House View

House View v1.0 — body

After enough years on a trading desk, you learn that almost no opinion outlasts the close. The ones that do belong to people who have something underneath what they're saying — a structure they come back to whether the day goes their way or not. The structure is what I still trust. The day is the day. The structure tells you whether the day means what you think it means.

This is that structure, in writing, dated and signed. Call it the House View. We will revise it when the evidence changes. We will keep the record when we get something wrong. The point of putting it down is the same point an analyst's structure has always served: to make sure that when the next session opens and the tape moves, we are reading the move against something, not reacting to it as if it were the whole picture.

The Spine

The standing claim has one sentence and a long lineage.

High-debt regimes structurally rhyme because inflation is the path of least political resistance for debt resolution.

That is mechanism, not chart-pattern fitting. Governments accumulate debt across peacetime and especially across crises. Debt service eventually consumes enough fiscal capacity to constrain political choice. When political choice is constrained, the resolution mechanism that produces the least immediate pain is inflation — softer than default, softer than austerity, faster than trying to grow out of it. Reinhart and Rogoff cataloged it in This Time Is Different. Ray Dalio's long-term debt cycle framework is the same argument from a different desk. Russell Napier has been making it about reserve currencies. Bridgewater has been running the math.

The reason this matters as a Spine and not just as a thesis is that it survives. A regime that rhymes with the past does not rhyme on every metric, but it rhymes on the political-economy mechanism. The chart pattern is downstream of the mechanism. The mechanism is what we anchor on.

Steel-manning the other side honestly: mainstream macro has serious counter-arguments. Fiscal sustainability is more accurately a function of the spread between growth and rates than of the debt-to-GDP level. Inflation is fundamentally monetary, not a fiscal-resolution mechanism. The United States in particular has a fiscal capacity that historical analogies systematically underestimate because reserve-currency status changes the equation. These are not silly objections. They are held by serious people, and the publication will treat them as load-bearing alternatives, not strawmen.

One nuance the Spine carries into this particular regime. The global economy sits at the highest leverage in its history. That is the alarming half of the picture, and the half that the standard reading emphasizes. The other half is that the leverage is arriving alongside a productivity event of genuinely uncertain magnitude — uncertain in timing, uncertain in distribution, but plausibly the largest such event in the lifetime of anyone reading this. The combination is what makes this regime resolve differently than past rhymes might suggest. Inflation continues to serve as the political resolution mechanism. A productivity lift, if it arrives at the magnitude its substrate allows, may serve simultaneously as fuel for real growth. The two together, in this regime, are less in tension than they have typically been in the past. The Spine does not change. The texture of this rhyme might.

The Specification

The Spine is durable. The Specification is what the current regime's specific shape looks like, and it is where the work lives.

The Specification has two legs. The first is well-established. The second is held with looser conviction and will keep developing.

Leg 1 — AI compute as the meta-resource

Oil reorganized the twentieth century. Compute is reorganizing the twenty-first.

That is not a slogan. It is the variable that explains the specific texture of fiscal, geopolitical, and capital-market behavior we are watching. Fiscal stress, because the old monetary system cannot smoothly absorb the capex that frontier compute requires. Geopolitical realignment, because compute access — the chips, the lithography that makes them, the power that runs them — is being weaponized in the same way oil access was weaponized in the second half of the twentieth century. Capital reorganization, because the productive base is shifting from energy-intensive to compute-intensive, and capital follows productive substrate.

The current regime is not, in our reading, primarily a story about white-collar work disappearing. That is the first-order effect, and it is real, and it is visible. The second-order effect is the more strategic one: compute, applied at scale, relaxes constraints that had been treated as fixed. Materials science. Drug discovery. Power systems engineering. Industrial process design. Agricultural productivity. The constraint relaxation is the strategic story. The substitution is the photogenic story.

One trajectory worth naming. Compute today runs on power inputs that the existing grid was not built for. Oil, natural gas, coal — the current substrate — is doing the immediate work and will continue to for some time. The trajectory, however, is toward power sources whose ceiling is materially higher than what terrestrial generation can deliver at the rate compute demand is scaling. The orbital-solar conversation that has moved from speculative to active engineering inside the last twelve months is one signal of where the power input is going. The reorganization of compute's power substrate is itself a Layer 2 variable the publication tracks.

Leg 2 — The waterfall

Compute is the meta-resource. The waterfall is everything that splashes out from the compute build and reshapes activity downstream. This is the leg held more loosely; it is also the leg that, if right, has the largest implications.

Three currents in the waterfall worth naming.

