# The Race to Become Infrastructure

_Three moves this season, an Anthropic listing, a Stripe bid for PayPal, and BlackRock’s push past fifteen trillion dollars, are versions of the same wager: that scale can be converted into control. A markets briefing._

**Category:** Strategy  
**Author:** Jaime Garcia, Founder, SnowRock  
**Published:** 2026-07-30T07:45:00-04:00  
**Reading time:** 18 min  
**Canonical URL:** https://snowrock.com/insights/evaluation-product

## In short

- Each of these companies is trying to convert scale into control, over compute and talent, over payment flows, over the plumbing of global capital.
- Scale is both the advantage and the trap. The same size that lets these firms distribute faster also concentrates the largest risks in one place.
- For an operator, the lesson is about dependence: when your suppliers race to become infrastructure, their ambition to be impossible to leave is the point, not a side effect.

Three of the most consequential moves in the market this season are, underneath the specifics, the same move. Anthropic is preparing to sell shares to the public. Stripe is bidding for PayPal. BlackRock has pushed past fifteen trillion dollars under management. Each is an attempt to convert scale into control.

## Anthropic Moves Closer to a Landmark Public Offering

Preparations for a possible fall listing are accelerating, but the artificial-intelligence developer would enter the public markets carrying extraordinary expectations, unresolved political exposure and an increasingly competitive economic model.

The company has begun taking several steps commonly associated with the final months of an I.P.O. process. Its advisers are preparing meetings with prospective investors. It is discussing additional credit facilities with banks that could later serve as underwriters. And it has already submitted confidential documents for a public listing. Taken together, those preparations suggest a market debut as early as this fall.

Such a listing would be significant even by the inflated standards of the artificial-intelligence economy. Anthropic is one of a small number of laboratories with the capital, technical talent and computing infrastructure required to develop leading general-purpose models, and a successful offering could value it at a level once reserved for the world’s most established technology businesses. It could also place Anthropic ahead of its closest private-market competitors.

OpenAI has reportedly begun its own confidential preparations but is considering a later listing. DeepSeek, the rapidly expanding Chinese developer, is evaluating a possible domestic offering as soon as next year. The sequencing matters because public-market capital is not unlimited. The first major independent laboratory to list could capture investor demand before competing offerings arrive, establish a valuation benchmark for the sector, attract institutional portfolios and potentially secure a lower cost of capital for the next stage of model development.

Yet timing alone will not determine which company ultimately creates the most value. Technology history contains many examples of businesses that entered the market after their competitors and nevertheless produced far stronger returns. Lyft went public before Uber, yet Uber became the more valuable company by building a broader platform and improving its economics. The same principle applies here. Investors may reward the first major listing initially, but long-term performance will depend on revenue quality, customer retention, capital efficiency and the ability to preserve technological differentiation. For Anthropic, those questions remain unsettled.

## The economics behind the AI M&A wager

Anthropic occupies a powerful but expensive position. Its models have gained traction among software developers, enterprises and technically sophisticated users, and the company has distinguished itself through an emphasis on model reliability, coding performance, enterprise adoption and safety-oriented development. But building frontier systems requires enormous amounts of capital. The largest laboratories must continuously purchase computing capacity, secure advanced chips, train new models, recruit scarce talent and operate increasingly complex infrastructure, and each new generation can require billions of dollars before producing a clear commercial return.

A public offering would give Anthropic access to a substantially deeper capital pool, which it could use to fund model development, expand data-center capacity, make strategic acquisitions or reduce dependence on private investors and commercial partners. It would also provide liquidity to employees and early shareholders, which matters in an industry where compensation frequently includes private equity whose value cannot easily be realized. In a single step the listing would strengthen the balance sheet, establish a liquid valuation for the shares, improve the company’s ability to compensate and retain employees, create acquisition currency, and let it finance a capital-intensive contest without returning repeatedly to private investors.

The scale of the prospective offering, however, could create its own burden. A valuation approaching a trillion dollars would imply that Anthropic is not merely a successful software company but a business on a path toward becoming one of the largest economic institutions in the world. Public investors would expect correspondingly extraordinary growth. Anthropic would need to demonstrate that demand for its models can expand faster than infrastructure expenses, that enterprise customers will remain loyal, and that competition will not force prices downward more quickly than usage rises. Those are formidable requirements.

## A Market Defined by Falling Prices

Anthropic’s greatest commercial challenge may be the speed at which artificial-intelligence capabilities are becoming less expensive. The leading laboratories release improved models frequently, and each release pressures competitors to raise performance, cut prices or both. Large customers can distribute their workloads across several providers, which limits the pricing power of any single company.

