Wrapping up the Summer…

I will start today’s comments with this overview by Market Screener this morning. It seemed to cover the highpoints well, my thoughts on those comments follows.

The final week of August will test three of the market’s strongest convictions: that Washington can contain the economic damage from Iran, that inflation will cool without a painful intervention from the Federal Reserve, and that artificial intelligence can justify the extraordinary sums being spent on it. Scott Bessent, Kevin Warsh and Nvidia will each provide part of the answer.

Stocks continued to advance over the summer, supported by resilient economic activity, enormous AI investment and investors’ stubborn refusal to stay pessimistic for long. But the rally has become less convincing since mid-August. Last week, all three major US indices declined as higher energy prices and rising bond yields hit technology stocks particularly hard.

The old concerns have returned in a more troublesome combination. The conflict in Iran is keeping oil expensive and adding to inflation pressure. The federal deficit is making investors more sensitive to the supply of government debt. AI spending remains vast, but the questions about returns are becoming more pressing. 

Treasury Secretary Scott Bessent takes center stage on Monday with what the administration has called an “economic D-Day” against Iran. Washington is promising its largest financial offensive yet, potentially including sanctions against countries and companies that trade with Tehran.

Iranian exports have already been disrupted and offers to China reduced. Further restrictions could tighten the oil market just as negotiations to reopen the Strait of Hormuz remain stalled. Brent crude rose more than 6% last week before falling 1.7% on Monday to about $91 a barrel. West Texas Intermediate dropped roughly 2% to $85.

More effective sanctions could push oil higher, add to consumer prices and increase the odds of another interest-rate hike. Traders already fully expect one quarter-point increase by the end of 2026, even though a relatively benign inflation report earlier this month reduced the likelihood of an immediate move.

More effective sanctions could push oil higher, add to consumer prices and increase the odds of another interest-rate hike. Traders already fully expect one quarter-point increase by the end of 2026, even though a relatively benign inflation report earlier this month reduced the likelihood of an immediate move.

Bessent must also reassure the bond market. The 30-year Treasury yield reached a 19-year high of 5.337% last week as investors worried about inflation and the government’s growing borrowing needs. The Treasury responded by increasing buybacks of long-dated debt, briefly producing a rally. By Friday, however, long-term yields had largely returned to where they began. On Monday, the 10-year yield eased to 4.714% and the 30-year yield fell to 5.248%. 

That leaves Federal Reserve Chair Kevin Warsh with an awkward assignment at Jackson Hole on Friday. Investors want to know how the Fed views the Treasury’s intervention, stubbornly high long-term yields and the risk that oil will reignite inflation. As one analyst put it, it is no longer obvious whether the long end of the yield curve is controlled by the Treasury, the Fed or the bond market. The least comforting answer is probably the most accurate: the bond market still gets the final vote.

The week’s other test arrives Wednesday, when Nvidia reports quarterly results. The company is now the world’s most valuable, making its earnings less a corporate update than a progress report on the entire AI investment cycle.

My Response to these comments are:

1. Keep in mind that Trump Market Manipulation remains the biggest factor as it has since his inauguration. He will not stop the nonsense on his own, and unless voters, not the Democratic Party Old Leadership, take over.
2. Bessent and Trump do not want the long end of the yield curve on interest rates to go higher and they are not going to change their view.
3. Today, later when I have some time, I am going to review the take on Warsh that I published on June 15, see our front page Understanding Warsh, Maybe there will be some things to thing about before he speaks on Friday. If anything jumps out I will speak to it in the coming days.
4. As to how much is being spent on the AI build out, I have to say that the more I use AI tools the more impressed I am in how they can make short-term decisions more efficient. It would seem to me that small companies will benefit the most from this, as the big companies will waste too much time thinking about big-unanswerable questions. And all the talk about people losing jobs due to AI will be found to be largely baseless as the quality of the questions asked of AI is the most critical part. That will take smart, practical people. AI does not know what it knows, it is primarily just a lot of data. It only knows what it knows when you force it to analyze something. The fact that information from the period before the internet is difficult for it to find will make it difficult to answer big picture questions.
5. And, another thing to digest on AI is that the platforms are not in any way equal. To me Grok may be the least useful with an underlying cultural agenda. For everyday work Microsoft’s CoPilot tends to be the most useful. High end elite users have gravitated to Claude, but that probably makes their overall economic impact for everyday usage impractical. The biggest loser could well be Google Search, as asking questions of AI tends to provide more solid results. This is probably largely because which Google answer comes up first is based on which answer paid the most to Google to be number one.
6. And, my last comment will be a repeat of what I said last week on Iran, Trump and Netenyahu lost their excursion into powerful limited war and now their only way to back out of the loss is to do something that looks like a win, like the sanctions, and hope that everyone forgets the debacle.