The market gave you two clean data points this week and one messy interpretation problem. AMD reported its best quarter ever on Tuesday night: $11.5 billion in revenue, up 50% year-over-year, with data center nearly doubling to $6.7 billion. It guided Q3 to $13 billion, above consensus. The stock fell 9% after hours anyway, because Wall Street had penciled in something bigger and Lisa Su didn't hand it over. That is not a thesis break. That is a valuation tax. Meanwhile, this morning LLY walked up to the plate and hit it out of the park in a way that left no room for the "yeah but" crowd: $23.0 billion in Q2 revenue, up 48% year-over-year, against consensus of roughly $20.9 billion. The whole week now pivots to Friday's July jobs number, and that one actually matters.
Where We Stand Through Wednesday
The S&P 500 closed at a record high Tuesday, surging 1.79% to 7,737, its first record close in two months. The Dow cleared 54,000 for the first time ever, up more than 900 points on the session. The catalyst was a two-fer: strong corporate earnings and falling oil prices as Treasury Secretary Bessent told CNBC he thought a Strait of Hormuz deal was imminent. Brent crude sank about 5.3% to settle at $79.36 per barrel, its lowest level since July 10.
The Nasdaq Composite popped 2.59% to close at 26,584.99, as Palantir Technologies rallied 29%. That move follows a massive earnings beat and guidance raise from PLTR, which we have been watching closely as a physical-AI and government-software bellwether. Arista Networks jumped 14% after the cloud-networking company forecast better-than-expected revenue for the third quarter, which reads directly to the AI infrastructure spending theme and reinforces what MSFT, AMZN, and GOOGL already told us about capex intentions.
On AMD: the results were genuinely excellent. Data center segment revenue more than doubled year-over-year to a record $6.7 billion, now representing 58% of total revenue, fueled by strong demand for EPYC CPUs and Instinct accelerators. The company's shares fell 8.94% in after-hours trading to $472.20 despite beating Wall Street estimates, suggesting investor concerns about valuation and the pace of AI product ramps following a 7% rally during regular trading. The non-consensus read here is straightforward: when a company posts its sixth consecutive quarter of 30%-plus revenue growth, guides Q3 to $13 billion above consensus, and still gets sold, the problem isn't the business. It's that the stock had already priced in more than the company delivered on the whisper. CEO Lisa Su stated the company is "still in the early stages of a multi-year AI adoption cycle" and guided data center to more than double again in 2027. AMD is a name we own in the AI hardware pillar, and this print did not change the thesis one degree.
On LLY: this morning's Q2 print was a different category of beat entirely. Worldwide revenue was $23.0 billion, an increase of 48% compared with Q2 2025, driven by a 60% increase in volume, partially offset by a 13% decrease in realized prices. The company reported earnings of $8.38 per share, beating estimates of $6.07 by $2.31. That is not rounding-error outperformance. As CEO David Ricks put it, "Lilly's momentum continues, as we delivered 48% revenue growth and raised our full-year guidance." LLY is a core position in our GLP-1 pillar, and a print like this, with volume up 60% even as pricing faces headwinds from the Trump drug-pricing framework, confirms the volume story is doing exactly what we expected it to do. Foundayo, the oral GLP-1, is still in early innings and not yet moving the revenue needle materially; the optionality on that launch is effectively free at this point.
What's On Deck: Thursday and Friday
This week's economic data centers on the labor market, with all eyes on Friday's release of the July jobs report. That is the only macro event that matters right now. Context: the June print came in at roughly 73,000 jobs with sharp negative revisions, and unemployment ticked to 4.2%. That was the number that knocked 40-plus basis points of September-cut probability back into market pricing virtually overnight. The next Federal Reserve FOMC decision is scheduled for September 16, 2026, which means Friday's NFP is the most important data point between now and that meeting. A second weak print confirms the labor softening trend and puts September cuts back on the table in a serious way. A bounce-back number reopens the "higher for longer" argument and tests whether last week's yield rally was real.
Thursday carries a dense economic calendar, including jobless claims, which the market will read as the freshest real-time leading indicator ahead of Friday's headline figure. Any claims number that comes in hot will amplify NFP anxiety going into the weekend.
The fund's posture on the macro question has not changed since last Saturday's letter: we treat the cuts-versus-recession tension as genuinely two-sided, and we are not chasing the rate whipsaw in either direction. What Friday's NFP will tell us is not whether to add or reduce duration. It will tell us whether the labor market is cooling in an orderly soft-landing way or in a way that starts to question demand for the AI buildout itself. So far, every hyperscaler's capex guidance, every data center earnings call, and every networking equipment number says that question has a clear answer. We are watching, not repositioning, on a single payroll print.
One name to watch Thursday specifically: LEU (Centrus Energy) is expected to report results around this date. The domestic HALEU enrichment thesis is a national-security and reshoring story independent of rate noise, and we want to see how management characterizes the federal contract pipeline and capacity timeline.
The AI hardware thesis delivered again this week. The market's post-AMD selloff is a valuation negotiation, not a demand signal. Friday's jobs number is the only thing that can actually move the needle on the broader macro frame.
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 in the AI hardware, GLP-1, and energy sectors. Past performance is not indicative of future results. All market data cited reflects publicly available information as of August 5, 2026.