Listen Get

The Earth Optimization Prize Fund

The pool of money that funds curing diseases whether the treaty passes or not

Abstract

Listen.

99% of humans are legally banned from the best-performing asset class. It is called venture capital. It makes about 17% (95% CI: 13%-22%) a year146. Your retirement fund makes 6.5% (95% CI: 5%-8%) after fees. The gap is not skill. It is a door marked “accredited investors only.” Your government decided only rich people are smart enough to invest in things that make money. This is called investor protection. It has protected you from having money. Thank you.

The Prize Fund walks through the door. Not to beat your 401(k) by a few percent. To take $70 trillion of retirement money your species has trapped in index funds and point it at the companies that cure diseases and raise incomes. And away from the ones billing $400 for a hammer.

There is a game attached. You put money in the pool. You vote for a 1% treaty that sends 1% of military spending to medical research. If the treaty passes and people live longer and earn more by 2040, the pool goes to the voters. You get a world with fewer diseases. If it fails, you get the pool back, which has been compounding, at a multiple of 9.03x (95% CI: 3.77x-20.2x). Your retirement account makes 2.57x (95% CI: 2.14x-3.07x) over the same period.

Either way, you come out ahead. That is the whole thing. People keep asking what the catch is. There is no catch. That is the catch.

This is not a money-market account. The fund holds real companies. They go up and they go down. The 15-year lockup is there so depositors do not panic-sell at the bottom, which is the main thing that ruins normal investors.

Keywords

war-on-disease, 1-percent-treaty, medical-research, public-health, peace-dividend, decentralized-trials, dfda, dih, victory-bonds, health-economics, cost-benefit-analysis, clinical-trials, drug-development, regulatory-reform, military-spending, peace-economics, decentralized-governance, wishocracy, blockchain-governance, impact-investing

The Thesis

The prize pool is not a jar of money on a shelf. It is an investment fund. It has three jobs.

One. Make money.

Two. Pay for the companies curing diseases, extending healthy life, building energy, and raising incomes. The companies that move the two numbers the prize is scored on.

Three. Stop paying for the opposite. Stop paying for the companies whose main product is their lobbying department. Stop paying for the defense contractors charging $400 for a hammer. Stop paying for the firms whose growth comes from buybacks.

Your index fund rewards market cap. Market cap rewards incumbency. Incumbency rewards lobbying. Productivity is optional. The crowd, given pairwise sliders and no lobbyists, points the money somewhere else.

This is a 15-year bet. It has downside. If you need principal protection, use IABs147. That is what IABs are for. This chapter is for skeptics.

Why Venture

Cambridge Associates tracks venture capital back 25 years. The gross return is about 17% (95% CI: 13%-22%). Your index fund makes 6.5% (95% CI: 5%-8%) after fees. The gap is not luck. It is access.

Venture pays more because you cannot pull the money out. That is called a lockup. The 15-year prize period is already a lockup. The premium is baked in.

In March 2026 a company called Fundrise listed a venture fund on the New York Stock Exchange. Anyone can buy shares. The fund is called VCX. It holds OpenAI, Anthropic, Databricks, SpaceX, Anduril, and Ramp. It charges 1.85% a year. This was impossible until recently. It is no longer impossible.

Everything below is measured against VCX.

Where the Extra Return Comes From

The crowd does not pick better individual companies than experts. Picking individual companies is a lottery. You need the one in fifty that returns everything. Crowds are bad at lotteries.

The crowd picks better sectors and better managers than experts.

  • SPIVA148: 88% of active large-cap funds underperformed the S&P 500 over 15 years. The professionals lose to a dartboard. You do not need to be smart. You need to not be worse than a dartboard.
  • Preqin’s venture data: the gap between the best firms and the worst is 5 to 15 percentage points a year. If the crowd merely avoids the bottom half (not finds the top quartile, just dodges the bottom), that is worth 0.5% (95% CI: 0%-1.5%) a year on its own.

That is the whole claim. The crowd does not pick companies. The crowd picks sectors and managers. Money flows to productivity instead of to buybacks.

