Socializing the Upside: The Record SpaceX IPO and the Case for a Social Wealth Fund

On 2026-06-12, SpaceX began trading on the Nasdaq under the ticker SPCX. It sold 555.6M shares at $135, raising $75B (with an $11.2B over-allotment option) at an implied valuation of $1.77T, roughly 2.9× the size of the prior record, Saudi Aramco (2019)‘s $25.6B. The stock opened at $150, touched $168.75, and closed at $160.95, up +19.2%. It was, by a wide margin, the largest initial public offering ever conducted.

This piece does two things. Part I reads the deal the way a careful investor would: as a valuation problem, a risk register, and a market event. Parts II through V use the IPO as the anchor for a public-economics argument: that when the public underwrites private enterprise through bailouts, subsidies, and procurement, it should hold a commensurate, durable equity claim on the upside, pooled in a Social Wealth Fund on the Norwegian and Alaskan model. The two halves are separable; you can accept the investment analysis and reject the policy, or vice versa.

---

I · The Event: investment, risk, and market analysis

Is the largest IPO in history priced like a business or like a moment? What does the historical record of mega-IPOs predict for SPCX over the lockup, index-inclusion, and one-to-three-year horizons, and what does an event this size do to the market around it?

The deal and its mechanics

The headline structure is a $75B primary raise (555.6M shares at $135) with an over-allotment (“greenshoe”) option of a further $11.2B. The greenshoe is the underwriters’ stabilization tool: they oversell the deal, then either buy back stock in the aftermarket if it sags (supporting the price) or exercise the option to cover the short if it rises. At $1.77T, SpaceX listed as roughly the seventh-largest U.S. company on day one, above Tesla.

It is the largest IPO ever in nominal dollars by a wide margin, and the league table below confirms it holds the record even after adjusting prior deals for inflation into 2026 dollars, clearing Saudi Aramco (2019) and the late-1990s telecom mega-listings in real terms.

Two structural facts dominate everything that follows. First, control: the offering is built on a dual-class share structure that leaves the founder with voting power far in excess of his economic stake, so buyers of the public float acquire cash-flow rights but not meaningful governance. Second, scarcity: the public float is a sliver of the company. On the share count implied by the $1.77T mark at $135, the floated stock is only about 4.9% of shares outstanding. A near-$1.8-trillion company with a low-single-digit free float tends to be volatile and prone to index distortions, themes that recur in the market-structure section.

Valuation: the core quantitative work

SpaceX reported FY2025 revenue of $18.7B and a GAAP net loss of $4.9B. With no positive earnings, a price-to-earnings ratio is undefined; the $1.77T mark sits at roughly 95× trailing sales. That multiple tells you little on its own because it bundles three very different businesses, so the right approach is to value each piece separately and add them up.

Sum-of-the-parts (SOTP) valuation: think of it as breaking the company into its component businesses, pricing each one against comparable firms, and totalling. SpaceX is really three things. Starlink is the profitable engine: $11.4B of revenue growing roughly 50% annually, $4.4B of operating income, 9M+ subscribers, with more upside from direct-to-cell and global broadband. Launch (Falcon and Starship) is a dominant, cash-generating franchise with steadily falling cost-per-kilogram. xAI, folded in via the early-2026 merger, is a loss-making frontier-AI bet with an $6.35B operating loss on $12.7B of capital spending, valued against private AI marks with a wide uncertainty band.

The interactive builder below lets you set each segment’s revenue multiple. The defensible base case lands at $529B; even a generous bull case reaches only $1.01T. The market’s $1.77T sits above the top of a reasonable sum-of-the-parts range.

Reverse DCF: what you must believe. A discounted cash flow model (DCF) asks what a stream of future profits is worth in today’s dollars, discounting future cash flows because a dollar today is worth more than a dollar in ten years. A reverse DCF flips the question: given today’s stock price, what future profit growth must the company achieve to justify it? Solving for the revenue growth rate (CAGR, compound annual growth rate) that justifies the $1.77T enterprise value, assuming a 20% steady-state profit margin and a 11% discount rate (roughly the rate of return investors expect for the risk), yields roughly 56% revenue growth sustained for a decade. That implies SpaceX becoming one of the largest-revenue companies on the planet by the mid-2030s. It is not impossible, but it is the growth you are implicitly betting on when you buy at today’s price. The grid below shows how that implied growth rate changes as you adjust the margin and discount-rate assumptions.