OpenAI remains a powerful rival with extensive consumer recognition, enterprise partnerships and a broad product ecosystem. Google can integrate its models into search, cloud computing, productivity software and mobile operating systems. Meta continues to support more openly available models that reduce the cost of deploying artificial intelligence independently. Chinese laboratories add further pricing pressure, and open-source, lower-cost models can be especially attractive to companies that do not require the highest available performance. Many customers care less about which laboratory wins a benchmark than whether a model is accurate, responsive and affordable enough for a specific business function.

This raises the possibility that advanced models gradually become commodities. Anthropic must therefore build defensibility beyond model quality, through enterprise integrations, developer tools, security features, proprietary workflows, customer data, switching costs and deeper participation in the software-development process. Its Fable system is central to that strategy. If Anthropic can become an indispensable operating layer for developers and companies, its commercial position will be far stronger than if it remains primarily a seller of model access.

Public investors will need to decide which kind of company Anthropic is becoming.

## Political and Regulatory Exposure

Anthropic’s preparations are also unfolding amid unresolved conflict with the federal government. The company remains engaged in litigation over its classification by the Pentagon as a supply-chain risk to national security, a designation that can damage commercial relationships, restrict access to government contracts and create uncertainty for customers in regulated or defense-related industries. It has also faced Commerce Department restrictions affecting the release of its Fable model; although it secured concessions that allowed the product to return to the market, the episode exposed its vulnerability to policy decisions outside its control.

Its relationship with the current administration remains difficult to predict. Artificial-intelligence companies operate at the intersection of national security, economic policy, censorship concerns, data-center development and strategic competition with China, and any laboratory seeking public capital must explain not only its product strategy but its exposure to government intervention. Investors will likely examine whether Anthropic can keep selling to federal agencies and defense contractors, whether export controls could restrict its access to chips or international markets, whether state-level regulation will raise compliance costs, whether disputes with the administration could affect partnerships, and whether the company’s safety policies will collide with government demands.

These risks are unusual for a conventional software offering. Anthropic is not simply entering a competitive industry. It is operating in a sector that governments increasingly regard as strategically important infrastructure, and political relationships may therefore influence valuation nearly as much as technical performance.

## Why Anthropic May Still Move First

Despite the uncertainty, Anthropic has strong reasons to proceed before its rivals. The public markets have reopened to large technology offerings, and investors remain eager for direct exposure to artificial intelligence, particularly businesses positioned as foundational providers rather than peripheral beneficiaries. A listing this fall could capture that demand while enthusiasm remains elevated.

Waiting carries several risks. The investment cycle could weaken. Volatility could return. A competitor could list first and absorb a substantial share of institutional demand. A disappointing debut could depress valuations across the sector. Anthropic may also believe that its current growth trajectory presents the strongest possible story, and that high revenue expansion, improving model performance and a favorable narrative create a narrow window in which a company receives a valuation it could not secure under less optimistic conditions.

An initial public offering is as much an act of timing as of financing. The company must enter the market after proving sufficient scale but before growth inevitably slows. It must raise enough capital to sustain expansion without asking public investors to accept an implausible valuation. And it must demonstrate technological leadership without suggesting that each new model requires unlimited spending. The next several months will show whether Anthropic has found that balance.

## Stripe’s Bid for PayPal Tests the Limits of Fintech Consolidation

A private payments company is attempting to acquire one of the industry’s earliest public giants, a rare transaction that would unite complementary strengths while imposing substantial financial and integration risk.

Stripe’s proposed acquisition of PayPal would reverse the usual order of corporate succession. PayPal helped establish digital payments as a mainstream category. Stripe emerged later, building infrastructure that let internet businesses accept payments with greater ease, and the younger private company has now grown large enough to pursue the public incumbent.

Stripe and Advent International have reportedly offered approximately $53 billion for PayPal, valuing it at roughly $60.50 a share, a 28 percent premium to its market price before reports of the approach became public. The bidders have assembled roughly $50 billion in committed bank financing.

The transaction would be unusual in several respects. Privately held companies rarely attempt to acquire public businesses of comparable historical significance; Stripe, however, has reached a reported valuation near $159 billion, giving it greater financial capacity than most private enterprises. It would also combine two large but differently positioned systems. Stripe processed roughly $1.9 trillion in payments last year, with strength in infrastructure for online businesses, software companies and digital platforms. PayPal processed roughly $1.79 trillion and maintains stronger consumer recognition, a large wallet network and the Braintree division.

Stripe would gain PayPal’s consumer accounts, merchant relationships, brand recognition and cash flow, and could integrate those assets with its own developer-oriented infrastructure and global processing systems. Advent would bring experience in financial-technology transactions, including its acquisition of Nuvei and its investment in the Brazilian payments company EBANX. The industrial logic is plausible. The price may be more controversial.

