In September 2014, Yahoo disclosed that it had received approximately $9.4 billion from selling part of its ownership in Alibaba, the Chinese technology company whose initial public offering had just broken a Wall Street fundraising record. The transaction involved 140 million American depositary shares priced at $68 each, including shares sold through the underwriters' additional-share option. Yahoo recorded the proceeds in a filing with the US Securities and Exchange Commission, providing a clear account of the cash realised after underwriting discounts and commissions. For Yahoo, the transaction had achieved its immediate purpose: an investment accumulated over several years had been converted into billions of dollars. For Alibaba, the listing established a public market for its shares and raised money for further expansion. Yet for someone purchasing the stock after trading began, the financial proposition was entirely different. Alibaba closed its first trading day at $93.89, approximately 38% above the offering price; by 9 November 2026, it was trading at $105.38. An investor who bought at that first closing price had earned approximately 12.2% in nominal price appreciation over more than twelve years, before distributions and costs. Yahoo's money had already been realised in 2014. ¹
The comparison does not mean that Yahoo's investment and the later public-market purchase carried equivalent risks. Yahoo acquired its stake much earlier, under conditions unlike those facing public investors in 2014. It does reveal an essential distinction often obscured in accounts of historic IPOs: the company raising new capital, shareholders selling existing shares and investors buying the newly traded stock may experience very different outcomes. A listing can offer an extraordinary exit for existing owners without delivering an equivalent return to later buyers. Equally, an unremarkable debut can become an exceptional long-term investment if subsequent earnings, cash distributions and growth justify the valuation. The size of the offering is a measure of how many shares were sold and at what price; it is not, in itself, a measure of wealth created for public shareholders.
That distinction acquired new urgency in June 2026, when SpaceX began trading on Nasdaq after the largest initial public offering on record. The company sold approximately $75 billion of shares before the underwriters exercised their additional-share option, far exceeding the sums raised in Saudi Aramco's 2019 listing or Alibaba's 2014 transaction. SpaceX's offering priced at $135 a share, but the stock finished its first trading session on 12 June at $160.95. On 9 November, it closed at $142.07. A hypothetical investor allocated 100 shares at the original offering price had a small paper gain at that November quotation, while someone who bought the same number at the first day's closing price was sitting on a loss of approximately 11.7%. Both investments concerned the same company, but their starting valuations were different. The result does not establish how SpaceX will perform over the next decade. It does show why a record-breaking fundraising achievement cannot be treated as a guarantee of an attractive investment. ² ³
The Raisina Post examined offering documents, SEC filings, corporate-action records, company financial statements and historical market prices for landmark IPOs across the United States, China, Saudi Arabia, Japan and Hong Kong, using the 9 November 2026 closing quotations as a common market cutoff. The purpose is not to crown a single investment winner from securities listed in different years and currencies, but to identify the decisions and structures that determined who benefited. In several cases, the money went primarily to an existing shareholder rather than the operating company; in others, the new shares financed corporate expansion. Among the companies that remained listed, long-term share-price appreciation ranged from extraordinary to modest, with cash dividends adding another dimension that a simple price comparison misses. Wider academic evidence also suggests that the celebrated winners are far from representative of IPO investing generally.
The transactions behind the headlines
The scale of the biggest IPOs
Where did the billions actually go?
An initial public offering can consist of newly issued shares, existing shares or a combination of both, and that difference determines where the buyer's money ends up. When a business issues primary shares, the proceeds generally enter the company after expenses and may support investment, acquisitions, repayment of borrowings or other corporate needs. When an existing owner sells secondary shares, the proceeds are received by that owner; the company gains a wider shareholder base and a public market, but the secondary sale does not itself inject money into its operations. Both structures are commonplace and legitimate. The problem arises when a headline about the total amount 'raised' leaves readers with the impression that every dollar is financing the company whose name appears on the stock-market screen.
