Bocconi Students Capital Markets
  • ABOUT US
  • Team
  • ARTICLES
    • Americas
    • APAC
    • EMEA
    • BSCM Analyses
  • Events
  • Alumni
  • JOIN US

The IPO Revival: A Strategic Choice or a Funding Necessity?

Although the US IPO market is recovering, the return of large offerings raises an issue that goes beyond the volume of deals: why are businesses deciding to list? For some, public markets provide shareholder liquidity and strategic flexibility; for others, they supply the capital required to maintain investment. This distinction is highlighted by Anthropic’s proposed listing, which combines significant infrastructure commitments with rapid growth. Drawing on historical data, company cases, and post-IPO returns, we examine how investors can differentiate between financial dependence and strategic choice, and why neither label alone determines whether an IPO is a desirable investment.​

From the Dot-Com Boom to the 2026 IPO Revival

An IPO can provide capital before a company becomes self-funding. The technology bubble pushed that exposure to its limit: in 2000, 81% of issuers in Ritter's operating-company sample had negative earnings before listing, so investors were buying losses in exchange for expected future profits. The risk lay less in the need for money than in the price paid for those expectations. The collapse that followed lowered that price and changed the choice facing smaller companies.

Between 2001 and 2009, a small company could list independently or sell to a larger business with established distribution and resources, and the second option grew stronger. Research by Gao, Ritter, and Zhu linked the decline in small-company IPOs to the advantages of operating inside larger organizations: an acquirer could bring products to market faster and make expansion more profitable than the company could alone. Going public therefore lost appeal for some founders even when the business itself remained worth financing. Selling was not the only alternative to a listing; private capital soon offered another.

During the 2010s, deeper private markets let companies grow without listing at all. Ewens and Farre-Mensa linked securities-market deregulation to greater funding for later-stage startups, which helps explain why fewer companies went public and why those that did were older. Founders responded accordingly: raising substantial capital no longer required a public offering, and a stronger bargaining position let founders delay listing whenever remaining private better served their interests. Investors inherited the consequences of that delay when these companies eventually listed.

Delay did not guarantee self-sufficiency, because additional private funding could finance years of expansion while postponing the need to generate cash internally; age and scale became less reliable shortcuts for financial maturity. A business could reach public markets with an established product and substantial revenue while still depending on new shareholders to cover its spending. For investors, an older issuer was not necessarily one that could support itself, and the income statement could hide the difference.
Reported profit can conceal a funding requirement, as Uber's 2019 listing showed. The company reported about $1.0bn in net income for 2018 while using $1.5bn in operating cash, because divestiture and investment gains drove the profit while operations ran at a loss. A shareholder reading only the income statement would have seen a profitable issuer; the cash-flow statement showed a business that was consuming cash. Accounting profit and the ability to pay operating bills answered different questions, and the next boom tested that distinction at scale.

The 2020 and 2021 boom reopened public markets to companies without profits: 80% and 75% of Ritter's issuers, respectively, were loss-making. Median technology IPO sales multiples remained below their 2000 level, with median offer-price sales multiples of 13.6x and 15.2x in his technology series, against 31.7x in 2000. That series includes direct listings and measures equity value against trailing sales. Despite these lower median multiples, the boom still carried substantial financial risk because issuers gained access to funding before their business models produced sustained profits.

The contraction came quickly after 2021: Ritter counted 38 IPOs in 2022 against 311 in 2021, and issuance had recovered only to 90 by 2025. A smaller market admitted stronger companies on paper, as WilmerHale found that profitability among US-incorporated issuers raising at least $50m rose from 23% in 2019 through 2021 to 33% in 2022 through 2024. The improvement describes who was admitted, not the wider population: companies unable or unwilling to list left the comparison without losing their financing needs.
The industry mix also shapes these averages. In WilmerHale's broader 2024 sample, only 10% of life-sciences issuers were profitable, compared with 61% of other issuers, because funding research before commercial revenue produces different financial statements than expanding an established consumer business. A higher share of biotechnology listings can therefore lower aggregate profitability without any comparable company weakening. Which companies go public matters alongside each company's financial condition, and the sources used to measure both are not interchangeable.

Coverage differences prevent these sources from forming one continuous series. Ritter excludes SPACs, REITs, banks, ADRs, and offer prices below $5 per share, among other categories; WilmerHale's definitions differ across its tables; PwC reports traditional IPOs separately from SPAC issuance and completed SPAC mergers. Figure 1 uses Ritter alone, and Figure 2 uses PwC alone, so a change in sample coverage is not mistaken for a change in companies' financial health.
Picture
Figure 1: Share of US operating-company IPOs with negative trailing earnings, 2000 to 2025 (%).

