The market had been waiting for three confirmations. It received all three inside five trading days. The AI infrastructure cycle cleared its most important hurdle when Microsoft, Alphabet and Meta raised capital expenditure guidance simultaneously, placing combined 2026 AI spending at roughly $725 billion annualised. Kevin Warsh’s path to Federal Reserve chairmanship went from contested to certain after the Justice Department dropped its probe and Senator Tillis withdrew his hold. And the energy market was simultaneously pushed higher by the US announcement that its naval blockade would remain until Iran agrees to a nuclear deal, and pulled lower by the UAE’s departure from OPEC+ effective 1 May.
Three questions answered. Three different directions. If you were positioned for a single narrative, this was a difficult week. If you were positioned for the spread between asset classes rather than the level of any one of them, it was a very instructive one.
S&P at record 7,230; AI capex cycle confirmed bullish (combined hyperscaler spend $725B annualised). Warsh regime change — Taylor Rule replacing forward guidance — is the multiple-compression risk ceiling. Size for uncertainty, not conviction.
Watch: META −7% after-hours on ROI concerns. Infrastructure (BE, TLN) and application layer (META) are now pricing differently. That bifurcation widening or compressing is the Week 17 signal.
PCE Core 3.2% YoY — highest since November 2023 — removes any probability of a June cut. Under Warsh’s Taylor Rule framework, the current data implies a terminal rate materially above the forward curve. No position in long-duration justified. Hold floating-rate.
Yield curve: 2Y 4.65% / 5Y 4.28% / 10Y 4.39% / 30Y 4.97% — bear steepening at the long end as regime change is priced in.
Moved from Green. Two structural forces now pulling in opposite directions: Hormuz supply constraint (US blockade until nuclear deal) vs UAE OPEC+ exit effective 1 May (potential +1.0–1.7m b/d + Saudi retaliation risk). Net direction genuinely unclear.
Watch: WTI $105. Saudi Arabia response to UAE exit is the decisive hinge. Copper: thesis intact despite Grasberg volume miss; -13.4% from Week 14 entry is a better entry, not a thesis breach.
Bitcoin $78,400 (−10.8% YTD) rebounded from $71,946 on hyperscaler capex confirmation — the named catalyst that would change the prior Red view arrived. Structural positive now present but Warsh regime tightening global liquidity creates a ceiling. Range-bound with modest upward bias.
Watch: ETH $2,270 (−23.5% YTD). Howell liquidity model remains the most reliable leading indicator for BTC direction; watch M2 global supply weekly.
| Scenario | Prob. | Trigger | 2Y | 10Y | Equity impact |
|---|---|---|---|---|---|
| Base: Taylor Rule, no 2026 cuts | 55% | NFP >150k + CPI ≥3% | 4.60–4.80% | 4.40–4.60% | −3 to −5% on multiple compression |
| Hawkish: H2 hike discussion | 25% | NFP >200k + PCE >3.2% | 4.80–5.00% | 4.60–4.80% | −7 to −10% |
| Relief: energy disinflation | 20% | WTI <$85 (UAE ramp) | 4.30–4.50% | 4.10–4.30% | +3 to +5% |
Series A, $105 million, led by Breakthrough Energy Ventures; $34 million seed led by Initialized Capital. Total raised: $139 million. Co-founded by Chad Rigetti, Idalia Friedson and Michael Keiser.
The quantum computing industry spent a decade making a single implicit assumption: that the path to commercial value ran through a general-purpose quantum computer. Build something large enough, stable enough, error-corrected enough — and the applications would follow. This assumption produced a generation of companies racing to hit qubit counts while enterprise customers waited, sceptically, for a machine that could actually do something useful on a Monday morning.
Chad Rigetti spent eight years building quantum hardware the conventional way at Rigetti Computing (NASDAQ: RGTI), the company he founded in 2013 and took public via SPAC in 2022. Then he resigned, took six months, and came back with a different question entirely. Not “how do we build a better quantum computer?” but “what problem does the enterprise actually need solved right now, and is there a quantum approach to solving it that does not require us to wait another decade?”
The answer he arrived at was the quantum abstraction layer. Sygaldry does not sell a quantum computer. It sells a quantum-accelerated server — a device designed to plug into existing data-centre infrastructure, run alongside classical processors, and handle the specific class of optimisation problems (logistics routing, portfolio construction, molecular simulation, cryptographic operations) where quantum approaches already outperform classical ones at realistic qubit counts. The machine is not general-purpose. It is not trying to be. That is the point.
