Actionable insights — Measuring in ounces, buying at a discount (AIA Weekly 8.29.26)

1. Change the unit of account — measure net worth in ounces, not dollars

1. Take any nominal series you care about — household net worth, a salary, a portfolio, a house price — over a long window (25 years or more, so the compounding is visible).
2. Divide it by the price of gold at each end date to restate it in ounces. Gold is chosen not as a trade but as a numéraire: a unit central banks cannot create.
3. Compare the two ratios. The nominal multiple tells you what the statement says; the ounce multiple tells you what actually changed in purchasing power. The gap between them is the debasement, quantified.
4. Use the result as a planning anchor, not a signal — it explains why the arithmetic feels wrong to people who "did everything right," and it sets the case for holding assets that cannot be printed.

• Any nominal milestone reported as progress (record net worth, record index level, record wages) — restate it in ounces before accepting it; a widening dollar-versus-ounce gap is the debasement accelerating.

2. Audit "we can grow our way out of it" by racing the two growth rates

1. Grant the claim's logic first — it is genuinely true that an economy growing faster than its debt shrinks the burden. Attack the data, not the premise.
2. Pull the two series for the same periods: real GDP growth and total debt growth, year by year, and put them side by side rather than citing either alone.
3. Compute the ratio. If debt is compounding at a multiple of output, the claim is arithmetically dead regardless of who makes it or how often.
4. Check whether it is a one-off or a pattern — "it's not a one-year one-off. This is how it is every year" is what turns a bad year into a structural verdict.
5. Add the refinancing overlay: a decade of near-zero-rate issuance now rolling into current yields means interest expense grows even if new borrowing stopped.

• The claim resurfacing near an election or a refunding announcement; a year where debt growth finally drops below nominal GDP growth (the first real evidence the claim could be true); interest expense as a share of revenue and of discretionary spending — Druckenmiller's markers, 4.5% of GDP by 2033 and 144% of discretionary by 2043.

3. Run the seven-indicator currency-crisis checklist — then apply the imminence discount

1. Score the country against the seven conditions that preceded prior collapses: government buying its own bonds · debt/GDP over 100% · interest above 15% of revenue · foreign holders reducing holdings · reserve share declining · financial oversight gutted · political interference in the central bank.
2. Benchmark against the actual historical cases rather than theory (Turkey 2018, Venezuela 2017, Argentina 2001, Sri Lanka 2022) so "meets the criteria" has a comparison set.
3. Then apply the exemption that changes the timeline: reserve-currency status. Roughly 70–80% of world commerce still settles in dollars, which buys years the comparison countries never had. Smith's rule governs — "there's a great deal of ruin in a nation."
4. Separate the two decisions the conclusion can inform. A certain-but-undated outcome sets allocation (own scarcity, plan across a generation); only dated, checkable events set positions. "It doesn't mean it's a tradable situation on Monday morning."
5. Discount anyone selling the collapse as a trade: "it's big clickbait. But is it investable, actionable? When's it going to happen? No one can know."

• Movement in the one mitigating variable: the dollar's share of trade settlement and of reserves. Also central-bank independence stories, foreign official holdings of Treasuries rolling off, and the Fed's own holdings rising again — indicator one turning back on.

4. Buy the physical trust off its discount to NAV — a mechanical entry with a computable edge

1. Decide the theme first (uranium), then choose the least operationally risky vehicle that expresses it — a closed-end trust holding the physical commodity, with no permitting, construction or management risk.
2. Track its net asset value continuously, not its price. The only variable you act on is the spread between them.
3. Set a standing buy band rather than a forecast: at roughly −10% to −15% to NAV, add. You are paying less than the metal is worth, so the edge exists before the commodity does anything.
4. Scale in — "I buy some" — so a wider discount is an opportunity rather than a drawdown.
5. Accept the trade-off explicitly: this will never be the 300%-a-year outcome, and that is the price of removing single-company risk from a theme you expect to take years.

• The trust's premium/discount series itself (published daily); a discount widening on a broad risk-off day rather than on anything uranium-specific — the cleanest version of the setup. Conversely, a persistent premium means the trust is issuing units and the edge is gone.

