@gregisenberg Greg, i recently came across this website https://t.co/L9654o0sl7
I think most of your ideas can be built here they are doing some cool things.
Thanks for your ideas they are brilliant and building just became easier
@DannyDayan5@DannyDayan5 i am trying to dm you but not able to. Your macro takes are amazing. Could we have a further conversation about a potential opportunity?
Luke Gromen argues the US is hitting a hard constraint where it must choose between keeping the dollar/Treasury market “safe” or actually winning the AI and great‑power race with China, and that this choice will culminate in a physical‑world, grid‑and‑resources crisis around 2026.
The core “impossible choice”
The US cannot simultaneously reshore industry at scale and preserve the real value of Treasuries; moving capital from Wall Street to factories forces rates up.
With debt at about 120% of GDP, the system cannot tolerate sustained long rates near 5% without breaking growth and markets.
Yield curve control (YCC) in some form is therefore inevitable if reshoring and AI capex go ahead, which implies long‑term financial repression and inflation as the release valve.
The post‑1971 model (offshoring + open capital account + foreigners recycling surpluses into Treasuries) enriched DC, global capital, and big banks while hollowing US industry.
The Trump–Bessent reshoring vision flips that script: higher tariffs, lower income taxes, foreign capital into US factories instead of Treasuries, and surplus countries saving in gold, not bonds.
In that world, US citizens and the industrial base win on a real basis, while long‑duration Treasury holders and TBTF banks lose via repression.
Because global capital is finite, every dollar foreign FDI puts into US plants is a dollar not buying Treasuries, adding upward pressure on yields.
The US net international investment position has swung from +10% of GDP in the early 1980s to about –85% today, meaning foreigners now own tens of trillions more US assets than Americans own abroad, and that stock cannot simultaneously fund factories and keep bond yields low.
Gromen’s bottom line: the US must choose “what do you want?”—real industrial capacity and AI leadership, or preservation of the bond market’s real value and dollar reserve status.
He expects, after a period of denial and bargaining, that policymakers will ultimately sacrifice the real value of the dollar and Treasuries to avoid losing militarily and technologically to China.
Grid, AI, and “physical” bottlenecks
The real bottleneck to AI is not chips but the electrical grid, rare earths, copper, silver, and skilled labor—things the Fed cannot print.
China’s power grid capacity surpassed the US around 2008–09 and has since grown by roughly one US‑grid equivalent in the past decade, while US grid capacity has been roughly flat.
Aggressive AI buildout is already running into grid hookup denials for 2030–31 data centers, with key land deals being cancelled because utilities know capacity won’t be available.
US hyperscalers are turning to Saudi Arabia and the UAE for data centers because those countries can build grid and generation faster than the US can.
That foreign siting strategy works near term but hands leverage to partners whose largest trading partner is China, potentially compromising US control over compute in a geopolitical crisis.
Massive grid and resource buildout would require a multi‑year, WWII‑style program: 5–6 years even at “warp speed,” and 10+ years at normal pace.
Doing this fast means very high wage and materials inflation, which collides with a bond market that cannot handle much higher nominal yields at current debt loads.
Gromen calls this “the mother of all crises” because for the first time the key constraint is real capital (grid, metals, engineers) rather than purely financial plumbing that can be extended and pretended away.
He frames it as a transition from decades of crises solved by “mark bad assets at par and print” to one where the limiting factor is physical capacity.
AI capex of several trillion dollars (he cites forecasts of about 3 trillion in US AI investment from 2025–28) will compete head‑on with Treasury supply for funding.
Inflation, deflation, and the 2026 “year zero”
Rising real capital costs and AI/grid capex are inflationary on the production side, raising input costs for power, copper, and critical materials.
At the same time, AI threatens to displace or undercut large numbers of white‑collar jobs, especially administrative and knowledge work, creating deflationary pressure through weak consumption and rising delinquencies.
He sees “real resource repression”: AI projects bidding resources away from households, pushing up energy and commodity prices while compressing consumer demand.
In a highly leveraged system, both inflation and deflation are bad for the bond market: inflation erodes real returns, deflation triggers defaults and forced selling of Treasuries by levered holders.
Hedge funds running basis trades have bought a large share (he cites about 37–40%) of long‑term issuance since 2022 using repo leverage; higher volatility or deflation shocks force them to de‑gross by selling cash Treasuries.
Banks, which now hold large Treasury portfolios due to post‑GFC regulation, face credit losses in deflation and may also have to sell Treasuries to recapitalize.
Foreigners, with trillions in dollar bonds but also dollar liabilities, are forced to sell Treasuries when the dollar spikes in a global dollar squeeze.
Gromen notes that recent “risk‑off” episodes already show this pattern: long yields dip briefly, then spike sharply higher even as equities keep falling, reflecting mechanical Treasury selling.