Scenarios. Three explicit scenarios, each with a stated set of beliefs and an assigned probability: bull, base, and bear. The enterprise values span a wide range, and the probability-weighted average comes to about $718B, against the $1.77T market mark. The gap is the price of the narrative. The bull case (Starlink as the global connectivity layer, Starship collapsing launch costs, xAI a top-tier lab) can reach the market number; the base and bear cases cannot. Buying SPCX at today’s price is a bet that the bull case is closer to the probable outcome than to a tail risk.

First-day pop and underpricing

The +19.2% first-day gain transferred roughly $14.42B of notional value from the company to the investors who received IPO allocations: money “left on the table” relative to where the stock actually closed. Underpricing of new issues is the norm, not the exception, and is usually a deliberate mix of compensation for early investors who provided pricing information and goodwill toward anchor institutions.

What is striking is how ordinary the pop was. Placed in the distribution of mega-IPO first-day returns, SpaceX’s +19.2% sits at only the 53rd percentile, below Snowflake, Airbnb, Reddit, and Alibaba, and barely above the +17% median. The biggest deal in history had a thoroughly middling debut. That is itself informative: the float was small and demand was real, but the deal was priced close enough to the moment’s fair value that it did not explode the way smaller, scarcer floats have.

The longer-horizon base rate is less kind. The post-IPO drift chart aligns each comparable at its listing date and tracks cumulative return over time, with the IPO date as the starting point so different companies can be compared on the same timeline. The mega-IPO record is bimodal and, on average, negative over one to three years. Jay Ritter’s long-run-underperformance evidence is clearly visible: a handful of winners like Arm and Reddit are offset by severe busts like Rivian, Coupang, and Lucid.

Risk analysis

The risk register scores 16 risks on likelihood multiplied by potential impact, rendered as a filterable heatmap. The top of the board is dominated by governance (super-voting founder control, key-person risk), valuation and sentiment (the 56% annual growth expectation leaves little room for error if the AI/space narrative shifts), and financials (the consolidated $4.9B loss, driven largely by xAI’s spending). Customer concentration, specifically the dependence on U.S. government revenue, is both a risk here and the empirical anchor of Part II.

Market structure, flows, and price discovery

Float scarcity is the structural story. With only roughly 4.9% of shares publicly traded at the IPO, three forces will dominate near-term price discovery.

• The lockup waterfall. Insider shares are locked up after an IPO and release in tranches over time: a typical 90-day partial release, the 180-day cliff, employee/RSU tranches, and the residual founder overhang. The 180-day cliff alone represents the single largest supply event, on the order of $703B of notional stock against a thin average daily volume. Research consistently finds large lockup expirations associated with negative price pressure; this is a scheduled, forecastable supply shock.
• Index inclusion. Nasdaq-100 fast-track eligibility could pull in passive demand quickly (index funds that track the Nasdaq-100 would need to buy). S&P 500 entry is gated by GAAP profitability over four trailing quarters, which the $4.9B loss currently fails, so the larger passive bid is likely delayed until SpaceX prints sustained profit. The dashboard sizes the passive demand each inclusion would create.

Broader-market and sector spillovers

Does a roughly $1.8-trillion listing move the names around it? The peer event study estimates abnormal returns for listed space, satellite, and defense peers, plus Tesla as a read-through to the Musk complex, around the SPCX debut. With only the debut session in the window so far, the estimates are necessarily thin; the living pipeline re-estimates them weekly as the event window fills in.

The method: fit a simple statistical relationship between each peer stock and the broader market over a pre-event period. Then, during the event window around the IPO, measure how each stock performed relative to what the market-wide move alone would have predicted. The gap is the abnormal return attributable to the event itself.

Synthesis: bull, base, bear

The base case is that SPCX is a great company at a demanding price. A defensible sum-of-the-parts ($529B) and a probability-weighted DCF ($718B) both sit well below the $1.77T mark. The reverse-DCF demands 56% annual growth for a decade. The first-day pop was unremarkable. The mega-IPO base rate is negative over one to three years. The float is scarce with a large scheduled supply overhang ahead.