Facebook's 2012 IPO demonstrates how large the difference can be. The final prospectus recorded 180 million shares newly issued by the company alongside approximately 241.2 million shares sold by existing shareholders. At $38 per share, the new stock represented about $6.84 billion of gross proceeds, while the stock sold by existing holders represented approximately $9.17 billion. Facebook explicitly said it would receive none of the proceeds from those secondary sales. Thus, more than half the shares in the $16 billion offering were sold by people who already owned them. The transaction supplied financing for Facebook while providing liquidity to existing investors, but those two activities benefited different recipients. Alibaba's offering in 2014 followed a similar combination of primary and secondary sales, and Yahoo's subsequent SEC filing establishes how much one major seller received. ⁴ ⁵
The distinction becomes especially clear in the flotations of Saudi Aramco, SoftBank Corp. and AIA Group. Saudi Arabia's 2019 transaction offered existing government-owned shares in its national oil producer, raising approximately $25.6 billion before the additional-share option. In Japan, SoftBank Group used the 2018 flotation of its telecommunications subsidiary, SoftBank Corp., to monetise part of an established holding. In Hong Kong, AIA's 2010 offering formed part of American International Group's restructuring after the global financial crisis and consisted of existing shares sold by an AIG-related owner. These companies operated in very different markets, and their transactions had distinct political and financial objectives, but all provided a means for pre-existing owners to realise value. The public shareholders acquiring those stakes were beginning investments whose outcomes would depend on operating performance, market prices and future distributions rather than the amount collected on listing day. ⁶
How much of each IPO went to existing owners?
SpaceX belongs to a different category because the 2026 sale involved newly issued shares, providing substantial capital for the business's expansion across space transportation, satellite systems and artificial intelligence. Its first 555,555,555 shares at $135 apiece amounted to approximately $75 billion. After the underwriters exercised their option, the company had sold 638,888,888 shares; multiplying that share count by the offer price gives roughly $86.25 billion in gross proceeds. In its subsequent Form 10-Q, SpaceX reported $85.675 billion in net proceeds after $575 million of underwriting commissions and expenses. An earlier announcement described approximately $85.7 billion as gross proceeds, a difference in terminology that can be reconciled through the disclosed share count, price and deductions. This is not evidence that any money disappeared or that the company's financial statements were falsified. It is a concrete example of why gross offering size, post-option sales and net cash received should not be described interchangeably in financial reporting. ⁷ ⁸
The offering also illustrates how public ownership can be separated from managerial control. SpaceX's differentiated voting rights allowed Elon Musk to retain overwhelming voting power even after new investors acquired publicly traded shares. That arrangement may provide continuity for an enterprise undertaking long-term, capital-intensive projects, but it also concentrates decisions about strategy, spending and governance. Public shareholders are participating in the financial performance of SpaceX without receiving proportionate authority over its direction. When evaluating such an offering, the headline proceeds should therefore be read alongside the voting structure, expected capital expenditure and the valuation investors paid for the shares.
Watch Nasdaq's original SpaceX IPO opening-bell coverage. The recording documents the June 2026 listing ceremony; it is linked to its publisher rather than reproduced here.
The price paid changed everything
An investor who receives an IPO allocation at the original offering price may begin with a materially different cost basis from someone who buys the same stock once public trading has started. SpaceX's first-session rise makes the distinction unusually visible. The $25.95 difference between the $135 offer price and the $160.95 first-day close, multiplied by the approximately 555.6 million shares in the initial offering, implies roughly $14.4 billion in theoretical paper gains on those offered shares. This calculation is often called 'money left on the table', but it does not show that allocated investors actually sold all their shares at the closing price, or that the underwriters could have placed the entire offering at that valuation. What it does show is how much the starting price of a public-market investment can change within hours of an IPO. For buyers entering at the first-day close, the subsequent share-price performance is measured from that higher figure, not from the more favourable offering price.
Alibaba provides the longer-term counterpart. By 9 November 2026, its $105.38 closing price was approximately 55% above the $68 offer price but only 12.2% above the first-day close of $93.89, before dividends and costs. Its commercial history was much larger than the capital gain experienced by that later buyer. The group expanded across e-commerce, cloud computing, logistics and international markets, while confronting competition, changing regulation and substantial investment demands. For the June quarter of 2026, Alibaba reported revenue of approximately RMB268.95 billion, up 9% year on year, even as operating income declined substantially. A successful and expanding operating company is not necessarily a successful purchase at every price. Whether the buyer ultimately earns an attractive return depends on how much growth the purchase price had already anticipated. ⁹
The same distinction is relevant to SpaceX's future. The company has an extraordinarily ambitious investment programme, but less than five months of public trading cannot establish whether its long-term cash generation will justify the valuation at which it went public. Rocket infrastructure, satellite networks and AI computing require substantial spending, while the company's broader corporate structure complicates comparisons with earlier financial periods. Public investors must assess earnings, financing requirements, potential dilution and the valuation paid rather than treating first-day trading excitement as evidence of durable returns. SpaceX's flotation was a financial milestone for the issuer; the verdict on the buyer's investment remains open.
