Why Trillion-Dollar AI and IPO Valuations Are Not Irrational: A Money Supply and Market Structure Perspective

Executive Summary

As SpaceX targets a $1.75 trillion IPO, Anthropic approaches a $900 billion debut, and OpenAI eyes a $1 trillion public listing, a predictable chorus of skepticism has emerged: these numbers are too big, the bubble is about to pop, the AI market is over-hyped. This talking paper argues the opposite — that these valuations, while historically large in nominal terms, are contextually modest when measured against the transformed financial environment in which they exist.

Trillion-dollar valuations are a small fraction of the capital pool from which they draw — a pool that has been fundamentally resized by monetary expansion, market consolidation, and compounding corporate earnings.

Three structural forces have fundamentally changed what a "large" valuation means in 2026: the dramatic expansion of the money supply, the sharp decline in publicly traded companies competing for investment capital, and the accelerating pace of corporate earnings growth. Together, they suggest that comparing today's valuations to those of prior eras is a category error — like using a 1990s map to navigate a city that has been rebuilt.

I. The Expanding Pool: Money Supply Transformation

The most fundamental and underappreciated driver of asset valuation is the sheer quantity of investable dollars in circulation. Critics of trillion-dollar IPO valuations typically compare them to precedent in absolute terms, ignoring that the monetary base from which those valuations are priced has grown dramatically.

The M2 Money Supply: A Structural Shift

U.S. M2 money supply — which includes cash, checking deposits, savings accounts, and money market funds — has undergone a generational transformation. As of early 2026, M2 stands at approximately $22.7 trillion. To put this in context:

Period	U.S. M2 Money Supply	Key Driver
January 2000	~$4.6 trillion	Pre-dot-com baseline
January 2010	~$8.5 trillion	Post-GFC QE begins
January 2020	~$15.4 trillion	Pre-pandemic baseline
February 2026	~$22.7 trillion	QE + fiscal stimulus + rate cuts

Two key facts emerge from this data. First, nearly 30% of all dollars currently in existence were created since January 2020 alone. Second, since the 2008 financial crisis, more than two-thirds of the existing money supply has been created — meaning the majority of all dollars ever printed have entered circulation within the lifetimes of many working investors.

Since 2009, M2 has grown by more than 160%. Since 2009, the True Money Supply (TMS) is up over 206%. Most of this expansion is permanent structural growth, not a temporary anomaly.

The Federal Reserve's return to quantitative easing in 2025 — adding $40 billion in monthly Treasury purchases — has pushed M2 back above its prior peak and continuing higher. This is not a short-term aberration. It is the new monetary reality within which all asset prices must be calibrated.

Global Money Supply: The Larger Picture

The U.S. story is only part of the equation. Global M2, when combining the U.S., Eurozone, China, and Japan, represents a pool of investable capital that dwarfs any prior era. International capital flows into U.S. equity markets — particularly into high-growth technology — mean that AI IPOs are not only drawing from domestic dollar pools but from global liquidity seeking the highest-return risk assets available.

When a company like SpaceX or Anthropic sets a valuation, it is pricing itself within a global capital market where tens of trillions of dollars are actively seeking deployment into growth assets. Seen this way, a $1–2 trillion valuation is not an extraordinary ask — it is a proportionate claim on a proportionately larger pool.

II. Fewer Vessels for More Water: The Decline of Public Companies

A second structural force amplifies the valuation pressure further: while the dollar pool has expanded dramatically, the number of publicly traded companies competing for those dollars has contracted sharply.

The Great Consolidation

At the peak in 1996, approximately 8,090 companies were listed on U.S. stock exchanges. By 2025, that number had fallen to roughly 4,300 — a decline of nearly 47%. The causes are well-documented: private equity buyouts, mergers and acquisitions, stricter regulatory compliance costs, and a cultural shift toward staying private longer (as demonstrated by SpaceX itself, which has remained private until 2026 despite being one of the most valuable companies on earth).

More capital chasing fewer publicly available companies is the mathematical precondition for elevated valuations. This is not a bubble dynamic — it is supply and demand applied to investable securities.