None of that says the stock falls. Narratives can stay expensive for years, and the bull case is possible. It says the risk-reward at the debut price is asymmetric to the downside, and that is the honest investment conclusion.

Predictions: the living part

To keep the analysis falsifiable, four explicit, dated predictions are frozen at publication and scored weekly against the realized tape. They are bets, with stated bands, and they will be marked right or wrong in public:

1. 3-month return vs the Nasdaq-100: modest underperformance.
2. 12-month return vs the Nasdaq-100: underperformance, reflecting the valuation gap and lockup supply.
3. 180-day lockup reaction: negative abnormal return around the cliff.
4. One-month peer spillover: slight positive read-through for listed space peers.

The tracker below shows the live price and the scenario DCF fair-value-per-share lines (reference levels, not price targets; the stock currently trades above even the bull-scenario line), plus the scorecard. Today it reads day 1: SPCX at $160.95, +19.2% versus the $135 IPO price, all four calls pending.

Strongest objections to Part I. (1) “A sum-of-the-parts using public-market multiples can’t value a private-quality monopoly.” Fair: Starlink and launch may deserve scarcity premia that no listed comparable captures, which is why the bull case can reach the market number. The point is not that the price is wrong but that it prices the bull case as the base case. (2) “Reverse-DCF is sensitive to assumptions.” True, which is why it is shown as a grid rather than a single number; across the plausible range of margin and discount-rate assumptions, the implied growth stays extraordinary. (3) “Mega-IPO base rates don’t apply to a category-defining firm.” Possibly; survivorship cuts both ways, and the comparable set deliberately includes both the busts and the successes.

What the data cannot see (Part I). Private-segment economics are disclosed at low resolution; the xAI mark is a range, not a number. The lockup schedule and exact float are modelled from typical structures until the filed prospectus is parsed line-by-line. The event study is descriptive, not causal: it cannot separate the SPCX debut from everything else moving markets that week. The predictions are judgements with stated bands, not model confidence intervals.

---

II · The Entanglement

How much of the value capitalized at IPO was underwritten by the public, through contracts, subsidies, guarantees, and spectrum, and what would the public hold today if, in past rescues, it had kept the equity it took instead of selling it back?

SpaceX’s public scaffolding

Pulling SpaceX’s awards from the federal spending database (USAspending.gov) returns $15B in identified contract obligations, split between NASA (Commercial Crew, Commercial Resupply, the Artemis Human Landing System) and the Department of Defense (Space Force and national-security launch). The cumulative timeline below shows the build-up.

An important clarification: contracts are payment for services rendered, not gifts. SpaceX launched the cargo, carried the astronauts, and delivered the capability; the government got what it paid for, often more cheaply than the alternatives. The argument is not that every federal dollar was a handout. It is narrower: the public bore enormous early risk, anchoring demand when no private market existed, funding capability development, granting spectrum and launch access, and captured none of the equity upside when that risk paid off in a $1.77T company. Add the $885.5M FCC RDOF broadband award (granted in 2020, then rescinded) and modest state and local incentives, and the ledger looks like this:

The retroactive counterfactual: what if we’d kept it?

The cleanest illustration of forfeited upside is not SpaceX; it is 2008. In the financial crisis, the public took real equity and warrants in exchange for rescue capital: roughly 92% of AIG, about 60% of General Motors, large stakes in Citigroup and the banking system. Then it sold those stakes back between 2009 and 2014, at the first politically convenient moment.

The standard story is that TARP “made a profit.” On the numbers, that is roughly true: of about $397B disbursed across the major programs, the public realized about $424B, a nominal gain of +$27B. But that framing answers the wrong question. The right question is: what if the public had kept the money invested instead of returning it to general revenue? Reinvest each program’s realized proceeds in a broad market index from its disposition date to today, and the retained portfolio would be worth about $2.30T, meaning the public forfeited roughly $1.88T of upside by selling early. The financial crisis drawdown was followed by one of the longest bull markets in history, and the public, having borne the downside, sat the recovery out.

