# Out of Debt: integrated simulation results

September 27, 2026 · Conditional policy stress test, not a forecast or investment recommendation.

**The plan can reduce debt-to-GDP under favorable fiscal and growth assumptions. The National Resource Fund helps, but it does not deliver the article’s 23% debt-to-GDP result by itself.** That result assumes the 2% debt-growth constraint has already been achieved.

I evaluated 14 named scenarios, 1,296 sensitivity combinations, and 20,000 paired random economic paths under three policies (60,000 policy trajectories). Deterministic horizons are 30 and 50 years; random paths cover 30 years. Nothing here estimates the real-world probability of legislative passage, resource availability, or AI adoption.

## Main results

All paths begin at a normalized 100% debt-to-GDP. The integrated central case reaches **82.7%** after 30 years; maintaining a 2.6% primary deficit instead reaches **144.0%**. A 1% primary surplus by year 10 reaches **59.1%**. These are debt held by the public divided by GDP; Fund assets are not subtracted.

| Scenario | Year 10 | Year 20 | Year 30 | Fund at year 30 |
|---|---:|---:|---:|---:|
| Article target: enforce 2% debt growth | 61.3% | 37.5% | 23.0% | $10.43T |
| Integrated central: primary balance by year 10 | 95.5% | 89.1% | 82.7% | $10.43T |
| Stronger discipline: 1% primary surplus by year 10 | 90.1% | 74.4% | 59.1% | $10.43T |
| No fiscal correction: 2.6% primary deficit | 109.3% | 127.2% | 144.0% | $10.43T |
| Weaker AI: 1.8% real growth | 106.8% | 112.0% | 117.0% | $10.43T |
| Higher financing cost: new yield 8.5% | 107.4% | 119.8% | 134.6% | $10.43T |
| 2% inflation; yields fall with inflation | 102.1% | 96.4% | 89.3% | $10.43T |
| Resource shortfall: $15B annual net receipts | 95.5% | 89.1% | 83.8% | $1.04T |
| Fund financed by borrowing | 98.5% | 93.5% | 87.7% | $10.43T |
| AI gains arrive gradually over 10 years | 100.3% | 93.5% | 86.9% | $10.43T |
| Resource receipts delayed 10 years | 95.5% | 89.1% | 83.4% | $4.74T |
| Finite receipts: contributions stop after year 20 | 95.5% | 89.1% | 82.8% | $8.65T |
| Adverse combination | 128.8% | 178.3% | 244.0% | $0.62T |
| Recession and portfolio crash in years 21–22 | 95.5% | 89.1% | 96.3% | $6.92T |

The first row solves for whatever primary surplus is necessary to enforce 2% debt growth. It is a target identity, not evidence that this fiscal adjustment is achievable. Other rows calculate debt from the budget. All dollar figures in this table are nominal.

## What it takes to reach the article’s 23% target

At the central financing assumptions, the 2% debt-growth target needs a **1.81% of GDP primary surplus in year 1**, approximately **$621B** in this normalized economy. The required surplus peaks at **2.68% of GDP in year 6**. Moving from a 2.6% primary deficit to a 1.81% surplus is a 4.41-percentage-point adjustment, roughly $1.51T using the model’s year-1 GDP.

This is substantially stronger than simply moving toward primary balance. A primary balance covers noninterest spending; it does not cover the interest bill. Primary targets in this model exclude the new resource receipts and Fund transactions, which are accounted for separately.

## Does investing beat reducing borrowing immediately?

The comparison below uses identical macroeconomic assumptions, fiscal adjustment and additional net resource receipts. The no-receipts case has no new resource development; the borrowed-capital case has no matching new resource revenue.

| Allocation | Debt/GDP, year 30 | (Debt − Fund)/GDP | Fund |
|---|---:|---:|---:|
| No new receipts | 83.9% | 83.9% | $0.00T |
| Invest receipts in Fund | 82.7% | 78.6% | $10.43T |
| Use receipts to reduce borrowing | 78.9% | 78.9% | $0.00T |
| Borrow to capitalize Fund | 87.7% | 83.6% | $10.43T |

The central Fund case has more outstanding debt than using resource receipts to reduce borrowing immediately. It has slightly lower net financial liabilities because it owns assets. The difference is only **0.36 percentage points of GDP** in this deterministic comparison. Investing is not an automatic improvement over avoiding Treasury interest costs.

