The Record

Physical-economy sectors — energy, industrials, materials, defense, infrastructure — will take a structurally larger share of investment and policy attention through 2035 than they did in the prior two decades. If capital spending in these sectors rolls over durably, this thesis is wrong.

OPEN — NOT YET SETTLED

Electricity is the master constraint of the decade; nearly every other system's growth rate will be gated by it somewhere.

OPEN — NOT YET SETTLED

The rebuild is slower than markets want but longer than markets price: multi-decade physical timelines meeting durable, bipartisan funding is the defining mismatch of the era.

OPEN — NOT YET SETTLED

Labor scarcity is permanent enough to make automation a structural theme rather than a cyclical one.

OPEN — NOT YET SETTLED

The connections between these systems will matter more than any single system: the biggest surprises of the decade will emerge at the seams — power prices meeting politics, defense meeting labor, AI meeting the grid.

OPEN — NOT YET SETTLED

Labor scarcity, not labor replacement, will be the dominant driver of U.S. automation adoption through 2030. If manufacturing wages stagnate and vacancies vanish, this thesis is wrong.

OPEN — NOT YET SETTLED

Warehouse and logistics automation is the fastest-compounding segment, because the return on investment is measured in months and the environments are controlled.

OPEN — NOT YET SETTLED

Reshoring at announced scale is almost certainly impossible without automation; the factories that succeed will be the most automated ones, which caps how many jobs reshoring "brings back" per press release.

OPEN — NOT YET SETTLED

Humanoid robots will find real niches this decade but likely stay under a few percent of robot deployments through 2030; general-purpose usefulness remains unproven.

OPEN — NOT YET SETTLED

The reliable economics sit with component and integration suppliers — the sellers of arms, sensors, and software — rather than with any single platform bet.

OPEN — NOT YET SETTLED

Copper demand growth this decade is close to a lock: grid spending, EVs, and data centers are all funded, physical programs already underway. The demand side of this thesis fails only if the entire buildout fails.

OPEN — NOT YET SETTLED

Mine supply is likely to undershoot demand growth *later* this decade, because the projects that could close the gap needed to start permitting years ago and mostly didn't. Note that this is a forecast about the 2030s: the balance for 2026 and 2027 is currently projected as surplus, and a thesis that needed next year to be tight is already wrong.

OPEN — NOT YET SETTLED

Structurally higher real copper prices are more likely than not over the next five years, though with violent cyclical swings along the way.

OPEN — NOT YET SETTLED

Substitution (aluminum in transmission and some wiring) and recycling will cap the upside — recycling is already a live reason the 2026 balance flipped to surplus.

OPEN — NOT YET SETTLED

Permitting reform is the wildcard: a genuine shortening of U.S. mine timelines would change the long-run picture, and is worth watching more than any price target.

OPEN — NOT YET SETTLED

U.S. electricity demand growth is a durable regime change, not a spike. Even if AI spending slows sharply, electrification and reshoring keep the line bending upward.

OPEN — NOT YET SETTLED

Transmission and grid equipment, not generation, will be the binding constraint for most of the next decade. Watch transformer lead times, not power plant announcements.

OPEN — NOT YET SETTLED

Firm power commands a rising premium. Nuclear restarts, life extensions, and long-dated corporate power contracts almost certainly continue, though new large reactors will disappoint on timelines.

OPEN — NOT YET SETTLED

Gas-fired capacity will be built in size through the early 2030s regardless of climate policy, because turbine backlogs, not politics, set the pace.

OPEN — NOT YET SETTLED

Rising retail electricity prices become a first-order political issue, likely inviting intervention in how data centers pay for grid upgrades.

OPEN — NOT YET SETTLED

Space is now economically an infrastructure sector, not an exploration story. The revenue base has shifted decisively toward services (connectivity, navigation, imaging) rather than launches themselves.

OPEN — NOT YET SETTLED

Launch cadence will likely keep rising over the next several years, and reusability means the marginal flight keeps getting cheaper. A sustained reversal in cadence would be evidence this thesis is breaking.