The result is extreme concentration. By early 2026, the top 10 S&P 500 companies account for approximately 41% of the entire index's weight — a level that has essentially doubled from the 18–23% range that prevailed from 1990 to 2015. Nvidia, Microsoft, and Apple alone account for nearly 20% of the S&P 500. Technology as a sector represents over a third of the index, and over 40% when communication services are included.

This is not irrationality. This is the market efficiently channeling an expanded money supply through a reduced number of available investment vessels. When SpaceX, OpenAI, and Anthropic go public, they will represent the first major expansion of the public investable universe in years. Institutional investors — pension funds, sovereign wealth funds, mutual funds — have been waiting for precisely this: large-cap, high-growth, liquid assets to absorb capital that has been sitting in money markets and private vehicles.

The IPO Supply Deficit

Total U.S. IPO fundraising for all of 2025 was approximately $45 billion. The three AI mega-IPOs — SpaceX, OpenAI, and Anthropic — could together raise over $240 billion in 2026 alone. Far from being a sign of excess, this represents a market finally getting the supply it needs to absorb five years of accumulated monetary expansion. The demand has been there; the supply has not.

III. Compounding Earnings: The Fundamental Justification

Beyond monetary and structural factors, the earnings picture provides a genuine fundamental basis for elevated valuations. The critique that AI companies are "loss-making" or "not profitable" misunderstands the trajectory and the structure of these businesses.

Anthropic: A Rapid Revenue Ascent

Anthropic's financial trajectory is arguably the most striking in the sector. Annualized revenue run rate expanded from $9 billion at the end of 2025 to over $30 billion by April 2026 — a more than threefold increase in roughly four months. Approximately 80% of that revenue comes from enterprise customers, creating a high-quality, recurring revenue base. This enterprise skew — compared to OpenAI's roughly 40% enterprise mix — is precisely why Anthropic is described as having the clearest path to profitability among the major AI labs.

SpaceX: Infrastructure at Scale

SpaceX's $18.7 billion in 2025 revenue reflects a diversified business spanning Starlink satellite internet (with over 9,500 deployed satellites and coverage of 99% of inhabitable earth), commercial launch services, and the newly integrated xAI/Colossus data center operations. The Anthropic compute agreement alone — $1.25 billion per month through 2029 — adds $15 billion annually in high-margin, contracted revenue. The adjusted EBITDA of $6.58 billion demonstrates a productive core business beneath the high-growth investment spending.

The Price-to-Addressable-Market Framework

Traditional price-to-sales or P/E ratios are blunt instruments for evaluating platform-stage companies with exponential revenue trajectories. A more instructive lens is the ratio of valuation to total addressable market (TAM).

Company	Target Valuation	Relevant TAM Estimate	Valuation / TAM
SpaceX	$1.75 trillion	~$30 trillion (space economy + AI infra)	~6%
OpenAI	$852B – $1T	~$15.7T (AI economic value by 2030)	~6%
Anthropic	$900 billion	~$8T+ (enterprise AI market)	~11%

At 6–11% of relevant addressable markets, and with revenue trajectories doubling every few quarters, these valuations are aggressive but not unanchored from reality. Compare this to the dot-com era, when companies with no revenue and no clear business model commanded nine-figure valuations based on page views.

IV. The AI Market: Size Relative to Dollars at Play

Critics cite AI market projections as evidence of irrational exuberance. The numbers are large: BlackRock's Investment Institute projects $5–8 trillion in AI-related capital expenditures through 2030. UNCTAD projects the global AI market reaching $4.8 trillion by 2033. Gartner estimates total worldwide AI spending will exceed $2 trillion in 2026 alone, rising to $3.3 trillion by 2029.

But these figures must be contextualized alongside the pool of global investable capital, which is measured in the hundreds of trillions. Global financial assets — equities, bonds, real estate, and alternative investments — are estimated to exceed $500 trillion when combined across asset classes. Against this backdrop:

▸   A $3.3 trillion annual AI spending figure by 2029 represents approximately 0.6% of total global financial assets.

▸   The combined valuation of SpaceX + OpenAI + Anthropic at listing (~$3.7 trillion) represents less than 1% of global financial assets.

▸   AI infrastructure capex of $5–8 trillion through 2030 is roughly 1–1.5% of global assets deployed over five years.