What happened to the shareholders?
Visa's 2008 flotation provides one of the most striking examples of long-term wealth creation among major IPOs. The payments company went public during a severe financial crisis, selling approximately 406 million Class A shares at $44 each and raising $17.9 billion before additional shares. Its first-day close reached $56.50, but the far more important development followed over nearly two decades. Visa completed a four-for-one stock split in March 2015, so an investor who bought 100 shares for $4,400 at the original offering price would later hold 400 shares. At the 9 November 2026 quotation of $422.16, those shares were worth approximately $168,864, a nominal price-only gain of about 3,738% before cash dividends, costs and taxes. The company's continuing operating performance helps explain the appreciation: for fiscal 2026, Visa reported net revenue of $45.9 billion and GAAP net income of $23.5 billion, while returning $25.6 billion to shareholders through dividends and repurchases. These are results of a business that expanded alongside global electronic payments, not gains that were guaranteed by the scale of its IPO. ¹⁰
Facebook's investment history followed a different route. Its 2012 debut was disrupted by trading problems and doubts about the company's ability to monetise mobile activity, and the stock later declined sharply from its offering price. Yet the company expanded its advertising operation, incorporated Instagram and WhatsApp, and eventually became Meta Platforms. At a 9 November 2026 share price of $641.37, its stock was approximately 1,588% above the original $38 offer price, before dividends. Even that extraordinary historical appreciation should not lead to uncritical reporting of current results. Meta's third-quarter 2026 revenue rose 26% to approximately $64.6 billion, while reported operating income fell to about $13 billion. The earnings release disclosed a $9.89 billion charge related to an agreement with US state attorneys general over litigation concerning teen mental health and social media. Meta said that excluding the charge, operating income would have been approximately $22.9 billion. That is the company's adjusted presentation rather than a separate audited measure, but it demonstrates why a headline percentage change in profit must be examined alongside the components of the financial statement. ¹¹
Saudi Aramco illustrates why dividends and bonus shares can be as important as the quoted stock price. The company sold shares at SAR32 in 2019 and traded at SAR26.10 on 9 November 2026. A direct comparison would overstate the capital loss because Aramco issued one additional share for every ten held in 2022 and repeated that arrangement in 2023. An investor who bought 100 shares for SAR3,200 and remained eligible for both issues would therefore hold 121 shares, worth approximately SAR3,158 at the November quotation. That represents a price-only decline of about 1.3%, but the calculation excludes substantial cash dividends. In its third-quarter 2026 announcement, Aramco reported adjusted net income of $31.2 billion and free cash flow of $28.2 billion, while declaring a quarterly base dividend of $21.9 billion. A proper total-return analysis must add cash distributions received and account for reinvestment where appropriate. An enterprise can remain exceptionally profitable and distribute enormous amounts to its owners without delivering spectacular share-price appreciation. ¹²
SoftBank Corp.'s 2018 offering offers a more conventional example of moderate price appreciation alongside distributions. The Japanese telecommunications subsidiary, which should not be confused with its investment-holding parent SoftBank Group, priced at ¥1,500 a share and disappointed investors on its first trading day. After a ten-for-one split in 2024, an original share corresponded to ten, making its ¥238.40 quotation in November 2026 equivalent to approximately ¥2,384 for the original holding. The split-adjusted price-only gain was therefore about 58.9%, with dividends contributing additionally. AIA's 2010 Hong Kong offering, undertaken as part of AIG's restructuring, was priced at HK$19.68; at approximately HK$78.55 in November 2026, it showed price appreciation of about 299%. Both outcomes were substantial but materially different from Visa's, illustrating why even among famous large offerings there is no single typical return.
A comparison of surviving securities must also address companies that no longer trade independently. NTT Docomo, whose 1998 flotation ranked among the largest, was subsequently acquired by its parent and delisted in 2020. Its 2026 investment outcome cannot be computed from a nonexistent 2026 market price; it requires the acquisition consideration and earlier distributions. The same caution applies to major Chinese bank offerings involving separate mainland and Hong Kong share classes, each with different offer prices and currencies. In July 2026, SK hynix raised approximately $26.5 billion through a US depositary-share offering, but the company was already publicly traded in South Korea. An enormous securities sale should not automatically be described as a first-time public listing. These classification issues help explain why rankings of the world's largest IPOs differ across providers.
