Three things happened this week that, taken together, tell you exactly where the thesis stands. First, the July jobs print (-23,000 against a consensus of +80,000) arrived Friday morning and, in one number, ended the Fed hike conversation that had been simmering since the July 29 hold. Second, the Strait of Hormuz diplomatic theater continued its now-familiar routine: Treasury Secretary Scott Bessent said Tuesday a deal could land "today or tomorrow," oil fell sharply, Iran published a draft plan Thursday that would permanently ban U.S. and Israeli vessels, oil bounced $3, and the week ended with Brent down roughly 9% regardless. Third, and most quietly, LEU reported Wednesday and delivered a 73% EPS beat, disclosed a record $4.5 billion backlog, and announced a new $900 million DOE HALEU enrichment contract. The S&P 500 closed at a record 7,757 on Friday, up 3.6% for the week, its strongest performance since April. The Nasdaq added 5.2%, powered by a chip-stock snapback. The week's noise was loud. The signal was louder.

The July jobs number did not break the economy thesis. It broke the hike thesis. Those are not the same thing, and conflating them is the most expensive mistake you can make right now.

The NFP Was Good News for the Wrong Reason, and the Market Priced It Correctly

The U.S. economy unexpectedly shed 23,000 jobs in July, following a downwardly revised 20,000 gain in June and coming in well short of the 80,000 increase forecasters expected. The revisions made it worse: May was cut by 66,000 and June by 37,000, leaving the two months combined 103,000 lower than previously reported. That is not a rounding error. That is a labor market that has been weaker for longer than anyone was willing to admit, and the picture gets a little more complicated when you notice that the unemployment rate actually ticked down to 4.1% as the labor force participation rate fell again, to 61.4%. People leaving the labor force pulls the unemployment rate down in a way that has nothing to do with the economy creating good jobs.

Markets read it correctly: stocks closed higher after weaker-than-expected labor data reduced the odds of a Federal Reserve rate hike next month. That is the right interpretation. The hike risk, which had briefly been the live conversation after the hawkish July 29 hold and Fed Governor Cook's Thursday warning that she was "prepared to act" because "inflation is too high" and the risks are "higher on the inflation side," collapsed on contact with the print. What replaces it is less comfortable: a labor market that may be softening not because tariff uncertainty cleared but because the post-AI-buildout structural hiring has not yet compensated for the sectors that are quietly contracting. The majority of net new jobs are still being created in a handful of sectors driven by structural trends like AI infrastructure and healthcare. That is a feature of our thesis, not a bug, but it is not the broad-based employment strength that would let the Fed declare victory.

The practical upshot for the week ahead: the hike is off the table, an early September cut is not obvious, and the next genuinely consequential data point is CPI, due Wednesday, August 13. If services inflation is still sticky, the Fed stays frozen. If it cracks, the conversation shifts decisively toward cuts, the two-year compresses further, and the growth-rate-sensitive names in the book get a secondary tailwind. We are not betting on either outcome. We are positioned for the structural thesis regardless, and the NFP simply removed the tail risk that mattered most.

AI Chips and Hyperscaler Capex: The Funding Debate Is Over

The earnings that really drove this week's tape were not from this week, but the market spent the week pricing them in properly for the first time. Amazon's cloud segment expanded 37% year over year in Q2, surpassing Wall Street's expectation for 31% growth, marking AWS's fastest expansion since 2021. Microsoft's Q4 fiscal 2026 report showed Azure up 43% year over year, capping the first full year in which Azure crossed $100 billion in revenue. Commercial remaining performance obligations swelled to $678 billion, up 84%, and Microsoft 365 Copilot passed 30 million paid seats. These are not "AI is coming someday" numbers. These are monetization numbers.