In early 2025, Rigetti had a working prototype and a waiting list of enterprise pilots. He also had a problem. The pilot customers wanted to integrate the device into their existing orchestration layer — their Kubernetes clusters, their HPC schedulers, their cloud APIs. Sygaldry’s hardware was ready. The software interface was not.
The decision was whether to delay the Series A by twelve months, finish the software layer properly, and launch into a market that would be twelve months more crowded — or raise now, ship the hardware with a provisional interface, and rely on customer integration teams to build the bridges themselves. Rigetti chose the latter. It was the right call for the wrong reasons: two of the four pilot customers had engineering teams capable of building the integration. The other two did not, and the relationship nearly collapsed before Sygaldry rebuilt the interface in six months under pressure.
What the experience taught Rigetti, and what the Series A pitch was ultimately built on, is that enterprise quantum value is not a hardware problem or a software problem. It is an integration problem. The companies that will win are the ones that solve the interface between quantum and classical infrastructure — not the ones with the most qubits. Breakthrough Energy Ventures, led by Bill Gates’s climate-tech mandate, backed Sygaldry specifically because the abstraction layer approach opens a path to quantum-accelerated materials discovery, a sector where the gap between what classical computers can simulate and what quantum processors can simulate is already commercially meaningful.
Quantum computing at scale introduces two structural governance risks that the industry has not yet addressed seriously. First, quantum-accelerated cryptographic capability will break most current encryption standards years before post-quantum cryptography is widely deployed — creating a window in which early quantum capability is a significant asymmetric advantage for its owners. The question of who owns that advantage — nation-states, a small number of technology companies, or a broader open ecosystem — is a governance question masquerading as a technical one. Second, the optimisation gains that make quantum computing commercially interesting (logistics, energy grids, financial modelling) will disproportionately benefit large organisations with the capital to deploy quantum infrastructure. The productivity gains will be real; their distribution will not be automatic. Neither risk is a reason to avoid the sector. Both are reasons to watch the regulatory environment closely, particularly the NIST post-quantum cryptography standardisation timeline and the emerging US and EU export control frameworks for quantum hardware.
Monday, 27 April. The week opened with a clarification that removed what was left of ambiguity in the energy market. The Trump administration confirmed that the US naval blockade of Iranian oil exports would remain in place until Iran agreed to a comprehensive nuclear deal — not a ceasefire extension, not a pause in hostilities, but a full nuclear framework. WTI moved from $94 to $99 on the announcement. By Tuesday it crossed $100 for the first time since 2022.
Tuesday, 28 April. Microsoft reported first-quarter results that answered the most important single question in technology investment: is the AI infrastructure buildout a real cycle or an expectation cycle? Azure revenue grew 40% year-over-year. Capital expenditure for the quarter was $34.9 billion. Full-year 2026 capex was guided to exceed the already-elevated 2025 figure. The binding constraint on Azure growth, Microsoft noted in its earnings call, is not demand. It is supply. The data centres cannot be built fast enough.
Wednesday, 29 April. Alphabet reported $109.9 billion in revenue, up 22% year-over-year. Google Cloud grew 63%. The company raised its full-year 2026 capital expenditure guidance to $180–190 billion, with a signal that 2027 capex would “significantly increase.” Later that evening, Meta reported revenue up 33% but guided to capital expenditure of $125–145 billion for 2026 — up from the prior range of $115–135 billion. The stock fell 7% after hours on concerns about return on investment timelines. The market, apparently, can simultaneously believe that AI infrastructure is the most important investment cycle of the decade and that this particular company is spending too much on it. The Federal Reserve also held interest rates steady at its April meeting — an 8–4 vote, the most internal dissent recorded since 1992. Three members opposed any signal of future cuts; one member called for an immediate reduction. It was the most divided Federal Open Market Committee in a generation, and the bond market barely registered it, because the earnings tape was louder.
Thursday, 30 April. The Justice Department dropped its criminal probe of Kevin Warsh. Senator Tillis immediately withdrew his hold on the confirmation. The Senate Banking Committee voted 13–11 to advance Warsh to the full Senate — the first fully partisan Federal Reserve chair vote in committee history. The Taylor Rule regime is no longer a probability. It is a schedule. PCE Core for March 2026 came in at 3.2% year-over-year, the highest reading since November 2023, with energy goods up 11.6% as Iran war costs pass through directly into consumer prices. Jerome Powell confirmed he would remain on the Federal Reserve Board after stepping down as chair.