​The 2026 recovery was real but narrower than its headline. From January through September, PwC reported 83 traditional US IPOs raising over $120.2bn, against 60 raising over $29.3bn in the same months of 2025. Deal count rose about 38%, yet proceeds roughly quadrupled because SpaceX's $75bn offering accounted for most of the total. Excluding that transaction, PwC reported approximately $45.2bn, around 54% above the comparable 2025 total. Figure 2 separates SpaceX from other proceeds, showing a recovery that went beyond one deal, but on a far smaller scale than the headline implies.
Breadth across sectors supports the same reading: PwC counted 22 biotech IPOs raising $6.9bn through September 2026, and its third-quarter leaders included data-center company Csquare at $1.1bn and Jersey Mike's at $1bn. Yet six sizeable offerings were postponed or withdrawn during September. Access remained selective. Public markets supplied more capital, but companies still had to justify their valuations and persuade investors that future cash generation could support the financing they received.
​
Picture
Figure 2: Traditional US IPOs, January to September 2025 and 2026 (number of offerings; proceeds in USD billions)

​When Going Public Meets a Funding Need
​

Three factors help assess how dependent a company is on external financing: its cash runway, the capital intensity of its business, and the gap between its commitments and available funding. The cash runway measures how many months a company can operate at its current burn rate. Less than 12–18 months of runway signals financial pressure, but it does not by itself make an IPO inevitable: a company can still raise private equity or debt, sell assets or cut spending, and the public market is only one of several possible sources of capital and liquidity. Capital intensity narrows those alternatives: a software company may cut marketing to approach breakeven, whereas a company building factories or data centers may find it harder to reduce investment quickly.
Rivian's November 2021 listing illustrates both sides. Its prospectus shows that the company had not yet generated material revenue from vehicle sales, lost $994m in the first half of 2021 and estimated a third-quarter loss of $1.2–1.3bn, alongside quarterly capital expenditure of $450–500m. It expected cumulative capital expenditure of about $8bn through 2023, while estimated cash stood at roughly $5.1bn at the end of September. The funding need was therefore large and persistent. Other sources, however, were still open: Rivian issued $2.5bn of convertible notes in July 2021 and $1.25bn of senior notes in October. Losses and capital intensity explain why Rivian needed substantial external capital, but not why it had to list at that particular moment; the timing also reflects strong investor demand for electric-vehicle stocks in 2021. The base offering of 153m shares at $78 raised $11.9bn, or $13.7bn gross once the underwriters exercised their option in full. By the end of 2022 the shares were more than three quarters below the offer price. 
Private funding and secondary sales can postpone a listing, but concentration limits and shareholder liquidity needs can make those alternatives less attractive as companies grow. WeWork's attempted listing in 2019 shows what happens when the offering itself becomes part of the financing plan. The company lost $1.9bn on $1.8bn of revenue in 2018, carried $47.2bn of lease obligations and had arranged a $6bn credit facility that was expected to close together with the IPO. After investors rejected its $47bn private valuation, the IPO was withdrawn and its largest shareholder had to rescue it. 

Anthropic: Funding the Next Phase of AI Growth
​

Anthropic’s proposed listing raises the question of financial necessity at an unprecedented scale. After a confidential submission in June, the company publicly filed its registration statement on Form S-1 on 26 October 2026; the number of shares and the price range have not yet been set. Press reports suggest that the offering could raise up to $100bn at a valuation of about $2tn, with Reuters reporting a target listing in mid-November. 

Several financial measures must be distinguished. Revenue, the GAAP figure, reached $16.3bn in the first half of 2026, up from $1.2bn a year earlier. "Run-Rate Revenue", which annualizes the latest 28 days of consumption, is an operating metric that the filing itself says is no substitute for revenue. The net loss of $80.7bn is largely an accounting effect: $72.4bn reflects the change in fair value of convertible notes, derivative liabilities and SAFEs as the company’s valuation rose. The operating loss for the first half was $693m, although the second quarter alone recorded $1.0bn of operating income. Operating cash flow was positive, at $3.9bn. At 30 June 2026 Anthropic held $112.4bn of cash and cash equivalents, $19.3bn of short-term investments and $3.2bn of undrawn credit, which management considers sufficient for at least the next twelve months. These disclosures do not suggest an immediate liquidity shortfall.

The longer-term financing challenge lies instead in future contractual commitments, most of which are not yet recorded as liabilities on the balance sheet. The S-1 discloses minimum spending commitments of $114.5bn with Google, $110.0bn with Amazon and $44.9bn with Microsoft, payable regardless of actual usage, together with about $161.2bn of equipment lease payments, excluding interest, with Broadcom. Agreements with SpaceX's AI unit, Nscale and Lambda depend on capacity being delivered or can be terminated, while additional TPU and AMD capacity, estimated at a further $145bn or more, is excluded from the contractual table. The filing also notes that revenue is mainly consumption-based and not tied to long-term contracts, whereas compute commitments extend over many years. 
​
Anthropic is therefore a hybrid case. Its revenue growth and positive operating cash flow resemble those of a strategic issuer, but revenue growth is not the same as financial solidity over the life of commitments that extend well into the 2030s. The IPO is less a response to a shortage of capital today than a way to broaden the funding base behind those commitments, and a key risk transferred to public investors is operating leverage: if demand grows sufficiently, those commitments could support profitable expansion; if growth slows, revenues may fall short of what is needed to cover those commitments.