This is an illustrative counterfactual, not a causal estimate. It assumes the proceeds could have been retained and indexed, ignores that some firms might have failed without ongoing public restructuring, and uses a price index that understates total return. But the direction is not in doubt, and the magnitude is large enough to survive any reasonable haircut. Drag the slider in the dashboard to see how much market underperformance it takes to erase the gap.

The bridge to a rule

Extend the same logic forward to the SpaceX-style pipeline of public support. Had a warrant-priced equity condition attached to the subsidy-equivalent portion of that support, not to arms-length service contracts but to the underwriting embedded in them plus outright subsidies, the public fund would hold a stake worth on the order of $12B in SpaceX alone today. The estimate is deliberately conservative and every assumption is exposed as a slider in the dashboard.

Strongest objections (Part II). Contracts are payment for services, not gifts: correct, and the analysis treats them separately, applying the equity logic only to the subsidy-equivalent slice and to outright support, never to a competitively-bid launch. Government equity distorts restructuring: a real concern, addressed in Part III via non-voting, arms-length stewardship. Hindsight bias: we know markets recovered; the counterfactual is illustrative precisely because no one could have known in 2009, which is why the right design is a standing rule that holds through cycles, not a market-timing bet. Retroactive application raises legal concerns: flagged here, engaged directly in Part III.

What the data cannot see (Part II). USAspending captures identified federal awards but not the full value of spectrum grants, regulatory forbearance, or the option value of being the government’s anchor customer. The TARP counterfactual cannot observe the world in which stakes were retained; bank behavior, dividend policy, and political pressure would all have differed. These are descriptive magnitudes meant to motivate a mechanism, not point estimates of a knowable quantity.

III · The Fund

If the public underwrites private companies through bailouts, subsidies, and procurement, what mechanism lets it hold a commensurate claim on the upside rather than forfeit it, pool that claim in a fund that pays a broad dividend, and do so without the pathologies of state ownership?

The proposal here is an equity condition on public support, with the proceeds held in a Social Wealth Fund. It is one mechanism, specified concretely below. It does not depend on the tax in Part IV; the two are separable policies that happen to be complementary.

It already exists, six times over

The Social Wealth Fund is not a thought experiment. Norway’s Government Pension Fund Global owns about 1.5% of every listed company on earth and runs on a transparent roughly 3%-of-fund spending rule. Alaska has paid every resident an annual dividend from oil royalties since 1982. Singapore (Temasek and GIC), Australia (the Future Fund), and Saudi Arabia (the PIF) all run large public investment vehicles with different mandates and different lessons.

The lessons are clear and they shape the design. From Norway: arms-length professional management, broad indexing rather than picking winners, an ethics council, radical transparency, and a spending rule that keeps the fund perpetual. From Alaska: that a universal, visible dividend is what makes a fund politically durable; citizens defend a fund that pays them. From Saudi Arabia’s PIF: the cautionary tale of concentration and politicization.

The equity condition: the funding mechanism

The rule: when a company receives a bailout, or material subsidy or procurement support above a defined threshold, the public receives a commensurate equity stake, transferred to the fund and held in principle in perpetuity. Specified precisely:

• Trigger and threshold. Direct bailouts and capital injections; loan guarantees (priced as the option they are); grants and subsidies above a defined threshold; large sole-source or cost-plus procurement; tax incentives above a threshold. The line is drawn explicitly between support that should convert to equity (bailouts, subsidies, guarantees) and arms-length payment for services (a competitively-bid launch), where an equity claim is harder to justify.
• Sizing the stake. A transparent, warrant-style formula: the stake’s value is a multiple of the support’s fair value, struck at the support date, exactly the mechanism the 1979 Chrysler rescue and the TARP bank programs already used. The public is compensated like any at-risk capital provider.
• Form of the equity. Non-voting or independently-stewarded shares by default. This neutralizes the “the government will run companies” objection while preserving the economic claim, using the same separation of cash-flow rights from control that SpaceX’s own dual-class structure uses on retail investors.
• Prospective by default, retroactive carefully. The clean version attaches to future support. Retroactivity is presented two ways: as the illustrative Part II counterfactual, and as a defensible “condition of continued eligibility” (you may keep receiving public contracts and subsidies, but new support now carries the condition), which sidesteps the worst of the constitutional takings problem because nothing already granted is seized.