## Why the Offer Could Be Rejected

PayPal has lost momentum in recent years. It faces intense competition from Apple, Google, Stripe and other providers, and its historic advantage as a trusted online checkout button has weakened as consumers move to mobile wallets and embedded payments. Management instability has added to investor concern; the company announced plans to replace its chief executive with Enrique Lores, the former head of HP, and has retained advisers to evaluate strategic alternatives. Those conditions make it vulnerable to an approach. They do not necessarily make the current offer sufficient.

PayPal’s shares have traded substantially above the proposed price during the past five years, and some analysts have called the bid opportunistic, questioning whether it adequately compensates shareholders for the company’s remaining assets, customer base and recovery potential. The market’s initial reaction revealed uncertainty, with PayPal shares closing below the offer price, a sign investors did not regard completion as assured.

Several outcomes remain possible. The board could reject the proposal as inadequate. Stripe and Advent could raise their offer. Another bidder could enter. Or PayPal could remain independent while pursuing a restructuring under new leadership. A counterbid from a large technology or financial company would complicate matters further; SpaceX has been discussed as a possible participant because it seeks a stronger position in payments and has access to substantial public-market capital, and Elon Musk’s history as a PayPal co-founder would lend such an approach symbolic weight, though symbolism alone would not establish strategic logic.

The board’s central question is straightforward: whether PayPal is worth more as an independent turnaround or as part of a larger platform. The answer depends on how much of its decline is structural. If the difficulties stem primarily from weak execution, new ownership could restore growth. If the core products have been permanently displaced, the offer may represent an attractive exit before further erosion.

## The Financing Risk

Even with committed lending, the acquisition would be financially demanding. A heavily financed transaction could impose substantial interest expense on the combined business, and Stripe would need to keep investing in its own technology while integrating PayPal’s infrastructure, customers, employees and regulatory obligations. Payments companies also operate under extensive legal and compliance requirements across many jurisdictions, and integrating systems without disrupting transactions or exposing customers to fraud would be technically difficult.

Success would depend on whether Stripe could extract value without damaging the assets it was buying. The upside is real: consolidating overlapping infrastructure, cross-selling Stripe services to PayPal merchants, expanding Stripe’s consumer presence, integrating Braintree more deeply, reducing duplicated administrative expense, and using PayPal’s cash flow to support long-term development. The failure modes are equally clear. The companies could lose customers during migration. Their systems may be harder to combine than expected. Regulators could demand concessions. Debt costs could reduce flexibility. And management attention could drift from innovation toward a multi-year integration.

## BlackRock’s $15.3 Trillion Expansion

The world’s largest asset manager is no longer relying primarily on index funds. Its next stage of growth is being built through active management, private credit, alternative assets, technology and the gradual conversion of securities into digital instruments.

BlackRock now oversees roughly $15.3 trillion in client assets, a scale that exceeds the annual economic output of nearly every country, and the figure reflects more than rising markets. During the second quarter, clients placed a net $192 billion with the firm; first-half inflows reached a record $321 billion, lifting organic growth in base-fee revenue by roughly 8 percent for the quarter.

BlackRock’s expansion illustrates the extraordinary economics of asset gathering. An investment manager does not need to own the capital it supervises; it earns recurring fees for allocating, administering and providing technology around that capital, and as markets rise and new money enters, revenue can increase without a proportionate rise in operating costs. BlackRock has become the dominant practitioner of that model. Its original ascent was built heavily on index funds and exchange-traded products, which remain foundational, but the company is deliberately expanding into higher-fee categories. Actively managed strategies drew roughly $53 billion in second-quarter inflows; private markets and liquid alternatives took in about $22 billion, up 50 percent from the prior year.

The firm increasingly describes itself not simply as an index manager but as a combined public-markets platform, private-markets investor and financial-technology provider. That distinction matters because traditional index products are highly scalable yet face persistent fee pressure, while private assets, active strategies and technology services generate more revenue per dollar of client capital. BlackRock’s objective is to preserve the scale of passive investing while layering higher-margin products around it.

## The Push Into Private Credit

Private credit has become one of the most important components of that expansion. The market has grown as banks reduced certain forms of lending and institutional investors searched for higher returns; private-credit managers provide loans directly to companies, often under agreements less liquid and less transparent than publicly traded debt. BlackRock strengthened its position by paying roughly $12 billion for HPS Investment Partners, gaining a major presence in private lending and expertise that would have taken years to build organically. During the quarter it reported roughly $6 billion in private-credit inflows.

The opportunity is substantial. Corporations need financing. Pension funds, insurers and wealthy investors seek yield. Banks face capital requirements that limit balance-sheet lending. Asset managers can step between those groups and earn fees for originating and managing loans. But the risks are becoming more visible. HPS restricted investor withdrawals after a surge of redemption requests, demonstrating a central vulnerability of private assets: funds may promise periodic liquidity while holding investments that cannot be sold quickly without steep discounts. When many investors try to withdraw at once, the structure comes under pressure.