Strategic Listings: Growth, Flexibility and Shareholder Liquidity
​

At the opposite end of the spectrum are companies that go public from a position of financial strength: they are profitable or close to it, able to fund their investment plans internally, and run a business model that has already proven it can scale. Google is the classic example. In 2003 it earned $106m of net income on $1.47bn of revenue, and in August 2004 it sold $1.67bn of shares through an auction designed to limit underpricing. $1.20bn of that amount, before underwriting discounts, came from new shares issued by the company; the remaining $0.46bn went to selling shareholders. Google was not facing an immediate funding shortfall; contemporary reporting attributed the timing largely to the SEC rule requiring firms with more than 500 shareholders of record to publish their financial statements. Facebook's 2012 IPO followed a similar logic. The company had earned $1bn of net income on $3.7bn of revenue in 2011, and $6.84bn of the approximately $16bn offering represented newly issued shares, before fees and expenses. The listing also gave Facebook the acquisition currency described by Brau and Fawcett (2006): when it bought WhatsApp in 2014 for around $19bn, most of the price was paid in Facebook shares. 

Arm's return to the public market in September 2023 shows that a strategic IPO is not always a capital raise. The company earned $524m of net income in its 2023 fiscal year, and every share in the $4.9bn offering was sold by SoftBank, its owner; Arm itself received no proceeds. A listing can therefore serve strategic objectives for the issuer and still transfer risk to public investors, in this case through a valuation of around 20x sales and a controlling shareholder with its own capital needs. 
​
Table 1 compares these issuers at the time of their offerings: among the strategic cases, the share of proceeds received by the company varied substantially, and its operations did not depend on the offering. Had the market window been closed, each of them could most likely have postponed its listing without threatening its operations. The ability to postpone a listing without jeopardizing operations is a useful distinction between strategic flexibility and financial necessity, although individual IPOs can combine both motives.
Picture
Table 1: Financial Position of Selected Issuers Around Their IPO or Filing
​¹ Includes the underwriters’ option; base offering $11.9bn. ² Includes $72.4bn of non-cash fair value changes; operating loss $693m. ³ Includes short-term investments or marketable securities. Proceeds percentages are calculated on a gross basis.
​

After the IPO: Financing Constraints, Valuation and Returns

Financing constraints can weaken shareholder returns when companies cannot fund worthwhile investment or must accept costly terms to obtain capital. A loss alone does not establish that problem: a company can lose money while retaining willing investors, just as a profitable business can struggle to refinance debt. The empirical question is therefore narrower than whether weak companies perform badly; it is whether restricted access to finance predicts weaker returns after allowing for differences in the business and its valuation.
Answering that question requires a reliable measure of restriction, and several common measures fail. Farre-Mensa and Ljungqvist found that firms labeled constrained by those measures did not behave as if their access to capital were restricted. Small size, limited cash, and low profitability may identify vulnerable businesses without proving that lenders or investors will refuse them funding. A return comparison built on those labels can therefore confuse operating weakness with a financing constraint before the analysis begins.

Although none of it isolates financing constraints, the general IPO evidence records relative underperformance at one, three, and five years. Over one year, Ritter's 1,299 IPOs from 2000 through 2009 returned an equally weighted minus 11.3%, against 8.4% for size-matched companies, with returns measured from the first trading close rather than the offer price. The result establishes a weak first year for that cohort, not a penalty for constrained issuers, because the comparison does not divide companies by financing access.
At three years, the benchmark determines whether positive returns amount to relative underperformance. Ritter's 1991 study followed 1,526 IPOs issued between 1975 and 1984 from the first trading close to the third anniversary or earlier delisting, reporting average returns of 34.47% against 61.86% for companies matched by size and industry. Investors gained money while falling behind comparable stocks, so a positive absolute return can still represent underperformance.

A separate study examining five-year holding periods also finds IPO underperformance: Loughran and Ritter found that IPOs from 1970 through 1990 returned about 5% a year, against about 12% for similar-sized nonissuers over corresponding holding periods. These are annualized figures, not cumulative five-year gains. Both older studies provide historical benchmarks, but neither offers a current estimate of the return cost driven specifically by financing constraints, and both depend on the chosen benchmark.
That dependence is the strongest objection to an IPO-specific penalty. Brav and Gompers examined venture-backed IPOs from 1972 through 1992 and other IPOs from 1975 through 1992 and found that giving larger companies more portfolio weight substantially reduced apparent underperformance. Small issuers without venture backing did perform poorly, but so did comparable nonissuing companies. This weakens the interpretation that listing itself explains the underperformance.