The thousands of IPOs nobody remembers
The celebrated successes of Visa and Meta might suggest that public investors in large offerings have generally done well, but such a conclusion would be vulnerable to survivorship bias. Successful companies remain visible, publish financial statements and generate retrospective accounts of fortunes made. Failed businesses, deeply depressed stocks and delisted issuers are far less likely to appear in histories of market milestones. A sample of famous winners will therefore exaggerate the typical experience of investors unless it is compared with a broad population that includes disappointing outcomes. The performance of large global offerings cannot be statistically inferred from a few examples, and the experience of the wider IPO market offers important context.
Jay R. Ritter, a finance professor at the University of Florida, has assembled a long-running dataset of US IPOs. His July 2026 statistical compilation examines three-year buy-and-hold returns for 9,195 US operating-company IPOs listed between 1975 and 2021, measured from the first-day closing price to the third anniversary or an earlier delisting or data cutoff. In that sample, 3,537 IPOs, or 38.5%, lost more than half their value, while a further 1,979, or 21.5%, had returns between negative 50% and zero. Combined, exactly 5,516 of 9,195 offerings, or approximately 60.0%, produced no positive three-year return. The median return was negative 25.7%, even as the average was positive at 21.2%, showing how comparatively few large winners can raise an average while the typical experience is much weaker. These figures include dividends and capital gains under the study's method and account for delistings. They are not measures of the performance of the global mega-IPO subset discussed in this investigation. ¹³
What happened to 9,195 US IPOs after three years?
Ritter's sample excludes categories such as depositary receipts, SPACs, banks, real-estate investment trusts and very low-priced offerings, and the results cannot be applied mechanically to SpaceX, Aramco or other individually selected foreign listings. The data nevertheless clarify why the offer price and first-day closing price must be distinguished and why headline investment winners should not be mistaken for representative outcomes. To determine whether mega-IPOs systematically outperform or underperform smaller offerings would require a consistent international dataset, comparable holding periods, delisting proceeds, cash dividends, corporate actions, market benchmarks and exchange-rate treatment. The evidence assembled here supports conclusions about particular companies and offering structures; it does not establish a universal investment rule.
The fortune realised before the market opened
The historical record shows that the companies, earlier shareholders and new public investors participating in the same IPO are not necessarily making the same financial bet. A business issuing new shares is seeking money at a valuation acceptable to its owners and advisers. An existing shareholder selling an established holding is seeking liquidity. A government may have broader fiscal or policy purposes. A public investor purchasing those shares expects future appreciation or distributions sufficient to compensate for the price and risks assumed. These interests can align, but their financial outcomes emerge at different times. Yahoo's Alibaba proceeds were realised in 2014, while later investors' returns depended on twelve years of subsequent market performance. SpaceX completed an unprecedented financing transaction in June 2026, while its shareholders still face a much longer test of whether the offering valuation is justified.
That separation is particularly important when a company has an impressive commercial record or a powerful cultural reputation. A stock offering transfers or creates ownership claims at a particular price; it does not guarantee that future growth will exceed what the market has already assumed. A business can become more valuable while its shareholders earn modest returns from an expensive starting valuation, or it can deliver substantial shareholder wealth through sustained earnings and dividends without the publicity surrounding a record debut. Visa's extraordinary appreciation, Aramco's cash distributions, Alibaba's comparatively restrained price performance and SpaceX's still-unresolved investment case all illustrate different sides of that proposition. The financial outcome for the seller can be known on listing day. For the buyer, the real test begins afterward.
Reporting note: This is a documentary draft based on public prospectuses, company announcements, financial statements and historical share-price records, with a common 9 November 2026 market cutoff. Offering sizes are initial values before overallotment where specified. Share-price returns are nominal, calculated in the securities' trading currencies, adjusted for the identified stock splits or bonus issues and exclude cash dividends, reinvestment, fees, taxes, inflation and foreign-exchange movements. The selection is illustrative, not a statistically representative global IPO sample. The separate Ritter research describes its defined US IPO sample. No company interviews, proprietary trading records or non-public allocation data are claimed.