Vol. II · 01 — Methodology
One model, three claims to defend.
The financial case rests on three claims: capex is large but bounded, token issuance front-runs that capex, and reactor-cheap energy makes the exports structurally high-margin. This volume defends each with numbers you can audit.
All figures are in USD billions unless noted, real (inflation-adjusted), concept-stage. The model is a deterministic 30-year discounted cash flow with a token-issuance overlay. "Revenue" rows in the cash-flow table are operating cash flow (gross revenue net of opex); capex is shown separately.
Engineering
Physics & systems.
Financial Model
You are here.
Story Bible
Lore, characters, beats.
This is an illustrative concept model, not financial advice, a valuation, or an offer of securities. Numbers exist to show the model's shape and that it closes — every input must be replaced with diligence-grade data before any raise.
Vol. II · 02 — Capex Breakdown
Where the $39B goes.
Two line items dominate — boring and structure — exactly as the engineering volume predicts. The reactor is large but a minority of total spend.
| Line item | Driver | Capex ($B) |
|---|---|---|
| Boring & excavation | TBMs, shafts, muck handling (Vol. I §11) | 12.0 |
| Structural lining & support | Lamé-sized linings (Vol. I §8) | 7.0 |
| Reactor & power island | SMR + steam cycle + grid-tie | 6.0 |
| Maglev spine + pods | Guideway, LSM, fleet | 4.0 |
| MEP / life support / water | Ventilation, scrubbing, MED desal | 4.0 |
| Shell fit-out | Residential + commercial cores | 3.5 |
| Compute & data infrastructure | Liquid-cooled vaults | 1.5 |
| Daylight optics + contingency | Sun pipes, misc. | 1.0 |
| Total capex | 39.0 |
Vol. II · 03 — Opex & Margin
Reactor-cheap energy is the margin engine.
Operating cost is dominated by reactor O&M/fuel, staffing, and maintenance. Because the city makes its own power and reuses ~60% of waste heat, energy — the largest opex line for any data centre, farm, or desal plant — is near-marginal. We model a blended operating margin m on gross revenue:
Vol. II · 04 — Revenue Streams
Five streams, one shared cost base.
Every stream is subsidised by the same cheap energy and shared infrastructure, so each carries high incremental margin. Mix at maturity (whitepaper Table 1):
| Stream | Pricing basis | Share |
|---|---|---|
| Tokenised tunnels | Lease + resale + usage fees | 38% |
| Compute & data | $/GPU-hr, undercut on power | 24% |
| Agritech | Premium year-round yield | 18% |
| IP & licensing | Per-city blueprint licence | 12% |
| Surplus power | $/MWh to surface grid | 8% |
The ramp is occupancy-driven: usage revenue scales with filled strata (whitepaper Eq. 8.2, R = N·f̄). Early years are token-sale heavy; late years are export-heavy.
Vol. II · 05 — Unit Economics
The per-unit numbers under the totals.
The defensible unit-economics claim is narrow and strong: AETHER does not need a pricing premium to win. It wins because its input cost — energy — is structurally below any surface competitor's. Premiums on agritech and habitability are upside, not the base case.
Vol. II · 06 — Token Supply & Treasury
Issuance as construction finance.
Container tokens are claims on commercial-tunnel capacity. Issuance is phased to track completed rings, so supply expands only as real, revenue-bearing space comes online — avoiding the speculative oversupply that breaks most asset tokens.
- Phased supply: tokens minted only against commissioned capacity.
- Treasury: a share of issuance + usage fees funds a maintenance/sinking reserve and buy-back stabilisation.
- Secondary liquidity: holders trade freely; the city captures a small transfer fee.
- Alignment: token value tracks real occupancy and yield, not pure speculation.
These instruments are securities in most jurisdictions and require qualified structuring (offering exemptions, custody, KYC/AML). Out of scope for this concept model.
Vol. II · 07 — 30-Year Cash Flow
The full annual model.