Friday, 1 May. Labour Day closures across Europe and Asia muted trading. The UAE’s departure from OPEC+ became effective. Markets closed the week with the S&P 500 at 7,230, WTI at $105, and the 10-year yield at 4.39%. The week that was meant to clarify three things clarified all three. The clarity, as it turned out, pointed in three different directions.
The appropriate portfolio posture for the Year of the Repricing is not directional exposure to any single asset class — it is positioning for the spread between them. Long energy infrastructure (TLN, BE) and short long-duration Treasuries is one expression of this. Long copper (supply structural, not cyclical) against a hedge on energy (UAE OPEC+ exit as the offsetting force) is another. The key error to avoid is treating the AI capex confirmation as a broad equity green light when the Warsh regime change simultaneously removes the forward guidance safety net that justified expensive multiples.
When Microsoft says that Azure’s growth is constrained by supply rather than demand, it is making a statement about the industrial economy as much as the technology economy. Data centres require land, power, cooling infrastructure, specialised transformers, and fibre. The lead time on a new hyperscale facility is eighteen to thirty-six months. The lead time on a new solid-oxide fuel cell deployment, per Bloom Energy’s Q1 results, is three to nine months. That gap — between the speed of demand and the speed of grid-connected supply — is where the behind-the-meter power thesis lives.
Torsten Sløk published a chart on 22 April showing that high-AI-adoption sectors are pulling ahead in new firm creation at a rate not seen since the early internet era. This is not a productivity story yet — it is a business formation story. The AI revolution is showing up at the foundational layer of the economy before it shows up in GDP statistics. By the time the statistics catch up, the early positions will already be crowded.
The enterprise question that matters most right now is not “what can AI do?” It is “which AI tasks are already being done that nobody has admitted to yet?” The answer, from every earnings call this week, is: more than the market was pricing six months ago, and less than the capital expenditure guidance suggests will be true in twenty-four months.
The Trump administration’s decision to anchor the naval blockade to a nuclear deal precondition was the most significant signal of the week — not because it was surprising, but because it removed the ambiguity. Markets had been pricing a “tactical pause” scenario in which the blockade was a negotiating lever that would eventually be withdrawn in exchange for something short of a comprehensive deal. That scenario no longer exists. The precondition is explicit, public, and has been restated by three senior officials. The resolution pathway is measured in months, not weeks. Iran’s parliament has been hostile to the nuclear framework; the IRGC and the Foreign Ministry have been operating on separate tracks since the ceasefire collapsed. Any resolution requires both tracks to align, which has not happened once in the past six weeks.
Force majeure clauses remain active. Three major shipping insurers have not removed them. The insurance signal continues to be the most reliable leading indicator — more reliable than diplomatic statements, more reliable than FM press releases, and considerably more reliable than social media.
QatarEnergy’s chief executive confirmed in late April that Iranian missile and drone strikes on Ras Laffan industrial city knocked out 12.8 million tonnes per annum of LNG production capacity. The repair timeline is three to five years, per the CEO’s own statement. This is not a Hormuz story. It is a Qatar story. It exists regardless of what happens in the Strait. Global LNG supply has a structural 12.8 mtpa gap for the next three to five years, and that gap does not resolve when the ceasefire does. The market has not priced this fully — it is treating the Qatar situation as collateral damage of the Hormuz narrative rather than as an independent structural shift in global LNG supply.
The UAE’s departure from OPEC+ should be read as a geopolitical statement as much as a supply decision. The UAE was attacked by Iranian missiles and drones. It has now exited a cartel led by its largest regional rival, Saudi Arabia, that has maintained Iran-adjacent solidarity on production policy. Saudi social media (Arabic-language) was unusually heated in the immediate aftermath of the announcement — the tone was not consistent with a managed negotiated exit. The signal from OPEC+ cartel cohesion is now as important as the signal from Hormuz operational status. The WMP has been watching the Strait; the UAE exit revealed we should also be watching the cartel.
UK Gilt: 4.52% (+26bps vs US 10Y 4.39%). German Bund: 2.78%. The sovereign spread divergence reflects implicit energy exposure scoring by geography — UK and European sovereigns are pricing in energy import risk that US Treasuries are not. This is the kind of cross-asset divergence that resolves slowly and then all at once.
The PCE Core reading for March 2026 — 3.2% year-over-year, highest since November 2023 — was treated as unambiguously bad news for rate expectations. The 10-year yield moved from 4.31% to 4.39% in the session. CME FedWatch moved to zero probability of a June cut. The narrative was settled by lunchtime.