Relative to no new resource receipts, the Fund lowers debt/GDP by about **1.2 percentage points** by year 30 through withdrawals beginning in year 21. Its remaining assets add about **4.1 percentage points of GDP** to the financial balance sheet. Fund assets are risky and are not deducted from the official debt measure. This is a partial financial comparison: unextracted resources, other public assets, environmental liabilities, and strategic benefits are outside it.

## Fund scale, purchasing power and withdrawals

With $150B deposited at each year end, a 7% nominal gross return, 0.25% annual fees and withdrawals of 3% of opening assets starting in year 21, the Fund reaches **$10.43T nominal**, or **$3.21T in starting-year purchasing power**, after 30 years. The year-30 withdrawal is **$297B nominal**, approximately **$92B in starting-year purchasing power**.

Under the article’s simpler accumulation-only arithmetic—7% nominal return, no fees, no withdrawals—the same fixed nominal contributions reach **$14.17T after 30 years** and **$60.98T after 50**. Those figures do not include budget support or establish that the contributions are available.

The 3% withdrawal rule is an illustrative policy, not a proven sustainable rate. With 4% inflation and a 0.25% fee, maintaining purchasing power after a 3% withdrawal requires at least **7.25% gross nominal return** in a smooth, no-new-contributions year. At 7%, real principal slowly shrinks once contributions stop. In the finite-receipts case it falls from $2.73T in year 20 to $2.54T in year 50, measured in starting-year dollars. Volatility adds further risk.

## Randomized stress test

These are synthetic draws from declared assumptions, not empirical forecasts, calibrated confidence intervals or real-world odds of success. Each policy uses the same 20,000 underlying shocks. Seed: 20260927.

| Fiscal policy | Median debt/GDP at year 30 | 5th–95th percentile | Synthetic paths below starting 100% |
|---|---:|---:|---:|
| Primary balance by year 10 | 83.2% | 56.8%–120.5% | 79.6% |
| Persistent 1.5% primary deficit after year 10 | 120.5% | 88.2%–166.4% | 16.7% |

Only 1.6% of the primary-balance draws reach below 50% debt/GDP, and none reaches below 25%. Zero observations in this sample does not establish impossibility. The article’s 23% path requires a materially different fiscal policy, not just a favorable investment draw.

The Fund’s year-30 median is $9.24T nominal, with a 5th–95th range of $4.49T–$20.34T. In the paired comparison, investing produces lower net financial liabilities than immediate borrowing reduction in 43.4% of these synthetic draws. This percentage changes with the assumed risk premium, returns and volatility.

## Assumptions and accounting

The starting GDP and debt are both $32T, chosen as a rounded scale to reproduce the article’s 100% starting ratio. They are not measured current balances. Fund assets start at zero. “Year 30” means 30 model periods, not an official fiscal-year forecast.

- Central real GDP growth: 3% from year 1. Inflation: 4% GDP-price inflation. The alternative AI case ramps from 1.8% to 3% over ten years.
- Starting effective interest cost: 3.3%, an assumption. The new-debt rate is inflation plus 2.5 percentage points (6.5% centrally), with 20% annual pass-through to the average cost of the debt stock. These are stylized financing assumptions, not a fitted Treasury maturity model.
- Primary balance moves linearly from −2.6% to 0% of current-year GDP by year 10, beginning the adjustment in year 1. Alternative rows change the terminal target. Surpluses are positive.
- Additional net eligible resource receipts: $150B each year, fixed in nominal dollars, after restoration costs and existing obligations. The Fund receives all of these receipts. Neither this amount nor resource-project feasibility is established by the model.
- Gross portfolio return: 7%; annual fee: 0.25 percentage points. All earnings remain invested until withdrawals begin in year 21. The 3% payout uses opening assets, capped by assets available after returns; new contributions arrive afterward.
- No taxes or spending are automatically changed by Fund gains. Each withdrawal reduces otherwise-required borrowing dollar for dollar.

For each year, with positive P denoting a primary surplus, N additional net resource revenue, C a Fund contribution and W a Fund withdrawal:

```
GDP(t)  = GDP(t−1) × (1 + real growth) × (1 + GDP-price inflation)
i(t)    = 0.8 × i(t−1) + 0.2 × new-debt yield(t)
Debt(t) = Debt(t−1) + i(t) × Debt(t−1) − P(t) − N(t) + C(t) − W(t)
Fund(t) = Fund(t−1) × (1 + gross return(t) − fee) + C(t) − W(t)
```

When all new resource receipts are invested, N=C: they create Fund assets but do not immediately reduce borrowing. Funding the Fund without new receipts increases borrowing. Withdrawals and contributions are transfers within the consolidated financial balance sheet; the code verifies they cancel from the net-liability identity. Interest on new within-year borrowing is omitted, as are other financing adjustments.