5. Size the speculative sleeve like a venture fund, not like a conviction bet

1. If you want the tenbagger, accept the asset class's real distribution: "you have to approach it as a private equity investor… like a venture capital firm would."
2. Build 8 to 10 researched names, not one. Underwrite the portfolio, not each position — "with the understanding that 80% of them are going to fail."
3. Expect the return to come from the tail: roughly one does okay, one "goes nuts and gives you a market-beating return," and that single outcome carries the basket.
4. Gate entry on work, not story: "you have to do your research… invest time… understand what's going on. And this is difficult." Exclude the low-quality shells outright.
5. Diagnose the failure mode before it happens — people "get wedded to one story and they go all in," and when that name isn't the winner they quit the sector entirely, forfeiting the distribution that made it worth playing.
6. If you can't or won't do the work, take the other branch: "just buy the metal on discount and sit back."

• Your own sleeve drifting to fewer than ~8 names, or one position growing past the size where its failure is survivable; "story" names that can't survive a technical read; the urge to average down into the 80% rather than let the tail work.

6. Price the incentive, not the deficit — what it takes to sanction a fifteen-year build

1. Start from the supply gap in physical units and dates (pounds per year, by year), taken from an operator's own published forecast rather than a promoter's.
2. Convert the gap into a clock: how long does one new project take from discovery to production in this commodity? If the lead time exceeds the years remaining to the shortfall date, the shortfall is already locked in — "you're already in the window."
3. Reframe scarcity correctly: the constraint is almost never geology ("there's plenty of uranium in the Earth's crust"), it is capital willingness.
4. Underwrite the project the way the sponsor must, listing every risk carried across the full build: permitting, First Nations/community, financing, operational, labour, equipment, cost overruns and contingency, and the host government's politics fifteen years out.
5. Solve for the commodity price at which a rational $3–5bn allocator says yes despite that ledger. That number — not the current spot — is your long-run target. "We're still not there."
6. Add confirmed demand arriving on the other side (reactor programmes) using rules of thumb you can compute: ~500,000 lb/yr per 1,000 MW plus ~1.5 Mlb for the initial core load.

• The first large greenfield project actually sanctioned (the price that triggered it is your revealed incentive price); long-term contract prices rather than spot; reactor build and licence applications converting into fuel demand; permitting timelines lengthening, which pushes the incentive price higher still.

7. Convert a reserve headline into a production schedule — reserves are not barrels delivered

1. When a resource number makes a headline, refuse to treat it as supply. Ask the only question that affects price: how many barrels (or pounds) reach a buyer per day, and starting when?
2. Find a comparable asset with a known history and use its ratio as the reality check — reserves in the ground versus actual daily production, and how many years and dollars it took to get there.
3. Audit the physical chain beyond the reservoir: processing, pipelines, storage, and above all export capacity. A country that cannot load tankers cannot sell oil however much it owns.
4. Convert the announced capital figure into a build schedule, then decide whether the market's reaction is pricing barrels that exist or barrels that are still a decade of construction away.
5. Keep the moral judgment separate from the position: "do I agree with it? No… but this is what's happening."

• Rig counts and service-company mobilisation (the first real signal); export terminal and tanker-loading capacity restored; contracts actually signed versus announced; and the temptation to sell oil exposure on a reserve headline.

8. Read a dominant producer's operational excuses as commercial strategy

1. When a producer with a large share of world supply reports disruptions, ask the incentive question before accepting the explanation: what does this producer gain if output stays below plan?
2. Check the contract structure. Where legacy contracts roll off into a rising market, a temporary shortfall converts directly into higher realised prices for years — the producer is paid to be constrained.
3. Note whether the stated causes are conveniently unfalsifiable from outside (reagent shortages, ground conditions) and whether they recur.
4. Split the conclusion in two: bullish for the commodity, cautious on the equity — a state owner who benefits from withholding is not aligned with a minority shareholder.

• Repeated guidance cuts from the same producer with different stated causes; term-contract prices rising while the producer reports difficulties; state ownership or renationalisation talk — the tell that commodity exposure should be taken somewhere other than that company's shares.