Normal, not radical

Taking equity for public support is ordinary. The 1979 Chrysler federal loan guarantee took warrants, and Treasury profited. TARP took warrants and equity across the banks, and the bank programs returned a profit. The UK took majority stakes in RBS and Lloyds. The novelty in this proposal is not taking the stake; it is retaining and pooling it rather than selling back at the first opportunity, as Part II showed the public has always done.

How big could it get?

The simulator below drives a fund from equity-condition inflows, compounds at a return calibrated to Norway’s long-run experience, applies a spending and dividend rule, and projects fund size and a universal citizen dividend. With a $60B/year inflow at a 5% real return and a 3% payout rule, the fund reaches about $3.04T in 30 years and pays roughly $364 per adult per year. It survives a 2008-style 35% drawdown at year 15 with about $2.55T still standing. The inflow figure is not heroic: it is a small fraction of annual federal procurement and subsidy flows.

A dividend on this scale is modest next to Alaska’s per-resident PFD, but Alaska’s fund is fed by one state’s oil; a national fund fed by the whole economy’s public-support flows operates at a different scale. A durable, broad-based claim on the upside the public already finances is buildable, with parameters drawn from funds that already work.

Strongest objections (Part III). Moral hazard: does guaranteed public equity encourage companies to seek out subsidies? Arguably the opposite: pricing the stake dampens the upside of capture, because the public shares in any windfall the subsidy creates. State ownership distorts markets: answered by non-voting shares, arms-length management, and broad indexing rather than directing firms. Political raiding of the fund: answered by a constitutional or statutory lock and a spending rule, the features that have kept Norway’s fund intact. WTO-subsidy concerns: the condition is a quid pro quo for support, not a new subsidy, and applies symmetrically to all recipients. Retroactive takings: handled by leading with the prospective rule and the “continued-eligibility” variant. The Meidner plan failed politically: which is exactly why the design centers a visible universal dividend (the Alaska lesson) rather than gradual dilutive collectivization.

What the data cannot see (Part III). The simulator is a deterministic projection, not a forecast: real returns vary, inflows depend on the political settlement, and a sequence of bad draws early would change the path. It cannot price the general-equilibrium effects of the rule on investment and firm behavior, which the objections gesture at and which economists disagree about.

IV · The Tax

Most great fortunes consist largely of unrealized capital gains that the income tax base, as currently designed, essentially never sees. This part examines whether you can tax accruing gains in a way that is administrable, liquidity-sensitive, symmetric (it credits losses as well as taxing gains), and constitutionally defensible.

One framing upfront: an accrual tax on unrealized gains is not proposed here as the primary mechanism for a Social Wealth Fund, nor as a definitive policy prescription. It is worth examining as one potential contributing tool because it addresses something structurally distinct from the equity condition in Part III: the incentive for concentrated wealth to compound indefinitely, untaxed and undisturbed. Whether such a tax is ultimately more beneficial than costly to society is an open empirical question, and this piece does not answer it. What it does is map the mechanism, the scale, and the design features that would make it more defensible.

Buy, borrow, die

The mechanism by which the largest fortunes escape income tax is not exotic. Hold an appreciating asset and never sell it. Borrow against it at low rates to fund consumption tax-free, using the asset as collateral (these are called securities-backed lines of credit, or SBLOCs). At death, your heir’s cost basis steps up to market value (IRC §1014), permanently erasing the embedded gain. Income tax is a tax on realization, and the very wealthy can simply never realize. The interactive example below shows the result: on a $1,000 stake compounding for two decades, buy-borrow-die pays $0 in income tax on the gain, while annual accrual would have collected a substantial sum.

The scale

This is not a rounding error. The stock of unrealized capital gains held by U.S. households is on the order of $56T, of which the top 1% holds roughly $24T, about 43% of the total. That is wealth that the income tax base, as currently constructed, essentially never sees.

The two design problems, and how they might be solved

Liquidity (the “dry tax” problem): owing tax on a gain with no cash to pay it. The design splits by asset type. Tradable assets like public equities can be marked to market annually; selling to raise cash is straightforward. For illiquid or privately-held assets, there are two options: a deferral charge at realization (an approach developed by economists Alan Auerbach and David Bradford, which involves paying an interest factor at the eventual sale date that replicates having paid accrual all along, without requiring annual valuation), or an election to pay in kind with shares, settling the tax liability without forcing a sale. In-kind payment removes the liquidity objection entirely, and if the Part III fund exists those shares are a natural destination; but the tax works regardless of where they go.