BlackRock’s scale may help it manage such episodes, but it does not remove the underlying mismatch. Private credit has grown rapidly during a period in which losses have stayed relatively contained; a sustained downturn would test underwriting quality, recovery values and the ability of funds to meet redemptions. The business can generate attractive fees in stable markets. Its durability will become clearer when defaults increase.

## Tokenization as the Next Infrastructure Layer

BlackRock is also positioning itself for a potentially more fundamental change: tokenization, which converts ownership claims on assets such as stocks, bonds or funds into digital units recorded on programmable infrastructure. In principle this could let securities trade continuously, settle more quickly and be divided into smaller ownership interests, altering parts of the market that have changed little in decades.

Traditional securities transactions pass through brokers, exchanges, clearinghouses, custodians and settlement systems. Each performs an important function, but the process can be slow and operationally complex. A tokenized asset could embed ownership records and transaction rules directly into a digital system, so that trades settle in minutes rather than days, fractional ownership broadens access to assets that are hard to divide, and markets operate beyond conventional business hours. BlackRock is among nearly forty firms planning to tokenize securities held through the Depository Trust & Clearing Corporation, a signal that the idea is moving beyond experimental cryptocurrency markets into the core infrastructure of institutional finance.

The opportunity for BlackRock is not merely to invest in tokenized assets. It is to become one of the institutions that administers them. The firm already operates investment products, risk-management systems and market infrastructure used by financial institutions worldwide, and if tokenization expands it could provide the funds, custody relationships, technology and distribution around the new architecture. Its greatest advantage may be trust: financial innovation often succeeds only after established institutions make it acceptable to conservative investors, and BlackRock has the scale and credibility to normalize technologies that would otherwise stay confined to specialized markets.

## The Institutional Concentration Question

BlackRock’s growth also raises broader questions about concentration in global finance. At $15.3 trillion, the firm influences capital allocation across public companies, governments, private businesses, infrastructure and credit markets; its technology supports other financial institutions, and its funds hold significant positions throughout the corporate economy. That scale produces efficiency, liquidity and broad access. It also creates dependence.

When a small number of asset managers supervise a large portion of global savings, their decisions can shape corporate governance, market structure and the availability of financing. Even passive investments carry influence, because the manager votes shares and determines how funds are constructed. BlackRock does not own its clients’ assets in the conventional sense; it acts as a fiduciary on their behalf. Still, the concentration of operational responsibility and voting authority within a single institution is economically significant, and its expansion into private markets, lending and tokenized infrastructure will deepen that role. BlackRock is becoming more than the world’s largest asset manager. It is becoming one of the principal operating systems of global capital.

## The Broader Market Shift

The developments surrounding Anthropic, Stripe and BlackRock reflect a common pattern. Scale is becoming both an advantage and a necessity. Anthropic needs public capital because frontier artificial intelligence demands extraordinary investment. Stripe is weighing a transformative acquisition because payments increasingly reward global platforms with deep networks. BlackRock is expanding into new asset classes because its existing scale lets it distribute products more efficiently than smaller rivals. Each is attempting to convert size into greater strategic control.

Anthropic seeks control over computing resources, talent and the infrastructure of machine intelligence. Stripe seeks control over a larger share of global payment flows. BlackRock seeks control over a broader portion of the systems through which capital is invested, lent and transferred. The market rewards these ambitions because technological and financial networks tend to reinforce themselves: more customers generate more data, more capital supports more investment, and greater distribution lowers the cost of introducing additional products.

But scale also magnifies failure. Anthropic could enter the public markets at a valuation its economics cannot support. Stripe could spend tens of billions acquiring complexity rather than growth. BlackRock could discover that private-market expansion introduced risks that were less visible in liquid index products. The institutions best positioned to shape the next era are therefore also accepting some of its largest concentrations of risk. None of these companies is content to participate in its industry any longer; each is trying to become the infrastructure the industry runs on.

## The SnowRock read

We watch these stories from the operator’s seat because the same logic that pushes a frontier lab toward a trillion-dollar listing eventually lands on the mid-market company deciding which AI vendor to trust. The thread running through all three is a caution about dependence. When your suppliers are racing to become infrastructure, the terms they offer today are not the terms you will live with in three years, and the switching costs they are quietly building are the strategy, not an accident.

So the advice we give the operators we work with is the unglamorous mirror image of these headlines. Assume the model layer keeps getting cheaper, and do not architect your business as though any single provider’s pricing or availability is permanent. Own your data and your evaluation, keep at least one credible exit in every critical dependency, and treat a vendor’s ambition to become your operating system as a reason for leverage rather than loyalty. The giants are spending fortunes to become impossible to leave. The quiet discipline that protects a smaller company is making sure it never has to.

---

Published by SnowRock. https://snowrock.com