OPEN — NOT YET SETTLED

Defense budgets are almost certain to deepen their dependence on proliferated commercial-style constellations this decade, blurring the line between commercial and military space.

OPEN — NOT YET SETTLED

Concentration is the system's biggest structural weakness: one company flew roughly 85% of U.S. orbital launches in 2025. Genuine second sources of both launch and broadband would make the infrastructure sturdier, and their absence keeps it fragile.

OPEN — NOT YET SETTLED

The durable model is subscription access to routine, chronic, predictable care. Where a condition needs ongoing management, cash-pay platforms compete well; where it needs coordination or acuity, they don't.

OPEN — NOT YET SETTLED

Cash-pay economics beat insured economics at the same price — no denials, collections, or concessions. That margin gap, more than consumer demand, pulls incumbents like the national labs into the channel.

OPEN — NOT YET SETTLED

Retail contraction plus manufacturer-direct distribution is unbundling the pharmacy counter: convenience shifts to shipping, clinical counseling loses its home.

OPEN — NOT YET SETTLED

Data regulation is the largest unpriced risk here. One high-profile failure could reset the category.

OPEN — NOT YET SETTLED

Oral GLP-1s shift the binding constraint from manufacturing capacity to payer willingness. The fight moves from supply headlines to formulary design.

OPEN — NOT YET SETTLED

Employer coverage is the least stable link, because employers bear the cost and rarely capture the benefit. Coverage fragments by indication rather than expanding uniformly.

OPEN — NOT YET SETTLED

The manufacturing buildout is durable regardless of which molecule wins. Sterile capacity and injector supply serve the broader biologics pipeline.

OPEN — NOT YET SETTLED

Food impact is real but gradual, appearing first in package size and category mix rather than aggregate volume.

OPEN — NOT YET SETTLED

Telehealth has stabilized as a durable slice of care concentrated in behavioral health, not a substitute for in-person medicine.

OPEN — NOT YET SETTLED

Ambient documentation AI is the first mainstream clinical use of generative AI, and it spreads because it attacks clinician time without touching liability-heavy decisions.

OPEN — NOT YET SETTLED

Returns stay muted until payment moves. A technology with no billing code and no risk-bearing buyer fails regardless of how well it works — which is why the winners will look less like standalone apps and more like features embedded in electronic records and insurers.

OPEN — NOT YET SETTLED

The supply shortage remains the binding constraint of the decade; even a full mortgage-rate normalization would restore transactions, not abundance.

OPEN — NOT YET SETTLED

State preemption of local zoning is now a durable, bipartisan trend — the era of purely local control over housing supply is ending.

OPEN — NOT YET SETTLED

Measurable results arrive late: 2023–2026 reforms won't visibly move starts before the late 2020s, and the gap between passage and production will be misread as failure.

OPEN — NOT YET SETTLED

Construction labor is the sleeper constraint — without workforce growth (including immigration), zoning reform legalizes homes no one is available to build.

OPEN — NOT YET SETTLED

The Sun Belt's population and construction lead over coastal metros persists through the decade; the drivers — permitting elasticity, employer relocation, retiree flows — are structural, not pandemic artifacts.

OPEN — NOT YET SETTLED

The *pace* of migration has permanently downshifted from 2021–2022 levels; expect steady inflows, concentrated in metros that stayed affordable (San Antonio, Oklahoma City, the Carolinas) over the famous boomtowns.

OPEN — NOT YET SETTLED

Sun Belt housing markets will keep underperforming on price and outperforming on volume — softening prices there are evidence the model works, not that it's failing.

OPEN — NOT YET SETTLED

Insurance, not water or heat, is the nearest-term threat to the model, because it converts diffuse climate risk into an immediate monthly payment.

OPEN — NOT YET SETTLED

Lock-in persists as a first-order force into at least the late 2020s; with roughly half of mortgages under 4%, no plausible near-term rate path closes the gap quickly.

OPEN — NOT YET SETTLED

The effect weakens gradually and predictably as low-rate loans age out — expect resale volumes to grind higher each year even without a rate rally.

