1. US nonfarm payrolls fell by 23,000 in July, versus expectations for an 80,000 increase. May and June payrolls were also revised down by a combined 103,000 jobs. The unemployment rate slipped to 4.1%, but largely because participation fell.

Markets consequently cut the implied probability of a September Fed hike from above 50% to below 50%.

This pushed the S&P 500 to a record close on Friday: 

* S&P 500: +0.62%

* Nasdaq: +1.30%

* Dow: +0.The power thesis continues to strengthen.

The US EIA expects electricity consumption to rise from 4,195 billion kWh in 2025 to 4,269 billion kWh in 2026 and 4,399 billion kWh in 2027, driven partly by AI data centres and electrification.

Grid constraints, turbine shortages and permitting delays are already becoming major barriers to AI development. 

For the week, the Nasdaq gained 5.19%.

2. Iran/Hormuz risk has worsened again today

This is the most important new weekend development.

Iran says it is close to an agreement with Oman over shipping arrangements in the Strait of Hormuz, but says that agreement alone will not reopen the waterway. Tehran is still demanding broader US concessions, including sanctions relief and compensation.

At the same time, the UAE accused Iran today of firing a missile at an ADNOC-linked vessel in the Strait. No injuries were reported in that incident, but it underscores that maritime risk remains very real.

Brent settled Friday at about US$83.55, up 1.3%, reflecting uncertainty over whether the war and shipping disruption are genuinely nearing resolution.

3. Next week’s market focus has shifted decisively to inflation

After the weak employment report, Wednesday’s US CPI is now the single most important scheduled macro event.

Consensus expectations are around:

* headline CPI: 3.4% YoY

* core CPI: 2.5% YoY

A benign CPI print would reinforce the argument that the Fed can remain on hold. A hot print could recreate the uncomfortable combination of weak labour demand + persistent inflation.

AI & semiconductor watch

The fundamental AI build-out remains strong despite recent volatility in AMD and other semiconductor shares.

$ON Semiconductor(ON)$  Onsemi said AI data-centre revenue is now its fastest-growing business and expects that revenue to more than double in 2026.

$Siemens Energy AG(SMEGF)$ also reported a record quarter, with AI data-centre demand and Middle East infrastructure investment contributing materially to growth.

The important distinction remains:

AI demand ≠ AI stock performance.

Many chip stocks already discount very aggressive future earnings. Physical infrastructure providers often have more visible order books and less dependence on winning a single accelerator architecture.

Selective stocks would be my recommendation for bargain hunting.