The Home Bias Tax

Japanese pensions hold Japanese stocks. American 401(k)s hold American stocks. Nigerian pensions hold Nigerian bonds. Every country is betting its own economy will beat the global average. This is arithmetically impossible for all of them at once. It is a poker table where everyone has the best hand.

The IMF and Vanguard estimate this costs about 0.8% (95% CI: 0.3%-1.5%) a year. A Nigerian depositor and a Norwegian depositor see the same pairwise sliders. The money goes where the crowd thinks it will do the most good, regardless of passport.

Six Ways the Fund Can Make Money

No single number. Six choices. Different returns, different fees, different downside.

Strategy Where the return comes from Fees Net per year Multiple by 2040 Downside Honest range on the multiple
Boring venture 11–13% (average venture, like VCX) −1.85% 9.5–11% 3.9–4.8× Moderate 2.2× – 7×
Tokenized with crowd voting 13–15% −0.5–1% 12–14% 5.5–7.1× Moderate 2.5× – 12×
Top-firm access 17–19% −1–1.5% 15.5–18% 8.8–12× Higher 2.5× – 25×
Leveraged 10–12% × 1.8 −2% 16–20% 10–15× Bad in crashes 1.5× – 30×
AI frontier secondaries 18–25% (Anthropic, SpaceX, OpenAI) −1% 17–24% 10–27× Very bad 1× – 50×
Crypto early-stage 15–30% −1% 14–29% 7–48× Very bad 0.5× – 80×

The ranges are wide because they are honest. You get one 15-year window. Historical venture vintages have ranged from minus 5% to plus 25% a year depending on which year you started. Narrow confidence intervals from a single-vintage model are a fiction your species is trained to trust.

The canonical 15.8% (95% CI: 9.25%-22.2%) number in the parameters file sits between the second and third rows. It assumes top-quartile access plus the sector-allocation edge plus the home-bias fix, minus the drag from scale. It is top quartile dressed as a base rate. The honest range is the table above.

Ten Ways to Hold the Money

An investment strategy is not a structure. The structure is the regulated shell that lets strangers put dollars in without anyone going to jail. Some shells cost nothing. Some cost millions. Some can hold only a small amount of money before the government says stop.

# Structure Ordinary people can buy? Time to launch Upfront cost Yearly fees Size limit Note
A Buy VCX on NYSE Yes, today 0 $0 1.85% none Works now. One manager picks the companies.
B Launch your own fund like VCX Yes 1–2 years $2–5M ~1.5% none Slow. Expensive. You control the picks.
C Launch a BDC Yes 1.5–2 years $3–6M ~2.0% none, can use leverage The only shell that can legally use leverage.
D Reg A+ Tier 2 Yes 1–1.5 years $1–3M ~1.0% $75M a year Size limit is fatal.
E Reg CF Yes 3–6 months $200–500k ~0.5% $5M a year 6,000 years to raise $30B.
F Reg D 506(c) Rich people only 6–9 months $500k–1M ~0.5% + carry none Does not fix the access problem.
G Tokenized fund on Securitize or Ondo Rich people now; maybe everyone later 9–12 months $500k–1.5M ~0.5–1.0% none Cheapest full stack. Depends on a bill passing.
H Interval fund Yes 1–1.5 years $2–4M ~1.5% none Gives up the lockup premium for partial liquidity.
I Buy VCX, wrap it in a claim token Yes 3–6 months $100–300k 1.85% + ~0.2% limited by VCX size Fastest way to get VCX on-chain.
J A fund-of-funds holding 20+ existing venture firms Rich people now 1–1.5 years $1–3M ~1.5–2.5% stacked none Sounds great. Fees eat it.

Option J is the one that sounds better than it is. You spread the money across Sequoia, Benchmark, Founders Fund, a16z, First Round, and twenty others. The crowd votes on which ones get more money. It should be perfect. Three boring financial problems ruin it:

  • The fees stack. You pay the wrapper 1 to 1.5%. Then each underlying fund takes 2% a year plus 20% of the profits. The net ends up worse than just buying VCX.
  • The best venture capital firms do not need new money. They say no. The list you can actually assemble is the second and third tier, which drags the returns toward average.
  • The top firms put “right of first refusal” clauses on their limited partners. You cannot turn the fund-of-funds into a tradable on-chain token without their permission, which they do not give.