OPEN — NOT YET SETTLED

Rates sustainably in the mid-5% range are the approximate unlock threshold: for millions of households the monthly gap compresses into "worth it for the right move" territory.

OPEN — NOT YET SETTLED

The most underpriced consequence is labor-market friction — reduced mobility acting as a quiet drag on wage growth and productivity, invisible in any housing statistic.

OPEN — NOT YET SETTLED

As long as most existing owners hold mortgages far below market rates, builders will keep capturing an outsized share of transactions — the advantage is structural, not cyclical.

OPEN — NOT YET SETTLED

The new-versus-existing price inversion persists while builders shift toward smaller homes and quiet incentives; it reflects mix and financing, not cheap production.

OPEN — NOT YET SETTLED

Buydown economics squeeze hardest at small private builders without mortgage arms or scale — pushing consolidation toward the large publics.

OPEN — NOT YET SETTLED

If mortgage rates fall decisively, the moat shrinks fast: resale supply returns as competition just as incentive spending loses its punch.

OPEN — NOT YET SETTLED

The defining feature of this cycle is collapsed volume, not falling prices — and it should persist as long as the gap between outstanding and prevailing mortgage rates stays wide.

OPEN — NOT YET SETTLED

A traditional price crash remains unlikely without a labor-market shock — forced selling, the fuel of 2008, is largely absent. Erosion in real prices is the likelier valve.

OPEN — NOT YET SETTLED

The freeze thaws gradually, not suddenly: through slowly rising inventory, life-event moves, and builder supply, rather than a single rate-cut moment.

OPEN — NOT YET SETTLED

Regions are diverging. Supply-friendly Sun Belt metros are thawing first — more listings, flat-to-falling prices, more sales — while tight coastal markets stay frozen longest.

OPEN — NOT YET SETTLED

Sub-tier suppliers — castings, forgings, energetics, motors — will outgrow the prime contractors' own revenues in percentage terms through the decade, because they're growing off a starved base.

OPEN — NOT YET SETTLED

Single-source chokepoints will cause at least a few high-profile program delays in the next several years; the map of those failures will direct the next round of investment.

OPEN — NOT YET SETTLED

Allied co-production is durable across political cycles — it's one of the few defense policies that saves money, adds capacity, and strengthens alliances simultaneously.

OPEN — NOT YET SETTLED

Software will take a steadily rising share of defense spending, and its margins will exceed hardware margins for the same strategic contribution.

OPEN — NOT YET SETTLED

Program-of-record status for commercial software platforms is the regime change: it converts episodic contracts into durable, compounding franchises.

OPEN — NOT YET SETTLED

Vendor concentration is the next fight — expect open-architecture mandates and funded challengers within two to three years as Congress reacts to Palantir-Anduril dominance.

OPEN — NOT YET SETTLED

Electronic warfare adaptation speed will become a formally tracked military metric within a few years.

OPEN — NOT YET SETTLED

Defense manufacturing capacity will keep expanding through the decade regardless of which party holds power — the funding coalition is durable even when execution embarrasses it.

OPEN — NOT YET SETTLED

Delivery rates will keep lagging announced goals by two to four years across munitions and shipbuilding; the gap between press release and qualified output is structural, not managerial.

OPEN — NOT YET SETTLED

Skilled labor, not capital, is the binding constraint — expect defense wages for welders and machinists to keep outrunning manufacturing averages, and automation investment to follow.

OPEN — NOT YET SETTLED

Cheap precision mass is now a permanent category of military power, alongside — not instead of — exquisite platforms.

OPEN — NOT YET SETTLED

Counter-drone defense will grow faster than drone offense in Western budgets over the next five years, because the West is currently on the losing side of the defender's math.

OPEN — NOT YET SETTLED

The winning counter-drone systems will be judged on cost per intercept, not probability of kill — guns, microwaves, and cheap interceptors over exquisite missiles.

OPEN — NOT YET SETTLED

Militaries that shorten their adaptation cycle will gain more advantage than those that buy any particular system.

OPEN — NOT YET SETTLED

Defense spending on autonomy will keep shifting from platform-first to software-first economics, with the data and autonomy layers earning the most durable margins.

