Strategic Conditions: Q2 2026
Executive Brief
What matters this quarter.
In April, one of the largest technology companies in the world raised twenty-five billion dollars in the bond market to help fund a capital program it once paid for out of cash flow. The rating agencies noted the change. The equity market largely did not. That single transaction is the quarter in miniature: the artificial intelligence buildout has stopped being a story about technology and become a story about balance sheets, credit, power, and the physical economy.
Planned AI infrastructure spending across the major platforms now exceeds seven hundred billion dollars for 2026, up from roughly six hundred billion entering the year. The technology is real and the demand signals are genuine. The risk is not fraud or fantasy. The risk is arithmetic: too many companies are being priced as winners against the same future profit pool, and an increasing share of the buildout is being financed with debt rather than operating cash. Credit weakens before equity admits it. The debt is always the earlier signal.
Meanwhile the physical constraints we flagged in the first quarter tightened rather than eased. Power availability, grid interconnection, transformers, switchgear, and water are now the gating assets of the digital economy. Refining margins ran well above sustainable mid-cycle levels this quarter, and several downstream equities are priced as if peak conditions were permanent. Chemicals remained in structural reset, with no credible evidence yet that capacity is actually leaving the system. Across all of it, the same discipline applies: distinguish what the cycle gave from what the structure will keep.
Key Findings
- The AI capital cycle has become a credit event in formation. With 2026 buildout plans above seven hundred billion dollars and a rising share financed with debt, the earliest honest signal of trouble will appear in credit spreads and covenant behavior, not in equity narratives. Watch the debt.
- Power is the binding constraint, and the value is migrating to the bottleneck. Generation capacity, interconnection positions, transformers, switchgear, gas infrastructure, and water rights now price the growth of the digital economy. Owners of those assets hold the leverage; renters of them hold the risk.
- Downstream energy equities are pricing peak margins as permanent. Refining cracks ran materially above the level the cost curve and inventories support through a full cycle. The gap between the tape and mid-cycle is the expectation burden an owner inherits at today's prices.
Conditions Dashboard
The read across the firm’s core domains.
Energy & Infrastructure
Data-center load growth is colliding with interconnection queues and equipment lead times. Gas-fired generation and grid hardware are the scarce assets. Refining margins are above mid-cycle; the equities assume they stay there.
Chemicals & Materials
The structural reset continues. Overcapacity is durable, regional cost curves keep diverging, and announced rationalization still exceeds actual capacity exit. Time-to-balance keeps extending.
Industrials & Supply Chains
Layoff programs framed as AI efficiency have not yet shown the productivity evidence that separates extraction from capability destruction. Policy volatility is pricing a hidden tax into supply-chain commitments.
Capital Allocation
Capital is abundant for anything with an AI narrative and selective everywhere else. That asymmetry is the opportunity: discipline is being repriced upward in the sectors capital has abandoned.
Leadership & Execution
The gap between announced transformation and installed operating cadence remains the most reliable predictor of which strategies survive the next two quarters.
§ 03
Energy and Infrastructure
Start with the queue. In most U.S. power markets, the wait for grid interconnection is now measured in years, transformer lead times are measured in quarters, and data-center developers are signing power agreements before they sign land. That sequencing tells you where the leverage sits. The scarce input is no longer capital or compute. It is the ability to deliver electrons to a specific site on a specific date.
Our first-quarter view held that midstream was in an infrastructure cycle rather than a commodity cycle, and that the durable positions were corridors and contracts rather than throughput. The second quarter reinforced it. Gas-fired generation has been pulled back into the growth story to serve load that renewables and storage cannot yet firm, and the pipes, storage, and compression that feed that generation are being re-rated accordingly. The transmission chain is direct: AI load growth converts to gas demand, gas demand converts to infrastructure utilization, and utilization converts to contracted cash flow for the assets already in the ground.