Microsoft, Google, Meta, and Amazon plan to spend roughly $725 billion combined on capex in 2026, up about 77% from roughly $410 billion in 2025, with analysts projecting the number tops $1 trillion in 2027. The funding debate, which consumed most of July's volatility, is over. Amazon and Microsoft's strong results, and the evidence they showed of AI monetization, along with guidance for cloud services growth acceleration, initiated the snapback rally in tech and the AI complex. For the AI chip leaders in our book, including NVDA, AMD, AVGO, and MU, the hyperscaler capex trajectory is contracted demand. The question was never whether they would spend. The question is whether the physical infrastructure can keep up with the check.

Power and Grid: GEV's Miss Is the Lesson, Not the Headline

GE Vernova reported Q2 on July 22 and slid on the day after earnings came in below the consensus EPS estimate. GEV is navigating a $176 billion backlog and recent contract wins despite challenges including weaker adjusted EBITDA. The market punished the near-term margin softness and moved on. What the market has not fully priced is the structural position: GE Vernova is the world's largest gas turbine maker, and demand is so intense that factory slots are already being booked for 2030. That is not a cyclical story. That is a capacity-constrained, multi-year revenue lock-in that the quarterly EBITDA line will not tell you.

GEV's Q2 orders reached $24.2 billion, and the backlog continues to expand driven by equipment growth in both Power and Electrification segments. Hyperscalers spending hundreds of billions on AI infrastructure need power faster than legacy electrical grids can deliver it, and rather than waiting years for grid interconnects, they are seeking natural gas turbines to generate reliable behind-the-meter power on-site. That is the thesis in plain language. The quarterly noise around EBITDA margins is the cost of owning the company that physically cannot build turbines fast enough. GEV is a core name in our power book, and the post-earnings dip changes nothing about the conviction.

The Hormuz situation reinforces the power thesis from a different angle. Brent crude ended the week near $82 a barrel, with both contracts tracking toward roughly a 9% weekly loss as diplomats traded optimism and Iran published conditions. Iran's apparent draft plan would ban U.S. and Israeli ships from transiting the strait, with other nations that have harmed Iran also restricted until compensation is paid. The volatility is not resolved, and until the strait is genuinely open, the energy-price floor for U.S. gas generation remains elevated. Firm-power names VST and CEG benefit from exactly that dynamic. Oil at $80+ is not a global economic catastrophe, but it is a floor under domestic gas generation economics, and that is a tailwind for the merchant power book.

Fuel Cycle: LEU Delivers the Week's Cleanest Beat

Centrus Energy reported Q2 2026 on August 5, posting EPS of $1.77 against analyst expectations of $1.02, a 73.5% beat. Revenue came in at $176.1 million, a 14% year-over-year increase, and the company reported a record backlog of $4.5 billion. The headline EPS number is interesting. The $900 million DOE HALEU enrichment contract disclosed alongside the results is the actual story.

Centrus won a $900 million DOE HALEU enrichment contract as the backlog reached $4.5 billion, with the 2026 outlook reaffirmed. We have been explicit about the frame for this position: it is a domestic reshoring and national security story, not a bet on any single SMR timeline. The DOE continues to fund HALEU capacity regardless of which reactor design crosses the finish line first, because the alternative is dependence on Russian enrichment for the next generation of U.S. nuclear capacity. That is not a policy anyone in Washington wants to defend. LEU is the only domestic HALEU enrichment operation at scale, and this week's contract award is the federal government putting another $900 million behind that logic. The quarterly EPS beat is pleasant. The contract is the thesis.

What's on Deck: August 11-15

The Hormuz situation deserves its own framing for next week. Conflicting accounts of bilateral engagement are muddying the progress of any Hormuz deal, with President Trump saying Iran will end "pretty soon" while Tehran accused him of staging "theater diplomacy." The market has been conditioned to rally on Bessent optimism and give back some of those gains when Tehran clarifies its actual terms. That pattern held perfectly this week. Until a signed agreement is in hand, the energy bid remains partially intact, and that is fine for our power and firm-energy names.