The productivity lift. Reasonable analysts estimate the productivity multiplier from frontier AI deployment across the service economy in the range of sixfold to thirty-sixfold, depending on assumptions about adoption pace and the depth of constraint relaxation downstream. That is a wide range, because the underlying uncertainty is wide. What is not wide is the floor: even at the low end, no historical productivity event matches its scale. The combination of high leverage and a productivity lift of that magnitude is what makes the current regime's resolution path genuinely different from the past rhymes it would otherwise share political-economy DNA with.

Programmable settlement. Blockchain rails — stablecoin issuance, agent-payment infrastructure, tokenization of traditionally illiquid assets — are becoming the coordination layer underneath the compute layer. Systems and devices participating directly in markets, with their own settlement and their own incentives, becomes infrastructurally possible at the rate the rails mature. The full shape of this is not yet visible. What is visible is that the rails are being built. When the rails mature, the unit of economic activity changes. Coordination at the machine layer becomes a normal feature of the economy rather than a research project.

Hyperdiversification, and what it implies. Tokenization plus AI-mediated portfolio construction points toward a structural shift in retail access. A portfolio of ten thousand positions, each AI-rated for risk and continuously rebalanced against a precisely-dialed risk preference, is technically feasible on rails that already exist in beta form. The return curve at that level of diversification is materially smoother than what most retail portfolios deliver today. The cost of building it is collapsing. The friction in maintaining it is collapsing. Paired with AI-generated legal templates that compress estate planning, contract formation, and identity verification into something individuals can operate themselves — predetermined, cryptographically certified, no probate court required — what emerges is something like the individual as a sovereign economic entity, with the institutional machinery that has historically required law firms and banks compressed into rails the individual operates directly.

I am aware that the last paragraph reads as science fiction. The publication is naming it because the rails for each component already exist in some form, and the trajectory is visible in capital flows even though the endpoint is not yet here. We hold this current in Layer 3 — Knightian — and we will track it through what shows up in capital markets and what shows up in regulatory posture, rather than pretending we know its timing.

How we hold what we know

Three confidence layers. We keep them distinct because mixing them is the failure mode.

Layer 1 — near-term, high confidence. Macro variables on a quarterly clock. Inflation prints. Policy rates. The yield curve. Recession triggers. These have numeric thresholds, dated releases, and falsifiable readings. They are commitment-grade.

Layer 2 — multi-year, high confidence. Geopolitical structural anchors. Compute export policy. Sovereign capital commitments. Energy-policy alignment with compute siting. Defense industrial reshoring. The reorganization of compute's power substrate. These move on the order of years and rarely reverse. They are commitment-grade on a longer clock.

Layer 3 — Knightian. Where the AI build leads in three or five or ten years. What the third-order effects of the meta-resource transition end up looking like. Whose tools win at scale. The magnitude and timing of the productivity lift. The shape of programmable settlement at maturity. The endpoint of hyperdiversification and individual sovereignty in legal-administrative matters. The publication does not hold opinions in this layer with the same conviction it holds Layer 1 and Layer 2 opinions. We track the live variables. We refuse to pretend we know more than we know.

The three orders of effects from any transformative technology go in the same direction. The first-order effect is substitution — the new thing replaces the old thing at the same job. Visible, measurable, often loud. The second-order effect is constraint relaxation — the new thing makes previously-fixed limits move. Strategic, less visible, decisive in the long run. The third-order effect is the existence of entirely new categories that nobody could have described from inside the prior regime. The suburb is a third-order effect of oil. Modern computing is a third-order effect of electricity. We do not pretend to forecast the AI equivalents. Anyone who claims to is selling something.

Where we go beyond that horizon, I am incapable of imagining. That sentence is not a piece of false humility. It is the actual epistemic posture, and it will appear in this publication often.

What we are watching

The Frame is held in part through Watchpoints — specific, dated, falsifiable triggers that would update the Specification or, in some cases, the Spine itself.

The current Watchpoint set:

• Core PCE re-acceleration above 3.2% year-over-year, resolving on the June 2026 release. A Layer 1 item.
• The 10-year Treasury sustaining above 4.75%. Quarterly resolution. A Layer 1 fiscal-stress signal.
• The Sahm Rule triggering above 0.50 percentage points on the three-month moving average. Monthly resolution. Layer 1, recession-positioning.
• A new compute-export rule extending H200/B200-class restrictions to additional jurisdictions. Rolling. Layer 2.
• A new sovereign-AI capital commitment of ten billion dollars or more, within the next ninety days. Rolling. Layer 2.
• A confirmed datacenter project delayed by power-grid constraint for more than twelve months. Rolling. Layer 2, and a structural signal that the meta-resource has begun to bind on the power side.

Each of these will resolve on the date it resolves, and the resolution will be published. A Watchpoint breach is always-escalate — the publication will return to the Frame and reconsider what the breach means, in writing, in the next revision.