Cash-flow research offers a clearer but narrower finding. Liu and colleagues studied 5,548 IPOs from 1980 through 2017 and associated larger cash shortfalls with weaker subsequent net cash flows, measured as operating cash flow less capital expenditure and scaled by assets. Their measure subtracts primary IPO proceeds from cash recorded after issuance, estimating a hypothetical funding gap rather than observing how the business would have operated without the offering. The published results reflect operating performance, not shareholder returns, and the measure's timing limits its use at listing.
Persistent cash consumption also does not automatically imply poor returns: funding can support valuable investment, while an apparently strong cash balance can conceal future obligations. What matters is what the spending produces relative to the price shareholders paid, which returns the analysis to valuation.

Valuation evidence supports examining price and profitability together: across more than 2,000 IPOs from 1980 through 1997, Purnanandam and Swaminathan associated richer valuations with weaker subsequent risk-adjusted returns. More expensive issuers also had lower profitability and stronger growth expectations, so investors could lose on disappointed expectations even when an offering successfully financed expansion. A profitable company bought too expensively can be a worse investment than a loss-making company bought cheaply.
None of these associations establishes causation, because sector, company size, initial valuation, and listing year can influence both financing conditions and later returns. More convincing return evidence would require a credible external change in financing access and a comparison group facing similar operating conditions.
​
Recent listings cannot close the gap yet: companies that went public in 2026 have no full one-year returns, much less three-year or five-year outcomes. Any valid comparison must retain failures, account for dividends and delistings, and keep offer-price returns separate from aftermarket returns. For constrained issuers, the available research does not establish a reliable return penalty across all three horizons, and general IPO averages cannot supply that missing comparison on their own. The evidence supports closer scrutiny of cash generation, financing access, and valuation, not a universal return penalty for every company that needs an IPO to keep growing.

A Practical Framework for Assessing IPO Risk
​

The preceding sections show that the distinction between a necessity-driven and a strategic listing is rarely visible in a single figure. Profitability does not settle it, since capital-intensive issuers can consume cash for years while remaining viable; the size of the offering does not settle it either, since large proceeds can fund either expansion or survival. What follows organizes the evidence an investor can actually obtain from a prospectus into five dimensions. It is not a scoring model. No weights are assigned, no thresholds are proposed as universal, and the dimensions are not independent of one another. The framework is the authors’ analytical proposal, intended to guide the reading of prospectus disclosures. 
Table 2 summarizes the five dimensions, the indicator each relies on, the signal that warrants further investigation and, equally important, what the indicator cannot establish on its own.
Picture
Table 2: A Framework for Assessing IPO Financial Risk

Conclusion: Capital Raised Is Not Value Created
​

An IPO is a financing event, not a verdict. The same transaction can mark the moment a company secures capital it could have raised elsewhere, or the moment it transfers to public shareholders a funding requirement it cannot meet on its own, and the two are not distinguishable from the headline figures. Profitability at listing has moved with the sectors admitted to the market as much as with the condition of the issuers. Proceeds measure money raised, not the returns that money will earn. Historical IPO underperformance does not, by itself, establish that financing constraints caused weaker returns. 
What survives is narrower, and more useful, than a rule. Liquidity, cash generation, commitments measured over a matched horizon, the use of proceeds and the price paid can each be assessed separately, from disclosure, before an allocation is accepted. Where those readings agree, they describe the transaction well. Where they disagree, as in the largest offering now pending, the disagreement is itself the finding: it identifies the assumption the investment actually depends on. The ability to complete an offering establishes access to capital. It establishes nothing about the business that will have to earn it back. 
​By Maria Carolina Baroni, Leo Shasha, Davide Franchini


Sources:
  • U.S. Securities and Exchange Commission (SEC EDGAR)
  • Jay R. Ritter - University of Florida
  • Harvard Law School Forum on Corporate Governance
  • The Journal of Finance
  • Journal of Financial Economics
  • The Review of Financial Studies
  • Journal of Financial and Quantitative Analysis
  • Finance Research Letters
  • National Bureau of Economic Research (NBER)
  • Centre for Economic Policy Research (CEPR)
  • WilmerHale
  • PwC
  • Renaissance Capital
  • Bloomberg
  • CNBC
  • Reuters
  • The Register

Contact us at [email protected]
Made by Bocconi Students Capital Markets
  • ABOUT US
  • Team
  • ARTICLES
    • Americas
    • APAC
    • EMEA
    • BSCM Analyses
  • Events
  • Alumni
  • JOIN US