Symmetry: the feature that makes accrual fair rather than confiscatory. A tax that grabs gains but ignores losses is a heads-I-win, tails-you-lose proposition that overcharges any volatile asset. The design is explicitly two-sided: when marked assets fall, the taxpayer receives a refundable loss credit, or carries it back with an interest factor. The illustration below runs a volatile six-year path: the one-sided tax takes $29 per $100, while the symmetric tax takes only $12. The government becomes a genuine partner in gains and losses, not just a fair-weather one.

Revenue

How much could it raise? The simulator multiplies the taxable unrealized stock by an expected return, a rate, and, critically, an avoidance and migration elasticity and the loss-refund drag that symmetry costs. At the $1B-net-worth tier (the design in the Wyden Billionaires Income Tax proposal) with a 25% rate and central assumptions, net revenue is on the order of $129B/year, roughly $1.29T over a decade. The high-end estimates are sensitive to the avoidance elasticity, which is why that elasticity is a slider you control rather than an assumption buried in a footnote.

Capital, incentives, and what the tax is really for

The standard objection to any wealth or accrual tax is capital flight: tax the wealthy and they leave, taking investment with them. This deserves more careful treatment than it usually receives.

First, public equity is not physically relocatable. A US taxpayer holding stock in an American company cannot move those shares to a tax haven to escape a mark-to-market tax; the liability follows the US taxpayer. Capital flight is most acute for genuinely mobile activities like some financial operations, and least acute for concentrated public equity held by US residents, which is the primary target here.

Second, the empirical record on high-income migration is sobering for the flight hypothesis. Research consistently finds that the behavioral response to high top rates is far smaller than intuition predicts. Professional networks, family, regulatory familiarity, and property all create friction that economic models often underweight.

Third, the lock-in argument cuts both ways. Critics note that accrual taxation removes the ability to defer indefinitely. But the current realization system creates its own distortion: wealthy investors often hold concentrated, undiversified positions not because those positions are optimal but because selling triggers tax. The result is capital frozen in place by the tax code, not by productive purpose. An accrual tax would reduce this lock-in, potentially pushing capital toward more productive uses and improving overall allocation.

The deeper case for thinking about accrual taxation is not primarily about revenue. It is about incentives. Under the current system, the most rational strategy for a very wealthy person is to never sell, borrowing against an appreciating asset to fund consumption without triggering tax. A founder holding $50 billion in a single stock position is not actively deploying that capital; the company already has the IPO proceeds. What the current code creates is a structural incentive to accumulate indefinitely in concentrated positions, untaxed and largely inactive.

An accrual tax, or even the credible threat of one, changes that calculus. Wealthy holders would face an annual cost to holding appreciated assets. Some would pay the tax and continue holding. Some would sell and reinvest elsewhere, paying capital gains at realization and re-entering the productive economy with more diversified capital. Some would pay in kind with shares flowing into public ownership. Some would donate. All of these outcomes involve the capital doing something; indefinite, untaxed accumulation in a single position becomes the less attractive option.

This is sometimes called a Pigouvian framing, after the economist Arthur Pigou, who argued for taxes that correct distorted incentives rather than just raise revenue. The goal is not for wealthy people to simply hand over money; it is to make the indefinite, non-productive accumulation of concentrated positions less attractive relative to alternatives. A secondary effect is that some of those alternatives, notably in-kind payment or realization and reinvestment, generate tax revenue or public ownership. But the mechanism is behavioral as much as fiscal.

The question, as with any Pigouvian instrument, is whether the correction is worth the cost. Valuation of privately-held assets is difficult and contested. Avoidance strategies will evolve. The constitutional path for taxing unrealized gains remains unsettled: Moore v. United States (2024) declined to resolve whether realization is a constitutional requirement for income taxation, and the question is live. These are real risks.