Random-draw specification: long-run real growth is normal with mean 3% and standard deviation 0.5 percentage points per path; inflation is centered at 4% with the same long-run uncertainty. Annual growth and inflation disturbances have stationary standard deviations of 1.5 and 0.8 percentage points, with AR(1) persistence 0.35 and 0.55. Growth is bounded at −8% to 10%, inflation at 0% to 10%. Returns are lognormal with 7% arithmetic mean and 12% log-return volatility before fees. Return innovations load +0.35 on growth innovations and −0.20 on inflation innovations. Net receipts are lognormal with $150B arithmetic mean and 35% log volatility, loading +0.4 on growth innovations. New yields include a 0.5-percentage-point independent shock and an extra 1 percentage point for each 100 percentage points of debt/GDP above the starting 100%. The primary balance changes by 0.4 times the cyclical growth disturbance. All these distributional choices are illustrative, not estimated from historical data. Arithmetic mean return is not compound growth.

The deterministic grid combines real growth of 1.8/2.5/3/3.5%; inflation of 2/4/5%; terminal primary balance of −2.6/−1.5/0/+1%; additional yield premiums of 0/1.5/3 percentage points; net receipts of $15/$150/$200B; and gross Fund returns of 4/7/8%. Grid entries have no probability weights. The recession case uses −3% real growth, 5% inflation, an extra 3% of GDP primary deficit, −30% portfolio return and halved receipts in year 21, followed by a partial recovery in year 22. The full inputs are in the code.

## Federal context and feasibility limits

CBO’s dated February 2026 outlook projects a 2.6%-of-GDP primary deficit for 2026 and 1.8% average real growth over 2027–2036. Those numbers anchor context, not the entire model. Later tariff updates are not incorporated here. [CBO outlook](https://www.cbo.gov/publication/62105), [subsequent updates](https://www.cbo.gov/topics/economy/outlook-budget-and-economy).

Interior reported $14.61B in FY2025 disbursements across recipients, including $5.01B to Treasury. The proposed $150B annual contribution is about 30 times that Treasury allocation and more than ten times total reported disbursements. This comparison does not estimate a future revenue ceiling, but it makes the funding gap explicit. Existing receipts cannot simply be counted again as new cash. [Interior FY2025 report](https://www.doi.gov/node/64051).

The Federal Reserve’s objective remains 2% PCE inflation. This exercise’s 4% GDP-price assumption is a scenario; PCE and GDP-price inflation are distinct. The paired 2% and 4% cases adjust financing yields as inflation changes, so higher inflation is not treated as free debt relief. [Federal Reserve strategy](https://www.federalreserve.gov/monetarypolicy/monetary-policy-strategy-tools-and-communications-statement-on-longer-run-goals-monetary-policy-strategy-2025.htm).

The exercise does not estimate resource reserves, extraction profitability, restoration costs by project, ecological recovery, tribal obligations, permitting feasibility, employment displacement, distributional effects, tax-base shifts, AI adoption costs, exchange rates, asset allocation, commodity concentration, Treasury maturity details, indexed benefits/TIPS, political governance or monetary-policy reactions. The primary-balance and yield assumptions summarize some fiscal effects; they do not prove them. No environmental or strategic benefit has been converted into fictitious financial cash flow.

## What the simulations imply for the plan

1. Treat the 2% debt-growth rule as a fiscal constraint that must be funded, with an explicit budget path.
2. Demonstrate additional net investable resource receipts before relying on $150B annually.
3. Compare each Fund allocation with the interest savings from reducing borrowing, using risk and fees.
4. Separate accumulation from distributions and test real purchasing-power preservation after resource depletion.
5. Present the model as a range of conditional outcomes, including weak growth, persistent deficits and higher yields.

## Reproduction and validation

Run `python3 simulations/run_simulations.py` and `python3 simulations/build_report.py` from the workspace root. The only simulation dependency is NumPy (2.4.2 in this run). The report uses standalone HTML/SVG with no network dependencies.

Six checks passed: the published debt identity, closed-form Fund accumulation, independent compound-interest debt calculation, equal-rate investment/debt-repayment equivalence, annual consolidated cash-flow identities, and the 1,296-case grid count. These check arithmetic and accounting; they do not validate economic forecasts.

Files: `results/results.json`, `results/scenario_paths.csv`, `results/sensitivity_grid.csv`, and `results/monte_carlo_terminal.csv`. The interactive HTML displays precomputed scenario paths; it does not silently rerun or change the Monte Carlo distributions.