Here is the contrarian read. Energy goods contributed 11.6% to the March PCE Core reading. This is a direct Iran war pass-through cost — not a structural inflation dynamic rooted in wage growth, services demand, or housing. Strip out the energy component and core PCE is running below 2.5%, comfortably within the pre-war range. The Warsh Taylor Rule framework is going to inherit an inflation number that is structurally inflated by a geopolitical shock that has a defined resolution pathway (a nuclear deal). When that pathway clears — whenever that is — the energy component reverses sharply, and the Taylor Rule output changes dramatically in a single quarter.
This is not a case for buying long-duration Treasuries now. It is a case for understanding that the rate path is not as structurally high as the current PCE print suggests. The fixed income Red call from Week 15 remains correct for the current window. But the horizon matters: the investor who locks in a short-duration position assuming 4.5%+ yields indefinitely is making a different bet from the one who is short long-duration as a tactical call against an energy-inflated PCE that will eventually revert.
With the Federal Reserve regime changing and PCE Core at 3.2%, the ERDR universe is functioning as designed: floating-rate instruments (Strategies 1 and 6) are outperforming fixed-rate long duration, while the private credit and specialist income strategies are benefiting from a higher base rate environment. The summary verdict: hold floating rate, reduce fixed income duration to zero, watch the energy-linked strategies for directional spread movement.
| # | Strategy | Current Yield | Spread vs IG | WoW Move | Thesis Status | Action |
|---|---|---|---|---|---|---|
| 1 | Active income fund + Lombard facility | 8.5–10% | +480bps | Stable | Floating-rate structure benefits from higher-for-longer; Lombard LTV covenant intact | ADD |
| 2 | IG / split-rated CLO tranches (BBB/BB) | 8–9% | +265bps | +5bps | CLO mezz performing; corporate credit quality holding; watch energy sector CLO exposure | HOLD |
| 3 | Listed infrastructure debt and equity | 6–7% | +190bps | +8bps | Energy infrastructure sub-sector supported by Hormuz premium; regulated utility sub-sector mildly compressed | HOLD |
| 4 | Business Development Companies (BDCs) | 10–12% | +720bps | +12bps | Floating-rate portfolios repricing upward with base rate; watch middle-market borrower stress if rates stay elevated into H2 | HOLD |
| 5 | Agency mortgage REITs | 13–15% | +540bps | Stable | Duration risk is the primary exposure; prepayment optionality negative in rising rate environment. Underweight vs prior guidance. | WATCH |
| 6 | Senior secured leveraged loans | 8–9% | +590bps | +7bps | Floating rate primary benefit; Warsh regime change extends the carry window; watch for leveraged buyout defaults in energy-exposed portfolios | ADD |
| 7 | Preferred shares and hybrid capital | 6–7% | +220bps | +10bps | IG-adjacent preferred holding up; bank hybrid capital spreads slightly wider on rate uncertainty. Neutral. | HOLD |
| 8 | Real asset royalties | 3–5% yield + capital appreciation | +440bps | +18bps | Energy royalties in strong carry; mining royalties supported by copper thesis. Premium capture on oil price above $100 is material for well-structured royalty structures. | ADD |
| 9 | EM hard-currency sovereign carry | 7–8% | +320–590bps | Wider | Gulf sovereign spreads specifically affected by Iran war risk. Broader EM carry still attractive but energy-importing EM sovereigns carry embedded oil shock risk. | WATCH |
| 10 | High-yield municipal bonds (US) | 5–6% (tax-equivalent ~8–10%) | +110–200bps | +6bps | Tax-equivalent yield still attractive vs taxable HY for high-bracket investors. Energy-dependent state/municipality credits require monitoring (Gulf Coast exposure). | HOLD |
| 11 | Private credit direct lending | 10–13% | +840bps | Stable | All-in floating-rate yields now the highest in the post-GFC era. Vintage 2025–2026 direct lending is structurally attractive. Illiquidity premium fully earned in current environment. | ADD |
| 12 | Trade and supply chain finance | 9–12% | +480bps | +22bps | Hormuz disruption has tightened trade finance spreads on Gulf routes as insurers have repriced. This is a direct carry benefit for strategies that can price the disruption premium. Self-liquidating structure remains the key credit protection. | ADD |
Three companies the current macro environment makes analytically interesting. Not a recommendation — a thinking framework. Every entry is scored after four weeks.
In the middle of a week that confirmed an energy war, a Federal Reserve regime change, and a $725-billion technology spending cycle, it is worth pausing on what has not changed. Global extreme poverty rates — the percentage of humans living on less than $2.15 per day — have fallen from 36% in 1990 to below 9% today. This is not despite the last thirty-five years of geopolitical turbulence, financial crises, and supply shocks. It is alongside them. The long arc of human progress has a remarkable tendency to continue regardless of what the tape is doing on any given Friday.