Forward Thesis by Pillar: Conviction and What Would Change It

AI Chips / Hardware (NVDA, AMD, AVGO, MU). Conviction: high. This week's hyperscaler capex confirmation settled the last structural objection to the thesis. Q2 2026 earnings season settled one of the biggest debates on Wall Street: the hyperscalers are converting their AI infrastructure spending into accelerating cloud revenue, not just capex line items. That is the demand side of the chip equation locked in for the next several years. The AMD post-earnings selloff last week, which we addressed in the Tuesday mid-week letter, was a valuation tax on a perfect quarter, not a thesis event. The master risk tell remains unchanged: a decisive weekly break in NVDA's long-term trend, not a single-session headline. NVDA is set to report on August 27. Everything between now and then is positioning, not fundamentals.

Power and Grid (GEV, VST, CEG). Conviction: high and getting higher. Combined hyperscaler AI capex is running at roughly $725 billion in 2026, and the binding physical constraint is not silicon, it is power and the wires that carry it. GEV's Q2 miss on EBITDA is not a thesis break; it is the margin profile of a company running its factories at the edge of capacity while expanding them simultaneously. The Hormuz situation keeps the domestic gas-generation economics firm, which is a secondary tailwind for the merchant power names in the book. What would change our view: a sustained reversal in data center build decisions, or a credible regulatory freeze on gas-fired generation. Neither is on the horizon.

Fuel Cycle / Energy Security (LEU). Conviction: increasing. Wednesday's 73.5% EPS beat and the $900 million DOE HALEU contract are the two cleanest data points we have gotten from this pillar in months. The frame is reshoring and national security, not reactor-timeline speculation. The federal government is funding domestic HALEU enrichment because the alternative is strategically untenable. A record $4.5 billion backlog means visibility on revenue that the quarterly EPS line does not fully reflect. What would change the view: a material change in DOE funding posture or a political reversal on domestic nuclear investment, neither of which is visible in the current legislative environment.

GLP-1 / Healthcare (LLY, NVO). Conviction: intact. LLY's $23.0 billion Q2 print, up 48% year over year against a consensus of roughly $20.9 billion, was addressed in Tuesday's letter. It requires no amendment. The metabolic drug franchise is the fund's deliberate off-AI diversifier, and it continues to earn that role. Nothing this week changed the read.

AI Software / Platforms (NOW, PANW, NET). Conviction: selective. The hyperscaler AI monetization results are a positive read-through for enterprise software adoption curves. PANW and NET are working. NOW remains the one in repair. The pillar benefits from the tailwind of proven AI ROI at the cloud layer, but software multiples are the first thing that gets trimmed when the two-year yield rises. A CPI surprise to the upside on Wednesday would be the near-term risk to watch.

Financials / Rate Regime (BAC). Conviction: steady. The NFP print removed the hike, which is the scenario that would have hurt BAC most. The "higher for longer but not higher" rate environment is where BAC is best positioned: net interest income stays supportive, credit quality is not yet under stress. The labor market softening bears watching, but the unemployment rate at 4.1% is not a credit event. What would change the view: unemployment moving materially above 4.5% with rising charge-offs, which is not the picture today.

The week's one-sentence summary: the hike narrative is dead, the AI buildout is funded and monetizing, LEU won a $900 million contract that nobody outside the fuel cycle read was watching for, and the Hormuz deal remains the market's most overtraded non-event. The record S&P close at 7,757 is not the story. The pillars behind it are.

This document is for informational purposes only and does not constitute investment advice or a solicitation to buy or sell any security. Nova One Capital and the Nova X Convergence Fund hold positions in securities discussed in this letter, including names referenced across the AI chips, power and grid, fuel cycle, GLP-1, and financials pillars. Past performance is not indicative of future results. All views expressed are those of the author as of August 8, 2026, and are subject to change without notice. Investors should conduct their own due diligence and consult a qualified financial advisor before making any investment decisions.