So Option J is a 10 to 20% slice, not the main vehicle, and only once the pool is big enough to negotiate.

Which Structures Can Deliver Which Strategies

Not every combination is legal. Reading this table is the fastest way to see why the decision has already been made for you in most cases.

Structure ↓ / Strategy → Boring Tokenized + voting Top firms Leveraged AI secondaries Crypto
A. Buy VCX ✓ - partial - partial -
B. Your own fund ✓ partial ✓ - ✓ -
C. BDC ✓ - - ✓ - -
D. Reg A+ capped capped capped - - -
E. Reg CF capped capped - - - -
F. Reg D 506(c) ✓ ✓ ✓ ✓ ✓ ✓ (rich only)
G. Securitize ✓ ✓ ✓ - ✓ ✓
H. Interval fund ✓ partial partial - - -
I. VCX + claim token ✓ - - - - -
J. Fund-of-funds ✓ ✓ partial - - -

Only the BDC can legally use leverage. Only Securitize can carry the tokenized, top-quartile, AI-frontier, and crypto strategies in the same shell. Only VCX works today without any regulatory lift at all.

What $100 Becomes

Failure-case payout on a $100 deposit, for the eight combinations worth considering:

Combination Multiple Dollar payout How hard to build How long
Buy VCX 4.3× $426 trivial 0 months
VCX + claim token 4.3× $426 small 3–6 months
Tokenized voting fund 6.3× $624 medium 9–12 months
Top firms via tokenized 10× $990 hard 12–18 months
Top firms via your own fund 10× $990 very hard 12–24 months
Leveraged BDC 12× $1,188 very hard 18–24 months
AI secondaries tokenized 18× $1,782 hard 12–18 months
Crypto tokenized 25× $2,475 medium-hard 9–12 months

On success, the deposit goes to the voters and the dollar column goes to zero. But the depositor lives in a world $6.74 million (95% CI: $2.62 million-$15.7 million) richer per person in lifetime earnings, which is larger than every number in the table combined.

Even the most boring row pays about four times in. The aggressive rows pay ten to twenty-five times in, with the downside spelled out below. The treaty passing is pure upside. The treaty failing still pays.

How to Build It

Build a barbell. Four slices. Let the crowd rebalance it.

The fund’s job is not to maximize returns. The fund’s job is to survive long enough to pay out. Survival means no single bad year kills the referendum. Plan everything around that.

Slice 1, Boring (60% of the pool). Buy VCX on the New York Stock Exchange. Wrap it in a claim token so depositors on Base get an on-chain receipt. Ship this in 3 to 6 months. Do not rebuild the portfolio. VCX already holds OpenAI, Anthropic, Databricks, SpaceX, Anduril, and Ramp. Wrap it and move on. Budget 9 to 11% a year net, about 4× over 15 years. No legislation needed. No headline risk.

Slice 2, Tokenized (20%). Stand up a Reg D fund on Securitize. Real per-deal voting. Tokenized shares. On-chain accounting. Budget 0.5 to 1% in fees. At launch this is accredited-only; it opens to everyone if the INVEST Act passes. Expect 12 to 14% a year, about 6× over 15 years. This is the slice where the crowd actually gets tested.

Slice 3, Frontier (10%). Buy pre-IPO shares in Anthropic, SpaceX, and OpenAI through the Securitize shell. Expect 10 to 18× over 15 years if the cohort holds. Keep it at 10%. A 70% drop in this slice becomes a 7% bruise to the pool, not a dead fund. Do not go above 10% even if the crowd wants to. This is the one sleeve where you overrule the vote.

Slice 4, Ballast (10%). Park it in Aave or Morpho stablecoin yield, or short-term tokenized Treasuries. Budget 3 to 5% a year. This slice exists to cushion rebalancing and nothing else.

Do not do these things.