Refining requires the opposite discipline this quarter. Cracks ran well above the level that cost curves and inventories can support through a full cycle, and several large downstream equities now embed those margins as if they were structural. Our own reverse valuation work this quarter found one major refiner-marketer priced for benchmark margins roughly fifty percent above sustainable mid-cycle. Nothing about that company's operations is broken. The price simply assumes the cycle is a plateau. Owners at these levels are not being paid to take risk; they are paying for the privilege of assuming it.
The discipline is the same one we applied in the first quarter, applied now in the opposite direction. Normalize through the cycle. Judge the asset on mid-cycle economics and the balance sheet on severe-case economics. If the investment only works at today's margins, it is not an investment thesis. It is a weather forecast.
§ 04
Chemicals and Materials
In March we wrote that chemicals was not in a routine demand downturn but a multi-factor structural reset: durable overcapacity, widening regional cost-curve divergence, and Chinese scale displacement extending the time-to-balance. One quarter later, the honest update is that nothing material has improved and the reset thesis has strengthened.
The pattern to watch is the gap between announced rationalization and actual capacity exit. Producers keep announcing reviews, impairments, and strategic alternatives. Far fewer have permanently closed units. Until announced exits become physical exits, the supply side has not adjusted, and every demand uptick will be absorbed by idle capacity rather than pricing. This is why the first apparent inflection in a structurally oversupplied sector is usually a trap for capital that confuses cyclical relief with structural repair.
For operators the implication is uncomfortable but clear: cost-curve position is the only defensible ground. Assets that are advantaged on feedstock, energy, and logistics will earn through the reset. Assets that depended on a demand recovery to justify their existence will keep consuming cash while their owners wait. For investors, the sector is becoming interesting for exactly that reason. The exit of capital is the precondition for the return of returns. We are watching for evidence of the former before underwriting the latter.
§ 05
Industrials and Supply Chains
The quarter's dominant industrial narrative was efficiency. Layoff announcements continued at a pace not seen since 2020, and an increasing share of them were framed as AI-driven productivity gains. The framing deserves scrutiny, because a workforce reduction is only an efficiency if two things rise together: output per remaining employee and the value of the enterprise. When neither moves, the cut was not efficiency. It was extraction dressed as strategy, and the bill for it arrives one to two quarters later in reliability, customer experience, and the quiet departure of the people who actually knew how the system worked.
Our advice to leadership teams evaluating these programs is to run the capability test before the cost test. Identify which functions carry the operating knowledge that does not survive a reorganization chart, and protect them explicitly. The companies that will compound through this cycle are the ones treating operational capability as an asset on the balance sheet rather than a line item in the plan.
On supply chains, policy volatility remained a hidden tax. Tariff uncertainty does not need to become tariff reality to impose cost; the uncertainty itself prices a premium into every sourcing decision, every inventory position, and every capacity commitment with a payback period longer than the political news cycle. The practical response is not prediction. It is designing supply positions that remain defensible across the plausible policy range, and paying for that flexibility knowingly rather than discovering its absence suddenly.
§ 06
Capital Allocation and Corporate Strategy
The central capital-allocation fact of the quarter is concentration of belief. Capital is abundant, cheap, and fast for anything attached to the AI buildout, and scarce, expensive, and skeptical for nearly everything else. Both halves of that sentence are the opportunity, and both carry the risk.
On the abundant side, the discipline is to fund the future without booking it before it is earned. The technology is real. The demand is real. What is not yet real is the profit pool large enough to justify every participant's current price simultaneously. History is specific about how this resolves: the dot-com era mispriced timing, the fiber overbuild mispriced capacity, shale mispriced capital discipline. In each case the technology succeeded and much of the capital did not. The companies that survived those cycles shared one trait. They could fund their build from a position of balance-sheet strength when the financing window closed for everyone else.
On the scarce side, sectors that capital has abandoned are where the arithmetic quietly improves. Payout behavior is one of the more honest tells. A dividend raise funded by organic free cash flow, made from a strong balance sheet, and consistent with a durable allocation pattern signals safer profits ahead, not necessarily higher ones. A raise funded by debt or asset sales is theater, and theater is expensive to the audience. We treat the raise as evidence, never proof, and validate it against funding quality before it changes any conclusion.
