Value Creation & Ownership
Broad equity ownership, but extreme concentration of market capitalisation in a handful of Big Tech groups.
Where does the USA stand, and where does it break?
Around 62 per cent of Americans own equities, mostly through 401(k) plans and broadly diversified index funds (Gallup 2025) — a level of participation in productive capital that is high by international standards. At the same time, concentration is extreme: the ten largest companies account for around 40 per cent of the S&P 500's market capitalisation (S&P Dow Jones Indices), and the wealthiest one per cent of households owns more equity wealth than the bottom 90 per cent combined (Federal Reserve, Distributional Financial Accounts). Ownership of productive capital in the United States is therefore simultaneously broad and highly concentrated: almost two-thirds own something, but the bulk of the value resides with a narrow elite — and a growing share of that value is held by a handful of AI and platform companies. The top one per cent holds around one-third of total net wealth (Federal Reserve). The homeownership rate stands at approximately 65 per cent, yet property — unlike equities — confers no direct participation in the value creation driven by AI. Capital-funded retirement savings through 401(k) and IRA accounts anchor broad sections of the population in the equity market, but to a highly unequal degree.
When AI raises productivity but replaces labour, the gains accrue as returns on capital to the owners of these systems — disproportionately to those who already own the most. A dual concentration compounds this: wealth is held predominantly by upper-income groups, and the index funds held by broad swathes of the population are themselves invested roughly 40 per cent in a small number of Big Tech companies — the very same companies whose AI is displacing jobs. Those with a 401(k) participate, at least in part, in the upside. Those who own nothing lose twice over: their wages and their share of capital. If the market capitalisation of AI companies continues to rise without a broadening of ownership, wealth inequality deepens precisely at the moment when labour income is weakening as a counterweight. In the United States, the central question of the AI economy is less whether value is being created than who owns it.
Paths for the USA
The United States widens the circle of owners — through seed capital for all (Baby Bonds), automatic savings accounts, a citizen fund, or strengthened employee share ownership plans (ESOPs). This is the lever most consistent with a capital culture, because it organises participation through ownership rather than redistributing returns after the fact. Employee ownership is well established in the United States, automatic 401(k) enrolment is expanding, and Baby Bonds programmes are launching in the first states. The effects materialise slowly, and without subsidies for lower earners participation remains skewed.
Trade-off: broad asset base versus slow impact
Feasibility: Automatic retirement saving and ESOPs are the most realistic building blocks
Competition authorities act against concentration among the large platform and AI companies — through antitrust law, merger control, and structural separation where appropriate. This targets the root of concentration by challenging the accumulation of compute, data, and market capitalisation in the hands of a few actors. Several major antitrust proceedings against technology companies are already under way in the United States. The weakness: cases take years, political will fluctuates, and structural separation does little to change who holds the shares.
Trade-off: competition versus the scale advantages of platforms
Feasibility: Necessary against the concentration of power, but not a participation instrument on its own
The state expands tax-advantaged savings accounts, introduces automatic enrolment in retirement plans nationwide, and supplements them with public subsidies for lower-income earners. This builds on the proven 401(k) system and lowers the entry barrier for those who currently own nothing. States such as California have led the way with automatic retirement accounts (CalSavers) for workers without employer-sponsored retirement plans. Without subsidies, the primary beneficiaries remain those who can already afford to save. The regressive effect of the tax concession persists.
Trade-off: broader participation versus fiscal targeting
Feasibility: Automatic enrolment plus subsidy is the most effective combination
The state secures a share of AI value creation through a fund, a data levy, or equity stakes, and distributes it as a citizens' dividend. This addresses the fault line directly, because it gives the public a stake in precisely the returns that would otherwise accrue to a handful of corporations. Alaska's natural resource fund demonstrates the principle. For the United States, the approach is contentious on economic-policy grounds: public equity participation is seen as interference in the market, and the demands of valuation, governance, and independence are formidable.
Trade-off: Collective participation versus market intervention
Feasibility: Conceptually compelling, institutionally unresolved
Trigger-based phases
Trigger: The share of the largest AI and platform companies in total market capitalisation continues to rise, while wealth inequality reaches new highs.
The AI rally drives up the portfolios of those who already own assets, leaving those without capital behind. Several states are expanding automatic savings schemes and Baby Bonds. The debate on capital participation is gaining momentum.
Trigger: The divide between capital owners and those without assets becomes the central political conflict, as AI profits drive further stock market concentration.
Calls for broader ownership and curbs on market power grow louder. Antitrust proceedings, Baby Bonds programmes, and automatic savings schemes compete as responses. The federal government must take a position.
Trigger: Sustained policy measurably broadens capital ownership or effectively constrains concentration.
The United States is deciding whether the value created by AI belongs to a narrow elite or is shared broadly. The outcome will determine whether the promise of widespread share ownership holds — or whether the wealth divide becomes entrenched.
Real-world pilot programmes
Status: In force (automatic savings enrolment for employees without occupational pension provision)
California requires employers without their own retirement plan to automatically enrol employees in a state-run savings scheme. Further states are following suit. The model demonstrates how participation in capital markets can be extended to those who have hitherto been excluded.
Status: In force since 2023 (seed capital for all children born through Medicaid)
The programme invests capital on behalf of disadvantaged newborns, making it available for wealth-building in adulthood. It is the most concrete attempt to broaden capital ownership from birth and equalise starting opportunities.
Status: Established (several thousand companies in employee ownership)
Employee Stock Ownership Plans make employees co-owners of their company and enjoy tax incentives in the United States, where they are widely used. They demonstrate a market-based path to supplementing labour income with capital income.
What we are watching
The metric measures the degree to which market capitalisation is concentrated in a small number of AI and platform companies. The threshold matters because rising concentration aggregates value creation — and with it capital returns — in ever fewer hands. A tipping point signals growing concentration risk even for broadly diversified index investors. For investors, diversification thereby becomes harder, not easier.
The metric measures how unequally productive capital is distributed. The threshold matters because a further rise in concentration means AI returns flow disproportionately to the top. A tipping point marks the entrenchment of a two-tier structure of capital owners and wage earners. Politically, this increases pressure for participation and redistribution instruments.
This metric shows whether ownership is genuinely broadening. The threshold matters because only rising participation among the lower half makes broad-based participation in AI capital real. An upward inflection would demonstrate that instruments such as auto-IRA schemes and seed capital are working. For society, it determines whether the promise of widespread share ownership is fulfilled.