  • Do not use leverage. The math looks beautiful until a crash prints “PRIZE FUND DOWN 60%” at the exact moment the referendum needs credibility. Leverage is the enemy of survival.
  • Do not hold crypto tokens. The moment the fund holds anything called a “token,” every government the referendum needs to reach writes the project off. Politically radioactive regardless of returns.
  • Do not build your own SEC-registered fund from scratch at launch. It takes 1 to 2 years and $2 to $5 million before it holds a dollar. By then the referendum is over. Wait until the pool is big enough to absorb the fixed cost, then revisit.

What to let the crowd vote on. Not individual companies. Asking a crowd to find the next Stripe in a pairwise slider is the same power-law error described above. Let the crowd vote on:

  1. How big each slice is. (Subject to the 10% cap on Frontier.)
  2. What sectors the Tokenized slice tilts toward. More biotech, less AI? More energy, less consumer?
  3. Which tokenized funds on Securitize the Tokenized slice allocates to.
  4. Which pre-IPO names land on the Frontier slate.
  5. What counts as Ballast.

The Boring slice passes through VCX unchanged. Fundrise picked the companies. The crowd picked Fundrise.

Timeline.

  • Months 0–6. Ship Boring and Ballast. 80% of the pool is already compounding. Configure the prize contract on Base to accept deposits, mint voting points proportional to each deposit, and route capital to the two live slices.
  • Months 6–18. Launch the Tokenized slice on Securitize. Accredited depositors get the full mix. Everyone else gets Boring and Ballast until the law changes.
  • Months 12–24. Open the Frontier slice.
  • Year 2 and beyond. If the INVEST Act passes, open the full mix to everyone. If it does not, the Tokenized and Frontier slices stay accredited-only indefinitely. The pool still works.

That is the launch configuration.

What it pays. The barbell nets about 12 to 13% gross a year at about 1% blended fees. About 5 to 6× over 15 years. Higher than Boring alone at 4×. Lower than the canonical 9.03x (95% CI: 3.77x-20.2x) headline, which assumes top-quartile access and a fully-grown Tokenized slice. The canonical is the target. The barbell is where you start. The crowd moves the mix between them as the pool grows.

If You Want a Different Shape

If the barbell above does not fit your constraints, pick the row that does.

  • Today, off-chain, retail. Buy VCX.
  • Today, on-chain, retail. VCX wrapped in a claim token.
  • Per-deal voting with on-chain liquidity. Securitize Reg D. Accept that retail access depends on a bill.
  • Manager diversification across the venture industry. Fund-of-funds. Wait until the pool is big enough to negotiate fee breaks.
  • Sovereign control over picks. Your own SEC-registered fund. Wait until you can spend $2–5M up front and 1.5% a year without eating into depositors.
  • Targeting 15%+ a year. A BDC with leverage. Accept 40–70% drawdowns.
  • Retail access with no legislative dependency. Only A, B, C, or H are available.

What Can Go Wrong

Plan for each of these. Do not let a depositor discover any of them from a press release.

  1. One window. You get one 15-year vintage. Historical venture returns range from minus 5% to plus 25% a year depending on the starting year. One shot. Build the pool assuming the vintage could be bad.
  2. Drawdowns. The aggressive slices can mark down 40 to 70% in a bad year. Depositors will see these marks on-chain in real time. Prepare the messaging before the mark, not after. The 15-year lockup stops panic-selling at the bottom, which is the main thing that ruins normal investors.
  3. Fundrise risk. VCX is Fundrise’s picks. If they are wrong, the Boring slice is wrong with them. VCX has no long track record. Monitor their deal flow. Be ready to move the Boring slice to an alternative if Fundrise’s selection visibly degrades.
  4. Legislative risk. The INVEST Act might die in committee. If it does, Tokenized and Frontier stay accredited-only indefinitely. Boring and Ballast still work for everyone. Write the depositor-facing copy so nothing assumes the bill passes.
  5. Tokenized discount. On-chain venture stakes can trade 30 to 60% below their real value in stressed markets. VCX itself trades at a discount to NAV regularly. Do not let depositors confuse the secondary-market price with the fund’s underlying value.
  6. Stacked fees. If you pick the fund-of-funds option, the wrapper fee plus each underlying fund’s 2-and-20 will quietly eat the returns. Model the stack before you commit.
  7. Crowd goes wrong. If the crowd is wrong at the sector level for three to five years in a row (overweighting crypto in 2021, say), the fund underperforms a dumb index. Shorten the rebalance cadence before you argue with the outcome.
  8. Breakeven shift. At a 6× failure payout instead of 11×, the dominant-assurance math still works, but the breakeven probability rises from 0.67% to about 1.3%. Still a winning bet for any referendum with real traction. The number just moves. Quote the new one.