The chart below tracks global extreme poverty over the past forty years. The slope is relentlessly downward. The Iran war does not show up in it, because it has not happened yet. The global financial crisis shows up as a slight pause. The COVID-19 pandemic shows up as a noticeable dip, the first meaningful reversal in three decades — followed by a resumption of the prior trend. The current energy shock will show up as something. It is not yet clear how large. What is clear, from the long view, is that the base rate of human economic progress is strongly positive.
The Soul of a New Machine — Tracy Kidder (1981). A team of engineers at Data General in 1978 races to build a new minicomputer before the competition does. Kidder spent a year embedded with the team and won the Pulitzer Prize for what he wrote. This is the book that best captures what it actually feels like to build something technically complex under extreme pressure — the obsession, the exhaustion, the moments of genuine doubt, and the particular kind of pride that comes from shipping something that should not have been possible in the time available. Forty-five years later, with every technology company on earth in a race to build AI infrastructure faster than its rivals, The Soul of a New Machine reads less like history and more like a dispatch from the present week.
Baseline: 1 January 2026. Prices as at Friday 1 May 2026 close (Labour Day closures: EU/HK/India use 30 Apr; Japan uses 28 Apr for Golden Week). 25 assets ranked best to worst YTD.
| # | Asset | 1 May 2026 | YTD % |
|---|---|---|---|
| 1 | WTI Crude | $105.00 | +66.1% |
| 2 | Baltic Dry Index | 2,730 | +37.9% |
| 3 | USD / TRY | 45.17 | +27.6% |
| 4 | Nikkei 225 | 59,917 | +15.6% |
| 5 | Russell 2000 | 2,813 | +13.3% |
| 6 | Nasdaq 100 | 27,710 | +9.9% |
| 7 | Gold ($/oz) | $4,661 | +7.4% |
| 8 | S&P 500 | 7,230 | +5.6% |
| 9 | Silver ($/oz) | $74.50 | +5.5% |
| 10 | Copper ($/lb) | $5.90 | +3.8% |
| 11 | FTSE 100 | 10,322 | +3.8% |
| 12 | Euro Stoxx 50 | 5,882 | +2.5% |
| 13 | HYG (High Yield) | 79.87 | +2.2% |
| 14 | Nifty 100 | 24,896 | +2.0% |
| 15 | MSCI EM | 1,600 | +0.3% |
| 16 | Swiss SMI | 13,136 | −0.8% |
| 17 | LQD (IG Corp) | 108.03 | −0.9% |
| 18 | DAX | 24,292 | −1.0% |
| 19 | Hang Seng | 25,777 | −2.1% |
| 20 | AGG (US Agg Bond) | 99.00 | −3.1% |
| 21 | USD / ZAR | 16.68 | −5.0% |
| 22 | TLT (Long Tsy) | 85.58 | −9.2% |
| 23 | Bitcoin (BTC) | $78,400 | −10.8% |
| 24 | Natural Gas | $2.784 | −20.8% |
| 25 | Ethereum (ETH) | $2,270 | −23.5% |
Notable moves: WTI +66.1% (Hormuz structural premium). BDI +37.9% (shipping disruption). TLT −9.2% price (bond bear market obscured by coupon; total return −0.75% YTD). Ethereum −23.5% worst performer. Nifty recovered to +2.0% from −2.1% last week.
The Week 16 ranking is the 2026 thesis made visible. WTI at +66.1% YTD, long-duration bonds in the bottom five, digital assets negative, equities in the middle: four asset classes already moving in different directions, four months in. The falsifiable claim at the heart of the Year of the Repricing — that at least three of the five major asset classes will move in different directions by December — has been answered ahead of schedule. What remains to be seen is whether the spread between them widens or closes in the second half.
| Company / Ticker | Entry price | Current price | Return | Original thesis (one line) | Score date |
|---|---|---|---|---|---|
| Freeport-McMoRan NYSE: FCX · first mentioned Wk 14 |
$65.49 19 Apr 2026 |
$56.68 1 May 2026 |
−13.4% | Supply story, not demand. No new mine online for 10–16 years. AI + energy transition + tariff front-running hitting simultaneously against fixed supply. Current weakness is a Grasberg volume miss, not a thesis breach. Re-entry signal: US-China tariff framework before 9 Jul 2026. | Week 18 17 May 2026 |