OPEN — NOT YET SETTLED

The primes-versus-startups fight ends in a barbell, not a takeover: primes keep exquisite platforms and munitions; new entrants take attritable mass and software.

OPEN — NOT YET SETTLED

Replicator matters more as precedent than as procurement — its fast-buying pathways will outlive the initiative itself.

OPEN — NOT YET SETTLED

Within three years, the binding constraint on the stack will be component supply chains and testing capacity, not funding or ideas.

OPEN — NOT YET SETTLED

Aggregate hyperscaler capex rises again in 2027, but the growth rate peaks by then — the shift from cash funding to debt and SPV financing is the tell that internal limits have been reached.

OPEN — NOT YET SETTLED

The buildout is now macroeconomically load-bearing: a genuine spending halt would subtract measurably from U.S. GDP growth and hit credit markets, not just tech stocks.

OPEN — NOT YET SETTLED

The infrastructure will outlive the boom that built it, echoing rail and fiber — but returns will distribute unevenly, favoring the scarce layers (power, land, chokepoint components) over the abundant ones.

OPEN — NOT YET SETTLED

The cycle's end, when it comes, will be announced not by a crash but by one sentence of softened guidance from a single hyperscaler.

OPEN — NOT YET SETTLED

Enterprise AI has durably crossed from pilots to production; the trend will not reverse even if the stock market's AI enthusiasm does.

OPEN — NOT YET SETTLED

Through 2027, ROI stays concentrated in text-heavy, high-volume workflows — support, sales development, coding, document operations — rather than spreading evenly across the enterprise.

OPEN — NOT YET SETTLED

A visible wave of failed and cancelled agent projects will coexist with compounding successes, and headlines will overweight the failures.

OPEN — NOT YET SETTLED

Labor effects arrive as attrition-absorbed restructuring — slower hiring in specific back-office roles — before they ever appear as layoffs, making them nearly invisible in monthly jobs data until 2027 or later.

OPEN — NOT YET SETTLED

OpenAI's restructuring solved its capital-access problem and created a concentration problem: it's now the single largest private counterparty in the AI buildout, and its health is systemically important to the whole chain.

OPEN — NOT YET SETTLED

The gap between committed compute spending and current revenue cannot be closed by consumer subscriptions alone; enterprise and agentic revenue must scale several-fold by 2028, or commitments will be renegotiated.

OPEN — NOT YET SETTLED

The investor-vendor circularity (Nvidia, Amazon, Oracle, SoftBank on both sides of the table) amplifies both the boom and any eventual correction.

OPEN — NOT YET SETTLED

A public listing, if it comes, will function as the market's referendum on the entire AI capex cycle — not just on one company.

OPEN — NOT YET SETTLED

Over the next five years, a disproportionate share of durable AI profits accrues below the model layer — in power, equipment, land, and credit — while the model layer stays fiercely contested.

OPEN — NOT YET SETTLED

Regulated utilities in data-center corridors are the least-appreciated structural beneficiaries, because rate-base growth compounds regardless of which AI company wins.

OPEN — NOT YET SETTLED

The middle of the stack (servers, networking, construction) grows enormously but concedes margin to concentrated buyers.

OPEN — NOT YET SETTLED

Private credit's AI exposure becomes a meaningful financial-stability question by 2028 — the same position, lender to a boom, that has defined every infrastructure cycle's eventual stress point.

OPEN — NOT YET SETTLED

Through at least 2027, advanced packaging and HBM — not GPU production — set the ceiling on how much AI compute the world can deploy each year.

OPEN — NOT YET SETTLED

Pricing power in the AI stack is migrating from chip designers toward the concentrated chokepoints: memory makers, packaging capacity, and electrical-equipment manufacturers.

OPEN — NOT YET SETTLED

Memory scarcity will keep leaking into non-AI markets (servers, PCs, phones), making DRAM prices an underappreciated inflation channel through 2026–27.