It is also worth noting what the empirical evidence does not show. The Nordic social democracies, which impose among the highest effective rates on capital in the developed world, consistently rank among the most innovative economies by any standard measure. Denmark, Sweden, Finland, and the Netherlands match or outperform the United States on the Global Innovation Index and similar rankings. The assumption that redistributive policy is inherently hostile to innovation is not supported by this evidence; the relationship is more complex and more contingent on what those proceeds fund and how well institutions function.

The honest summary: accrual taxation on concentrated wealth is an interesting mechanism for addressing a real distortion. It could contribute revenue toward a Social Wealth Fund or other programs. Whether it is net beneficial to society over the long run depends on empirical parameters, principally avoidance elasticities and effects on investment, that economists still disagree about. This piece maps the mechanism and describes how it could be designed responsibly. It does not prescribe.

Strongest objections (Part IV). Valuation of private assets: the tradable/non-tradable split plus the deferral charge sidesteps most valuation disputes, since illiquid assets are never marked annually. Capital flight and expatriation: the empirical record on high-income migration finds far smaller effects than intuition predicts; the design uses a narrow, high-threshold base with an exit-tax backstop; and crucially, US-held equity cannot be physically relocated to avoid a US taxpayer liability. Administrability: the base is a few thousand taxpayers and reuses existing brokerage reporting. Constitutionality: Moore v. United States (2024) upheld the Mandatory Repatriation Tax and expressly declined to decide whether realization is constitutionally required; the income-tax framing is the most defensible path, and the live risk is mapped, not hidden. Lock-in and incentives: accrual taxation actually reduces the realization system’s lock-in distortion, which is itself a source of capital misallocation.

What the data cannot see (Part IV). Behavioral elasticities are uncertain and the revenue range is wide; the unrealized-gains stock is estimated from distributional accounts, not a tax return census; and the constitutional question is unsettled. On whether accrual taxation is net beneficial to society, this piece does not take a position. It describes the mechanism, the scale of the base, and the design features that would make it more defensible.

V · The Argument

Why is a public claim on the most powerful productive assets legitimate, not merely useful? The policy parts lean on a normative claim, and it deserves to be argued, not assumed.

Luck versus desert

The intuition behind concentrated fortune is that it is earned: that the founder of a $1.77T company deserves the proceeds because he created the value. There is something to this, but less than the intuition supposes.

Start with the philosophy. Rawls’s “natural lottery” observes that the talents and circumstances that let someone build an empire, including intelligence, temperament, the country and era and family they were born into, are themselves unearned; from behind a veil of ignorance, not knowing where you would land, you would not endorse a distribution that lets the lottery’s winners take nearly everything. Dworkin sharpens it into brute luck versus option luck: we may let people keep the fruits of gambles they chose to take, but not the windfalls of circumstances they did not choose. Even Hayek, no egalitarian, conceded that market rewards track value to others, not moral desert; the market is a useful signal, not a moral scoreboard.

Then the empirics. Robert Frank’s Success and Luck documents how large a role timing and chance play at the extreme tail of outcomes, where tiny initial advantages compound. Intergenerational-mobility evidence (Chetty and colleagues) shows that where you end up is powerfully predicted by where you started: the “self-made” story understates inheritance, networks, and timing. And the folk intuition that the supremely successful could simply do it again runs into regression to the mean and survivorship bias. For every founder who compounded one improbable success into an empire, many equally able people hit the same wall and are never heard from. None of this denies skill or effort. It denies that skill and effort alone account for the tail, and therefore denies that the tail’s rewards are a pure moral entitlement immune to public claim, especially when, as Parts I and II showed, the public underwrote the runway.

”The rich are the job creators”

The strongest practical objection to taxing or claiming on great fortunes is that it kills the goose: the wealthy are the job creators and the engine of investment. The demand-side rebuttal is that jobs are created by customers. A business hires when demand justifies it, and broad purchasing power is what generates demand. The evidence on where job creation actually comes from points not to the wealthy as a class but to young firms: Haltiwanger, Jarmin, and Miranda find that it is firm age, not size, that drives net job creation, with startups and young high-growth firms doing the heavy lifting, not incumbents or the personal consumption of the rich. A Social Wealth Fund that indexes broadly and pays a citizen dividend does not defund that engine; it redirects a slice of the returns to capital, returns the public often helped create, toward the public, while leaving the price system and the incentive to build a successful firm intact.

