Steel stocks jump after Trump teases plans for 25% tariffs on steel and aluminum

Momentum stocks dragged the market lower Friday, with stocks like Palantir Technologies, SoundHound AI, Rocket Lab, Robinhood Markets, and GE Vernova continuing a recent slide.

(Robinhood Markets, Inc. is the parent company of Sherwood Media, an independently operated media company.)

The iShares MSCI USA Momentum Factor ETF opened 1% higher and built on those gains before reversing hard early in the session to trade 1% lower as of 11 a.m. ET.

If it closes at these levels, this fund that holds US stocks with the best risk-adjusted trailing returns will have completed a so-called “bearish engulfing candle pattern.” As the name suggests is, this is considered to be a negative technical signal that occurs when, the day after a security rises, it ends up opening above the previous day’s closing price and closes below the previous day’s opening price.

ETFs that track major US stock indexes are higher and short-term yields are falling after the August jobs report continued to confirm the trend of labor market cooling, calcifying bets on a Federal Reserve rate cut this month.

Non-farm payrolls rose by just 22,000 in August, while economists had expected an addition of 75,000. The unemployment rate ticked up to 4.3%, in line with estimates. Revisions to the past two months were also negative, but not as severe as in the July report.

The SPDR S&P 500 ETF was up 0.3% to session highs in the minutes following the release, while two-year US Treasury yields fell below 3.5%.

A report and market reaction like this suggests traders are embracing the idea that the softening in the US labor market is primarily driven by supply-side factors in light of major changes to net immigration, as recently argued by economists at the St. Louis Federal Reserve Bank, and isn’t a worrying sign that the US economy is on the verge of a recession.

With revisions, June’s non-farm payroll growth is now -13,000. That’s the first month of net job losses since December 2020. And the underemployment rate (or U6, which includes the unemployed, those employed part time who want a full-time job, and those who want a job but aren’t looking for one currently) rose to 8.1%, its highest level since October 2021.

Some see this data as much more concerning than the market reaction implies.

“Since a month or two ago, policy hawks, growth bulls (I call them wrong), have been arguing two things. First, sequential growth should perk up because the weakness in the summer was all a function of uncertainty around Liberation Day. Second, focus on the ratios because the unemployment rate is still low,” Neil Dutta, head of US economics at Renaissance Macro Research, wrote. “Both of these views were wrong as we now know. Employment growth is still cooling (there is no uptick in hours either) and the unemployment rate is rising. Bye Felicia!”

For the stock market, AI has been the rising tide that lifts any boat that can loosely be seen as flying its colors.

But in the genesis of the AI trade this morning — the powerful chip designers of the picks and shovels for this gold rush — there’s a little bit of a zero-sum element at play:

Broadcom is flying up double digits on the reported addition of OpenAI as the major customer that’s ordered $10 billion in custom chips, significantly improving its 2026 revenue outlook in the process.

Meanwhile, Nvidia is down 3% and No. 3 US chip player Advanced Micro Devices is faring even worse, as this news comes one day after analysts at Seaport cut that stock to neutral, saying that its AI accelerator business hasn’t gained much traction yet. The Street had been very optimistic about the prospects for its new line of chips.

AMD and Nvidia both reported quarterly sales that exceeded expectations, with guidance for revenues in the current quarter that were also ahead of estimates. Nevertheless, both stocks fell after reporting results. To get a positive reaction as a major AI chip designer this earnings season, it seems you need to have done something so good for your company that it actually hurts your competitors’ outlooks.

As we’ve noted, Nvidia’s data center revenues are extremely concentrated, with just three customers (one of which is suspected to be OpenAI) making up over half of direct hardware sales. And despite the chip designer’s protestations to the contrary, the AI boom is more supply-constrained than demand-constrained. So it makes sense that hyperscalers aiming to equip themselves with state-of-the-art technology are looking to do so from a variety of major suppliers.

In its latest conference call, Nvidia CEO Jensen Huang downplayed the threat of custom chips (or ASICs) muscling in on his turf, and highlighted several of the perceived advantages of choosing his company’s products:

“One of the advantages that we have is that NVIDIA is available in every cloud. We're available from every computer company. We're available from the cloud to on-prem to edge to robotics on the same programming model. And so it's sensible that every framework in the world supports NVIDIA. When you're building a new model architecture, releasing it on NVIDIA is most sensible.

And so the diversity of our platform, both in the ability to evolve into any architecture, the fact that we're everywhere, and also we accelerate the entire pipeline. Everything from data processing, to pre-training, to post-training with reinforcement learning, all the way out to inference. And so, when you build a data center with NVIDIA platform in it, the utility of it is best. The lifetime usefulness is much, much longer.”

“Because our performance per dollar is so incredible, you also have extremely great margins. So, the growth opportunity with NVIDIA's architecture and the gross margins opportunity with NVIDIA's architecture is absolutely the best. And so there's a lot of reasons why NVIDIA is chosen by every cloud and every startup and every computer company. We're really a holistic, full-stack solution for AI factories.”