For boards and owners, the operative question this quarter is the expectation burden: what does our current valuation, or our current plan, require the world to deliver? If the answer is perfection sustained for years, the correct move is usually to harvest something while conditions are generous and rebuild the reserve that lets you act when they are not. Liquidity is not idle capital. It is the option to be decisive in the quarter when everyone else is forced to sell.
§ 07
Leadership and Execution
Volatile quarters reward a specific leadership behavior that looks passive from the outside: reducing decision frequency while increasing decision quality. When conditions whip between narratives, the instinct is to steer with every headline. The better discipline is the one taught by black ice. Stabilize first. Make small corrections. Do not jerk the wheel. A leadership team that changed strategic direction three times this quarter did not demonstrate agility; it demonstrated that it never had a thesis.
The second execution discipline worth naming is the commitment ledger. Every management team made promises last quarter: capacity actions, cost programs, capital returns, guidance. The most informative research a board or an investor can run right now is simply to line up what was promised against what was delivered, and to weight future promises accordingly. Institutions reveal themselves through delivery, not through communication.
Finally, the no-action decision deserves rehabilitation. In a quarter where capital was loud, the leaders who declined to chase, declined to over-hire into a narrative, and declined to reprice their own discipline were making an active choice that will not show up in any announcement. Activity is not a key performance indicator. Alignment between capital, capability, and conditions is.
§ 08
Key Strategic Moves
The moves below are the quarter's translation from reading to action. Each is stated for a leadership team or capital partner deciding in the next ninety days.
- Underwrite power before capital. Any strategy adjacent to the data-center buildout should verify generation, interconnection, and equipment access before committing capex. The constraint is physical, and contracts for it are being written now.
- Re-test energy exposure against mid-cycle. Where downstream margins or the equities that own them assume current conditions persist, harvest or hedge the difference between the tape and the sustainable level.
- Treat large AI vendor and infrastructure commitments as credit exposure. Review counterparty balance sheets and financing structures the way a lender would, because an increasing share of the buildout is borrowed.
- Sequence chemicals re-entry on physical evidence. Actual capacity exit, not announced reviews, is the trigger. Until then, cost-curve position is the only thesis.
- Rebuild the opportunity reserve while conditions are generous. Selling into strength is uncomfortable and correct; the reserve is what converts the next dislocation from threat into opportunity.
§ 09
What We Are Watching
The signals below are the ones most likely to change our view between now and the third-quarter report.
- AI-related debt issuance and private-credit exposure to compute borrowers. Spread behavior and covenant amendments will tell the truth before equity prices do.
- Interconnection queue reform and transformer and switchgear lead times. Any material easing changes the bottleneck thesis; continued tightening deepens it.
- Chinese chemicals run rates and physical capacity closures, as distinct from announced rationalization.
- Dividend and buyback funding quality across energy and industrials. Payout actions financed by debt or asset sales are early balance-sheet tells.
- The pace of refining crack normalization against what downstream equities have priced. The gap can close from either direction; only one of them is kind to today's buyers.
§ 10
Questions Leaders Should Ask
These are the questions we would put on the agenda of any board, owner, or investment committee this quarter.
- What does our current plan or valuation assume will persist, and what is the actual evidence that those conditions are structural rather than cyclical?
- If our AI-related spending had to justify itself as owner earnings within three years, what would we cut first, and why have we not examined that already?
- Where does our growth depend on power, interconnection, equipment, or people we have not yet secured?
- Which of our efficiency programs can demonstrate rising output per employee and rising enterprise value, rather than falling headcount alone?
- If forced selling appeared in our sector next quarter, are we positioned to act, or only to react?
Related Engagement
Discuss how these conditions affect your m&a & capital decisions.
Source & Methodology
This report draws on public company filings and SEC data, Federal Reserve and FSOC financial stability publications, BIS reporting on private credit, and January 2026 sector datasets from Aswath Damodaran, read through the firm's operator-investor lens. Valuation references are directional and method-driven rather than security recommendations. No client-confidential material is used. Views are as of late June 2026 and will be updated in the Q3 report.
Next Quarter