaled $960 billion year-to-date through late July, with Goldman Sachs projecting full-year 2026 volume at approximately $1.3 trillion.^[2] NVIDIA alone increased its buyback authorization by $80 billion.^[2] Berkshire Hathaway, under Greg Abel, repurchased $4.5 billion in Q2 2026.^[9] BASF launched a new €1 billion buyback program starting in August 2026, part of a larger €4 billion repurchase plan running through 2028.^[9]

Snyder’s assessment is that $1.3 trillion in buybacks “is sufficient to offset the combined potential supply from direct corporate issuance and lockup expirations” in 2026.^[2] Jay Ritter, Professor Emeritus at the University of Florida, added that “given the overall size of the US stock market, the likelihood of serious indigestion remains low.”^[2]

But there is a counter-signal. NDR noted in a late-July report that buybacks are falling from their 2025 peaks — a potential warning for S&P 500 firms, particularly in technology, where AI capital expenditure is competing with repurchase capacity for cash.^[9] If AI capex continues to scale, the buyback machine that has been absorbing issuance for two decades may shrink even as the supply of new shares grows. That is the tension.

Goldman’s Snyder acknowledged the trajectory: “The supply-demand balance is clearly moving in a negative direction.”^[2] He also noted the self-correcting mechanism — “if supply becomes so excessive that the market starts to wobble, that itself constrains future issuance and self-corrects”^[2] — which limits the probability of a disorderly outcome but does not change the direction of travel.

The 2027 Supply Calendar

The structural test arrives in 2027. Lockup periods for the 2026 IPO cohort — which includes SpaceX, SK hynix (SKHY), Cerebras (CBRS), Bending Spoons (BSP), and Neutron Holdings (LIME) — will expire on rolling six-month schedules.^[2] SpaceX alone has more than 7.5 billion Class A shares (mostly held by employees and investors) and more than 5.5 billion Class B shares (mostly held by Musk and executives).^[1] After the full series of 2026 unlocks, more than 5 billion Class A shares could be publicly available for trading.^[1] Musk’s shares — more than 40% of the company — unlock in June 2027.^[1]

The scale is not trivial, but it is not overwhelming either. Goldman’s 1%-of-market-cap framing puts it in historical context. The real question is whether the buyback engine continues to run at $1.3 trillion annually while AI capex demands accelerate, and whether the companies entering the index can justify the valuations assigned to them at IPO.

What to Watch Next

1. SpaceX’s next unlock tranche. The staggered schedule means additional share releases through late 2026. Watch the float-to-volume ratio: if daily trading volume does not expand proportionally, even phased unlocks will compress the stock. The September Nasdaq 100 rebalance — which will reweight SpaceX based on its expanded float — is a concrete date to circle.

2. Q3 secondary-offering volume. Q2 2026 set a five-year high.^[3] If Q3 matches or exceeds it, that confirms the supply pipeline is structural, not opportunistic. Watch the mix: ATM offerings (like BigBear.ai’s) are a different signal than priced follow-ons.

3. Buyback announcement rate. Goldman’s $960 billion YTD figure is a record pace.^[2] But if the NDR signal about waning repurchases is correct, the H2 2026 buyback announcement run-rate will slow even as issuance accelerates. That divergence is the core tension.

4. OpenAI and Anthropic IPO timing. Ninety One estimates these two could raise $200-250 billion combined.^[8] If either files an S-1 before year-end, the 2027 lockup calendar gets dramatically heavier.

5. IPO performance dispersion. Of 225 YTD listings, the return distribution is wide — from Swarmer’s +632% to Green Circle Decarbonize’s -89%.^[5] If the median IPO return turns negative, it signals the market is struggling with supply absorption, per Snyder’s framework.^[2] Watch the median, not the mean.

The base case is that the US market absorbs the supply. Buybacks at $1.3 trillion dwarf issuance at $700 billion, and issuance at 1% of market cap is historically unremarkable. The risk case is not a 2026 problem — it is a 2027-2028 problem, when lockups expire in waves and the buyback engine may be running on less fuel. The SpaceX unlock on August 6 was the first data point. The stock’s 6% rally that day suggests the market is not yet choking. But 88% of SpaceX’s shares are still locked. The supply test has barely begun.