The numbers cited as evidence of a bubble are, proportionally, a modest reallocation of global capital toward the most transformative technology of the era. The pool is not merely large — it is orders of magnitude larger than the claims being made against it.

The analogy is apt: the glass may look impressive until you understand it is being drawn from a pool that dwarfs it by three orders of magnitude. Concern about the glass is understandable; concern that the glass will drain the pool is not.

V. Addressing the Bear Case

Intellectual honesty requires acknowledging the legitimate concerns embedded in the skeptical view. This paper does not argue that AI valuations carry no risk — only that the systemic bubble framing misdiagnoses the structural environment.

Concentration Risk

The top-10 S&P 500 concentration at 41% of index weight is historically extreme. If a few key AI companies disappoint, drawdowns could be sharp. However, concentration driven by genuine earnings growth and dominance in transformative markets differs structurally from concentration driven by speculation and narrative.

Profitability Timelines

OpenAI's $14 billion net loss in 2025 and its $436 billion in total obligations are genuine concerns. Markets are pricing in an aggressive monetization trajectory that must materialize. The counterpoint is that Anthropic's rapid enterprise revenue growth demonstrates that AI can translate to high-quality recurring revenue — the question is timeline, not feasibility.

Monetary Policy Risk

The expanded money supply that supports elevated valuations could contract if the Fed pivots aggressively to fight renewed inflation. History suggests that rapid M2 contraction pressures all risk assets. Investors should hold this risk — but it is a macro risk that applies to all asset classes, not a specific AI sector pathology.

VI. Conclusion: Calibrate the Context, Not Just the Number

The instinct to declare trillion-dollar valuations "too high" is intuitive but analytically incomplete. Valuation is not an absolute measure — it is a claim on future cash flows, discounted at a rate set by the cost of capital, and denominated in a currency whose supply has more than doubled since 2009.

When the money supply doubles, when the number of investable companies contracts by half, and when corporate earnings compound at accelerating rates, the denominator in every valuation calculation changes. A trillion-dollar company in 2026 is drawing from a pool of capital that simply did not exist in 2010. The pool is larger. The glass is proportionately smaller.

SpaceX, Anthropic, and OpenAI represent genuine infrastructure companies at the foundation of the most significant technology transition since the internet — perhaps since electrification. They are not lottery tickets. They are early claims on platforms that will intermediate trillions of dollars of economic activity for decades. Measured against the capital available, measured against the markets they address, and measured against the earnings trajectories they are already demonstrating, the concern is not that these valuations are too high. The concern should be that investors who dismiss them on the basis of nominal size alone may miss one of the most significant wealth-creation events in the history of public markets.

The question is not whether a trillion-dollar valuation is a big number. The question is: big relative to what? Measured against the transformed monetary environment, the answer is: not particularly.

Appendix: Key Data Points

Metric	Value	Source / Context
U.S. M2 Money Supply (Feb 2026)	$22.7 trillion	Federal Reserve H.6 Release
M2 growth since 2009	+160%	Mises Institute / Federal Reserve
M2 created since Jan 2020	~30% of current total	Federal Reserve data
U.S. listed companies (1996 peak)	~8,090	Historical exchange data
U.S. listed companies (2025)	~4,300	~47% decline from peak
Top 10 S&P 500 weight (2026)	~41%	S&P Dow Jones Indices
SpaceX IPO target valuation	$1.75 trillion	S-1 / Bloomberg / Reuters
Anthropic IPO target valuation	~$900 billion	TechJournal / IG International
OpenAI IPO target valuation	$852B – $1 trillion	SEC filing / Reuters
Anthropic revenue run rate (Apr 2026)	>$30 billion annualized	IG International
SpaceX 2025 revenue	$18.7 billion	SpaceX S-1
AI capex projected through 2030	$5–8 trillion	BlackRock Investment Institute
Global AI market (UNCTAD, by 2033)	$4.8 trillion	UNCTAD Technology Report 2025
Gartner AI spending forecast (2026)	>$2 trillion	Gartner / Vention Research
AI projected economic value (by 2030)	$15.7 trillion	PwC / AI Cloudbase

This talking paper is prepared for discussion purposes and represents a macro-analytical perspective. It does not constitute investment advice. All valuations and market projections are sourced from publicly available filings, research reports, and financial media as of May 2026.