OPEN — NOT YET SETTLED

Liquid cooling shifts from exotic to default for new AI capacity, pulling pumps, plumbing, and thermal engineering permanently into the data-center economy.

OPEN — NOT YET SETTLED

U.S. electricity demand growth is real and durable, not a forecasting blip — but the pace utilities are planning for will prove too high in some regions, because queue requests are partly duplicated and speculative.

OPEN — NOT YET SETTLED

Firm generation is the binding constraint through at least 2030; turbine capacity and nuclear timelines can't expand as fast as announced demand.

OPEN — NOT YET SETTLED

Nuclear restarts and hyperscaler power contracts stay rare and expensive — strategically important, but not enough megawatts to close the gap. Gas fills most of it.

OPEN — NOT YET SETTLED

Consumer electricity prices in data-center-heavy regions rise faster than the national average, making cost allocation a first-tier political fight by 2027.

OPEN — NOT YET SETTLED

Physical investment will outgrow financial engineering for the rest of this decade. Buybacks won't vanish, but their era as the market's marginal buyer is over.

OPEN — NOT YET SETTLED

The savings-investment seesaw keeps the floor under interest rates higher than the 2010s taught everyone to expect. Sub-2% 10-year yields belong to the old regime absent a crisis.

OPEN — NOT YET SETTLED

Market leadership broadens toward the suppliers of the buildout: electrical equipment, grid infrastructure, engineering, select industrials and materials, alongside the tech giants funding it.

OPEN — NOT YET SETTLED

Every building era overbuilds something. By late decade, expect visible excess in at least one lane, AI compute being the obvious candidate, without that invalidating the broader regime, just as railroad bankruptcies didn't unbuild the railroads.

OPEN — NOT YET SETTLED

No currency displaces the dollar's core roles within a decade. The plumbing of trade invoicing, FX markets, and offshore debt changes at generational speed.

OPEN — NOT YET SETTLED

Reserve diversification into gold and mid-sized currencies continues regardless. Expect the dollar's reserve share to keep grinding lower by fractions of a point per year, not collapsing.

OPEN — NOT YET SETTLED

The economically meaningful risk is a rising cost of U.S. borrowing as foreign official demand for Treasuries flattens while issuance grows. This connects directly to the fiscal story: the "exorbitant privilege" shrinks at the margin exactly when America leans on it hardest.

OPEN — NOT YET SETTLED

Weaponized finance is the main accelerant. Each expansion of sanctions makes dollar workarounds more valuable; the pace of de-dollarization is substantially a U.S. policy choice.

OPEN — NOT YET SETTLED

Factory construction has peaked for this cycle but will plateau far above the pre-2021 norm, roughly double the old baseline, rather than round-trip. The regime changed; the spike didn't.

OPEN — NOT YET SETTLED

The binding constraints on reshoring through 2030 are electricity availability and skilled trades, not capital. Projects will slip years for want of megawatts and electricians.

OPEN — NOT YET SETTLED

The economic gains will stay geographically concentrated. A dozen metros capture most of the value; "industrial renaissance" will feel true in Phoenix and abstract in most of the country.

OPEN — NOT YET SETTLED

Measured productivity from these plants arrives late. Expect a stretch where costs are visible and output isn't, and expect that gap to be used politically against the whole project.

OPEN — NOT YET SETTLED

The era of "invisible" tightening is over. With the RRP empty, any future balance-sheet drain hits bank reserves directly, so markets are more liquidity-sensitive now than at any point since 2019.

OPEN — NOT YET SETTLED

The Fed's December bill-buying marks the practical floor of the balance sheet. The trend from here is gradual growth, a structural tailwind for asset prices, whatever it's officially called.

OPEN — NOT YET SETTLED

Treasury's bill-heavy mix will continue; it's the path of least market disruption. A forced shift back toward long bonds is the underpriced risk that would drain the harbor fastest.

OPEN — NOT YET SETTLED

Expect more sudden, newsless volatility around tax dates, quarter-ends, and refunding weeks, the moments the plumbing is most exposed.

OPEN — NOT YET SETTLED

The U.S. labor market is in a stall, not a collapse: job growth near 36,000 a month is below what even a slow-growing workforce likely needs over time.