He expects 2026 to split into “BC” (before the big liquidity event) and “AD” (after “year zero” when authorities formally cap yields and effectively monetize deficits).
His base case is that some time in 2026—possibly as early as Q2—the US will be forced into a formal or de facto YCC/mega‑QE regime that triggers a huge nominal boom but real losses for bondholders.
Gold, dollar, and reserve structure
Gromen is extremely bullish on gold, arguing it is the only asset that wins in both inflation and deflation under current sovereign debt conditions.
In deflation, collapsing tax receipts versus fixed interest and entitlement outlays force governments toward printing or equivalent measures, raising sovereign credit risk and benefiting gold.
In inflation, repression of real yields and bond debasement also favor gold, especially for surplus countries and commodity exporters that need a store of value for finite production.
Central banks have been steadily increasing gold holdings; measured in oil, the gold price has risen from about 6 barrels per ounce in 2007 to over 70, and he expects this ratio to rise further.
He argues that gold is already larger than Treasuries within global FX reserves by market value, and that another few years of buying could make gold the largest single reserve asset ahead of the dollar.
On historical metrics, US official gold at market value is still a much smaller share of foreign‑held US debt than in past episodes, leading him to see room for at least a 3–5× move in gold prices over the coming cycle.
He believes China and Russia hold far more gold than their official data suggest, via state banks and other vehicles, positioning them for a shift toward a neutral reserve asset.
His long‑run vision is a world of fiat currencies floating against a much higher‑priced gold anchor, with the dollar still important but no longer the undisputed sole reserve asset.
He sees US shale oil strength as a partial support for the dollar, but notes that US production growth is highly sensitive to price and capital discipline; sub‑$60–65 oil threatens new capex.
He thinks US policy on oil (mixed signals between “drill, baby, drill” rhetoric and pushes for low prices) reflects the same unresolved “what do you want?” trade‑off between cheap energy and investment incentives.
Bitcoin, tech, and market signals
Gromen has turned near‑term bearish on Bitcoin after being long from sub‑30k, noting that it continues to trade like a high‑beta tech stock rather than a neutral macro hedge.
He flags rising real capital costs, AI credit growth, and bond‑market stress as negative for speculative tech and thus for Bitcoin in the next leg of the cycle.
He is troubled that Bitcoin has failed to make new highs versus gold in this cycle, breaking prior patterns where each cycle saw much higher BTC‑to‑gold ratios.
He points to large OG whale selling, Tether’s increased gold exposure, and growing concern about quantum computing risk as additional headwinds.
Using his technical triggers, he notes that similar breakdowns in the past have led to median drawdowns of around two‑thirds, with no prior false signals, and now expects a substantial decline before reconsidering.
He still views Bitcoin as “the last functioning smoke alarm,” reading current weakness as a warning of rough macro conditions in early 2026 rather than just an idiosyncratic story.
On equities, he expects a very rough “BC” phase where rising real capital costs and liquidity strains hit risk assets broadly, followed by an “AD” phase where nominal indices soar but underperform gold in real terms.
He suggests trades like long US infrastructure (e.g., grid‑related plays) versus short semiconductors over the next few months, reflecting near‑term grid scarcity and chip oversupply, but not as a multi‑year secular pair.
He anticipates extreme financial repression in the endgame—capped long yields, negative real rates, and high inflation—benefiting real assets while punishing holders of long‑duration bonds.
Throughout, he frames policy and market outcomes through a simple lens: double‑entry bookkeeping and real resource constraints will eventually override political rhetoric, forcing the US to choose between bondholders and strategic industrial/AI dominance, and he expects it to choose the latter.
Michael Bury is arguably the worst investor of our generation. Here are 12 of his most smooth brained predictions.
Jan 2017 – Predicted a global financial collapse and WW3 were imminent.
Sep 2019 – Claimed index funds were the next CDOs, ready to implode like 2008.
Dec 2020 – Shorted Tesla, said its price was “ridiculous” and destined to crash.
Jan 2021 – Reiterated that Tesla’s valuation would implode soon. shares kept doubling.
Jan 2021 (late) – Argued GameStop’s rally wouldn’t repeat. it exploded again weeks later.
Feb 2021 – Warned the entire stock market was “dancing on a knife’s edge.”
Feb 2021 – Claimed inflation would destroy Bitcoin. it hit new highs later that year.
Feb 2021 – Called Robinhood a “dangerous casino” it kept growing users and revenue. Up almost 1000x
Mar 2021 – Said Bitcoin was a speculative bubble. it kept rallying.
Jun 2021 – Predicted the “mother of all crashes” and sold everything. market kept going like David goggins
Sep 2022 – Forecasted massive stock failures ahead. indexes finished the year higher.
Aug 2023 – Bet $1.6 B on a total market crash, tweeted “Sell,” later admitted “I was wrong to say sell.” Do we know how much he lost?