A 40% mark in year 5 is survivable. The fund recovers by year 15. The failure case still pays. Depositors who cannot tolerate a year-5 mark should be routed to IABs instead.

The Incentive Problem

Capital providers get 4 to 25× on treaty failure and zero on success. Voters get paid on success. Put them in the same vehicle and you have capital allocators with a direct financial interest in the referendum losing. That is a short position on your own mission.

The fix is boring. Every dollar deposited mints voting points proportional to the amount, at deposit time, not only on referral. Every depositor now holds both sides. Skeptics still deposit. Believers still refer. Nobody at the table profits from the referendum losing.

Gating deposits on a voting action would work too, but it adds friction at the exact step where you want none.

Laws Moving Your Way

Track these. Your launch plan depends on which ones land.

  • INVEST Act (H.R. 9459, passed the House in December 2025 by a bipartisan vote). Expands the accredited-investor definition to cover professional licenses, education, and an SEC-administered exam. Removes some restrictions on closed-end funds holding private funds, which directly helps structures B, C, G, and J.
  • Fair Investment Opportunities for Professional Experts Act (H.R. 3394, passed 397 to 12). License- and education-based accreditation.
  • Equal Opportunity for All Investors Act. An SEC exam pathway to accredited status with no wealth test.
  • DTC digital-twin program (2026). On-chain versions of SEC-registered securities, which lowers the transfer-agent cost for tokenized shells.
  • Securitize, Ondo, Superstate. Operating SEC-registered tokenized funds today. Talk to them before you design anything tokenized from scratch.

Structures A and I need none of these to launch. Structures B, C, and H get cheaper and broader if the INVEST Act passes. Structures G and J need at least one of these bills for retail access. Plan for both the pass and the fail case.

PRIZE vs IAB vs Conventional Retirement

If the prize fails, your deposit comes back at 4 to 25× what you put in, depending on which strategy the fund targets. Your retirement account makes 2.57x (95% CI: 2.14x-3.07x) over the same period. If the prize succeeds, your deposit goes to the voters, and you live in a world $6.74 million (95% CI: $2.62 million-$15.7 million) richer per person.

Your species has a law called ERISA. It requires the people managing your retirement money to act in your interest, not theirs. A fiduciary cannot put your retirement money into a fund whose best-case outcome means the money goes to someone else. So this is not a retirement account.

Instrument Success Failure Retirement-compatible?
PRIZE deposit Pool routes to voters (deposit forfeited) 4× – 25× depending on strategy No
IAB 272% perpetual revenue share Principal at risk Yes
Conventional retirement 2.57x (95% CI: 2.14x-3.07x) over 15 years Same Yes

If your retirement account wants exposure to the treaty succeeding, IABs147 are the instrument. Standard securities. Nothing exotic. If you want skin in the game and you can tolerate drawdowns, PRIZE is the instrument. A conditional bet on civilization with a very attractive consolation prize.

The Self-Fulfilling Property

A normal investment fund bets on which companies will succeed. The Prize Fund puts money into companies that cure diseases, build infrastructure, and raise incomes. Those are the prize targets.

Allocate well, the targets get hit, the pool goes to the voters. Allocate badly, the fund still compounds and the depositors split it.

Someone will call this circular. It is not. It is a loop. Deposit. Allocate. The world improves. You get paid.

The objection assumes the fund sits there passively. The fund is the largest active capital allocator in human history, pointed at two numbers everyone agrees on.