OPEN — NOT YET SETTLED

The 4.2% unemployment rate overstates health; participation-driven declines are masking weak hiring. Judge this market by hiring and quits rates, not the headline rate.

OPEN — NOT YET SETTLED

Payroll revisions will likely keep skewing negative while the stall lasts. Treat each first print as a ceiling, not a floor.

OPEN — NOT YET SETTLED

Because energy-driven inflation delays Fed support, the risk of a policy mistake, easing too late into a weakening job market, is the highest it's been this cycle.

OPEN — NOT YET SETTLED

The 10-year yield will stay structurally higher this cycle than the 2010s norm. A durable return below 3% would likely require a recession or renewed Fed bond-buying, not just rate cuts.

OPEN — NOT YET SETTLED

Mortgage rates cannot meaningfully fall unless the 10-year falls. Fed cuts that don't move long yields won't unfreeze housing.

OPEN — NOT YET SETTLED

Equity valuations are more exposed to long yields than to Fed meetings. A sustained move above roughly 5% would test the market's AI-earnings math in a way nothing since 2022 has.

OPEN — NOT YET SETTLED

The swing variable is term premium and Treasury supply, not the policy rate. Watch the fiscal side for the next big yield move in either direction.

OPEN — NOT YET SETTLED

Fiscal policy is now a co-equal market force with monetary policy. Treasury's quarterly refunding decisions will keep moving markets the way Fed statements do, whoever chairs the Fed.

OPEN — NOT YET SETTLED

Deficits stay structural for the rest of the decade. Under current law, the debt ratio rises even in the good scenarios, and no plausible election outcome changes the programs driving it.

OPEN — NOT YET SETTLED

The path of least resistance is quiet: nominal growth and inflation running somewhat above interest rates, plus rule changes that deepen captive demand for Treasuries. Not default, not austerity.

OPEN — NOT YET SETTLED

This judgment is falsifiable. Sustained primary surpluses, or interest costs falling well back below defense spending for years, would mean the regime call was wrong.

OPEN — NOT YET SETTLED

Munitions and missile production will keep expanding for years regardless of election outcomes — the funding coalition behind it is durable and bipartisan.

OPEN — NOT YET SETTLED

Qualified output will lag appropriations by two to five years, because the constraints are machine tools, energetic materials, and skilled workers, not money.

OPEN — NOT YET SETTLED

Sub-tier suppliers — energetics, rocket motors, castings — will remain the binding chokepoints, and the highest-leverage place to watch for both risk and growth.

OPEN — NOT YET SETTLED

Allied demand will exceed this industrial base's ability to serve it through the decade, keeping order books full even if U.S. procurement plateaus.

OPEN — NOT YET SETTLED

Secured power, not chips or capital, remains the decisive competitive variable in AI infrastructure through at least 2028.

OPEN — NOT YET SETTLED

The price premium for entitled, powered land over raw land keeps widening as long as interconnection timelines stay measured in years.

OPEN — NOT YET SETTLED

The financing layer — special vehicles, leases, private credit — is where any stress in the buildout will surface first, before it shows in tech earnings.

OPEN — NOT YET SETTLED

Transformer and turbine lead times ease only gradually; the industrial supply chain cannot be surged the way software can.

OPEN — NOT YET SETTLED

Institutional ownership of single-family homes stays in the low single digits nationally, but its concentration in Sun Belt metros and entry-level price bands will keep giving it outsized influence on the prices first-time buyers face.

OPEN — NOT YET SETTLED

Build-to-rent, not resale buying, is the growth engine from here — the institutional footprint in *new* construction will expand faster than its share of existing homes.

OPEN — NOT YET SETTLED

Legislation targeting institutional buyers will keep advancing because it polls well, and will keep underdelivering on affordability because it doesn't add supply.

OPEN — NOT YET SETTLED

If entry-level construction meaningfully recovers, the institutional opportunity shrinks with it — the bid is a creature of scarcity, and scarcity is the variable to watch.

OPEN — NOT YET SETTLED