Economic democracy

The deepest argument is about power, not money. We accept that political power should be democratically accountable. Yet most people spend a third of their lives, half their waking hours, inside firms that are, in Elizabeth Anderson’s phrase, private governments: hierarchical authorities that direct their lives in detail, with few of the accountability mechanisms we demand of the state. Robert Dahl’s A Preface to Economic Democracy asks the obvious question: if democratic accountability is right for the polity, why does it stop at the workplace door? Concentrated ownership of productive assets is concentrated governing power over the people who work with them and the communities that depend on them.

A Social Wealth Fund does not by itself democratize the firm, but it is a step: even non-voting public ownership creates a public claim and a public interest in how concentrated economic power is exercised, and it can be paired with governance channels, worker or citizen representation, of the kind Germany’s codetermination (Mitbestimmung) has run for decades. The empirical literature on codetermination finds broadly neutral-to-positive effects on productivity and stability, undercutting the assumption that any dilution of pure shareholder control is economically ruinous.

Strongest objections (Part V). Nozick’s entitlement theory: if each step in acquiring the fortune was just (voluntary exchange, no fraud), the holding is just, and redistribution violates rights. The response is that the steps were not purely private: public R&D, procurement, spectrum, and rescue underwrote them, so a public claim is not a seizure of purely private product but a recovery of a public contribution. The Hayekian knowledge and incentive argument: concentrated private control allocates capital better than any public body could, and dulling the reward dulls the drive to build. This is the most serious objection, which is why the design keeps the price system, keeps the founder’s upside large, uses non-voting stakes, and indexes rather than directs. Public-choice skepticism: government will capture and misallocate the fund. Answered, as in Part III, by Norway-style arms-length governance, a constitutional lock, and a visible dividend that gives citizens a stake in defending the fund’s integrity.

What remains contested. Strip away the empirical disputes and a real values disagreement remains: how much weight to give desert-based property claims versus equal claims on a partly-collective product. The data can discipline the job-creator and capital-flight questions; it cannot settle whether the tail of fortune is deserved. This piece argues it is largely not, but flags that this is, in the end, a moral judgement on which reasonable people differ, and the policy design is built to be defensible even to those who weigh desert more heavily.

Appendix · Methods, data, and reproducibility

Single source of truth. Every headline figure lives once, in data/facts/ipo_facts.yaml, and is read by both this prose (via an injected spacex_facts.json) and the dashboards. Prose and dashboards therefore cannot disagree on a number. Every figure traces to a primary source in the repository’s SOURCES.md.

Pipeline. Six numbered stages, fetch → clean → model → score → dashboards → facts, run end-to-end in a few seconds. Stage 00 pulls from keyless sources (Yahoo Finance for SPCX and 30-plus peers and benchmarks, FRED for CPI, USAspending.gov for SpaceX’s federal awards), validates each pull, and falls back to the last good snapshot on failure so a bad fetch never overwrites good data. The whole thing re-runs weekly via GitHub Actions and commits any changes, which triggers the site’s existing deploy.

Methods by part. Part I: valuation by sum-of-the-parts, a reverse-DCF that solves for the implied revenue CAGR, and an explicit scenario DCF; first-day-pop percentile against a curated mega-IPO set; market-model abnormal returns (estimation window then cumulated CARs) for the peer event study; a lockup and index-flow supply-demand model; a likelihood-times-impact risk register. Part II: USAspending aggregation plus an illustrative TARP “kept-it” counterfactual (realized proceeds reinvested in a market index from disposition to today), explicitly descriptive, not a causal estimate. Part III: a deterministic fund-growth simulator calibrated to GPFG’s long-run real return and spending rule. Part IV: an accrual-tax revenue model with avoidance and loss-refund parameters, plus buy-borrow-die and symmetry illustrations.

Honesty rails. Every part carries a Strongest objections response and a What the data cannot see note. Curated or illustrative inputs (the comparable set, TARP figures, the wealth-distribution buckets) are labelled as such. Analyst assumptions (multiples, scores, elasticities) are exposed as dashboard sliders rather than hidden.

The living tracker. Four falsifiable Part I predictions were frozen at publication into predictions.yaml; the weekly run scores them against the realized tape and appends to a public changelog, so the calls are graded, not quietly forgotten.