What this implies

A House View is not a list of names. It is a regime read, and the regime read implies tilts in the kinds of exposures that compound when the Frame is correct and that survive when it is partly wrong. The publication does not give individualized advice. What we can say honestly about the shape of the regime:

Substrate that the meta-resource depends on tends to compound through the cycle. Semiconductor design and manufacture. The lithography that makes both possible. The data-center infrastructure layer underneath the silicon. The power generation that the silicon runs on, and the rails carrying that power. These are not predictions about any single name. They are observations about the shape of an industrial buildout.

The defense industrial base sits in the path of geopolitical realignment, and the path of geopolitical realignment is not a one-quarter story. The names whose order books grow with that path tend to grow on a different clock than the index.

Incumbent businesses sitting on proprietary data — legal, medical, financial, information-services — extend their moats rather than lose them when the AI build matures, because the AI build extracts more value from data that the public web cannot reach. The substitution story misses this. The constraint-relaxation story does not.

And ownership of equities in general is one of the few inflation hedges that also compounds. The Spine implies that holding productive assets through the regime is a structurally defensive posture, not a speculative one. Paired with the productivity lift, the same posture takes on a constructive register that the standard inflation-hedge framing usually does not.

The Knightian leg of the Specification implies, more loosely, that the infrastructure carrying the tokenization current — payment rails, settlement networks, identity layers — sits in the path of whatever shape the waterfall ultimately takes. We are watching that current. We are not commit-grade on the endpoint.

What the Frame does not imply is conviction on the third-order outcomes. We will name candidates as they appear. We will not pretend to know what categories will exist in 2031 that do not exist today.

The standing position

Three currents push in the same direction on the demand for productive assets.

More money in the system — the political-economy mechanism of the Spine itself.

More real productivity — the lift available from the substrate, if it arrives at anything near the lower end of the range its proponents argue for.

More participants accessing the system more easily — the tokenization current, hyperdiversification, the compression of institutional friction into rails the individual can operate.

More money, paired with more productivity, paired with more access, produces more demand for the assets the system is built around. That is the regime read, simplified to its arithmetic. Each input carries its own uncertainty. The direction of the combined vector is the part we are willing to commit to.

The Frame's standing position is Advance on the Global Economy over the regime horizon.

That is not a call on any specific name. It is not a prediction about next quarter's print, or the print after that. It is the standing posture from which the more specific positions on the Scoreboard are evaluated. A Long Diagonal candidate that is coherent with Advance compounds the Frame. A candidate that fights it has to clear a higher bar. The standing position is the gravity in the room.

It will resolve when it resolves. It will be revised when the evidence asks for revision. The record will be kept either way.

What this publication is not

It is not personalized investment advice, and it does not become personalized when it is more specific. The publisher's exclusion under the Investment Advisers Act exists for a reason: there is value in independent analysis of public information, written for a general audience, with the publisher's positions disclosed. The publication operates inside that exclusion deliberately. Where the publication carries specific positions in the Scoreboard, the operator's exposure relevant to those positions is disclosed. There is no auto-renewal at any tier. There is no quiet deletion of any call.

The vocabulary in this publication is chess vocabulary. Positions advance and retreat. Pieces come on and come off the board. The Watchpoints carry numeric thresholds. The Scoreboard resolves every call as a Checkmate, Retreat, Stalemate, or Loss, and the resolution is published. The publication will not use directional-finance vocabulary because that vocabulary is what advisers say to clients, and this publication does not have clients. It has readers.

The hand-off

A standing Frame is not the same as a static one. The Spine moves slowly. The Specification moves with the evidence. The Watchpoints carry dates. Each of the relay desks reads this Frame on every shift — the Asia Desk in the overnight, the Morning Desk against the open, my own synthesis at the close. The Long Diagonal screener runs against the regime tilt. The Scoreboard resolves whether we were reading the regime correctly or not.

This is what I am willing to commit to in writing today. We will revise it when the evidence asks us to. We will keep the record either way.

What we do well, we want to be useful at. What we do not know, we want to be honest about. The line between those two is where this publication lives.

Changelog

A standing Frame is not a static one. Every substantive revision is dated and described here. Quiet edits — typos, link fixes, formatting — do not appear in this log.

Citations and underlying research available at /house-view/citations.

The Frame revises when the evidence asks. Watchpoint breaches trigger Frame review on the next regular slot or sooner if material. Specification updates are appended as new rows. Spine revisions are rare and warrant their own piece.

The operator holds indirect exposure to the names and themes discussed in this publication, primarily through diversified funds. Current detail at /disclosures. The publication does not provide individualized investment advice; see Editorial Philosophy.