Year-by-year, USD billions. "OCF" is operating cash flow (gross net of opex). "Net" = OCF − Capex. Cumulative crosses zero around Year 13.
| Yr | Capex | OCF | Net | Cum. |
|---|---|---|---|---|
| 1 | −5.0 | 0.0 | −5.0 | −5.0 |
| 2 | −5.0 | 0.5 | −4.5 | −9.5 |
| 3 | −4.0 | 1.0 | −3.0 | −12.5 |
| 4 | −2.5 | 1.8 | −0.7 | −13.2 |
| 5 | −1.5 | 2.7 | +1.2 | −12.0 |
| 6 | −2.5 | 2.0 | −0.5 | −12.5 |
| 7 | −2.0 | 2.4 | +0.4 | −12.1 |
| 8 | −2.0 | 2.8 | +0.8 | −11.3 |
| 9 | −1.5 | 3.2 | +1.7 | −9.6 |
| 10 | −1.0 | 3.6 | +2.6 | −7.0 |
| 11 | −1.5 | 3.8 | +2.3 | −4.7 |
| 12 | −1.2 | 4.4 | +3.2 | −1.5 |
| 13 | −1.0 | 4.8 | +3.8 | +2.3 |
| 14 | −0.8 | 5.4 | +4.6 | +6.9 |
| 15 | −0.5 | 5.6 | +5.1 | +12.0 |
| 16 | −0.8 | 6.2 | +5.4 | +17.4 |
| 17 | −0.7 | 6.8 | +6.1 | +23.5 |
| 18 | −0.6 | 7.2 | +6.6 | +30.1 |
| 19 | −0.5 | 7.6 | +7.1 | +37.2 |
| 20 | −0.4 | 8.2 | +7.8 | +45.0 |
| 21 | −0.5 | 8.8 | +8.3 | +53.3 |
| 22 | −0.5 | 9.4 | +8.9 | +62.2 |
| 23 | −0.4 | 10.0 | +9.6 | +71.8 |
| 24 | −0.3 | 10.6 | +10.3 | +82.1 |
| 25 | −0.3 | 11.2 | +10.9 | +93.0 |
| 26 | −0.4 | 11.8 | +11.4 | +104.4 |
| 27 | −0.4 | 12.4 | +12.0 | +116.4 |
| 28 | −0.4 | 13.0 | +12.6 | +129.0 |
| 29 | −0.4 | 13.4 | +13.0 | +142.0 |
| 30 | −0.4 | 13.4 | +13.0 | +155.0 |
| Σ | −39.0 | 194.0 | +155.0 | +155.0 |
Vol. II · 08 — NPV, IRR & Payback
The headline metrics.
Discounting the annual net series (Table II.3) gives the returns. Computed at an 8% real discount rate:
Vol. II · 09 — Sensitivity Analysis
What moves the answer.
NPV is most sensitive to the occupancy ramp (how fast strata fill) and the operating margin (how cheap the energy really is) — both downstream of engineering execution. Capex overruns hurt but are bounded by the phased, token-funded build.
| Driver | Swing tested | NPV impact |
|---|---|---|
| Occupancy ramp speed | ±2 yr to fill | High (±) |
| Operating margin m | 0.55 ↔ 0.65 | High (±) |
| Discount rate r | 6% ↔ 10% | High (∓) |
| Capex overrun | +25% | Medium (−) |
| Token price p₀ | ±30% | Medium (±) |
| Reuse fraction φ | 0.5 ↔ 0.7 | Medium (±) |
| Power export price | ±40% | Low (±) |
Read: the project lives or dies on filling the city and keeping energy cheap — i.e. on the engineering in Vol. I, not on financial engineering. That is the right place for the risk to sit.
Vol. II · 10 — Siting Economics
Rock quality is a capex lever, not a detail.
Per Vol. I §8, competent host rock carries load by arching and slashes lining demand — the single largest swing in the $19B boring+structure block. The ideal site combines: high-RMR rock, low seismicity, a desalination-friendly water source, and grid proximity for power export. Siting is therefore a financial decision as much as a geological one.
Vol. II · 11 — The Raise
What the first cheque buys.
The capital stack blends a patient anchor equity partner (the "thinks-in-centuries" investor), phased token issuance against completed capacity, and project debt secured on contracted export revenue. The first tranche funds the path to first revenue — not the whole city.
- Phase A (this raise): production whitepaper, full 3D engine, film treatment, site selection, and a single-stratum engineering pilot.
- Phase B: reactor + first commissioned rings; begin token issuance against real capacity.
- Phase C: downward expansion self-funded by issuance + early export OCF.
Vol. II · 12 — Assumptions & Risk
The inputs, on the table.
Principal financial risks: slow occupancy ramp, capex overrun on boring, token regulatory friction, and margin erosion if energy reuse underperforms. Each maps directly to an engineering risk in Vol. I §12 — the financial and engineering risk registers are two views of the same list.
Replace every figure here with diligence-grade data, a stochastic (Monte-Carlo) overlay, and qualified legal/tax structuring before treating any of it as investable. This model demonstrates the concept closes — nothing more.