"Beethoven's music tends to move from chaos to order, as if order were an imperative of human existence." ~ Daniel Barenboim

Good evening,

Fundstrat Head of Research Tom Lee's mantra of late – “Just get through August” – has borne fruit, as the bell rang at 16:00 ET today to end the last trading day of the month. Much of the last week of the last month of summer was dominated by expectations, and then digestion of, earnings from Nvidia, the poster child for the popularity and power of AI, which is overturning electricity, chip, server, real-estate, labor, and other markets. AI may even be writing this very note. (One never knows).

Lee pointed out that it is a sign of market strength that NVDA fell -6% post-EPS and yet the S&P 500 managed to end on a positive note. In a research note from earlier this week, Lee reminded us that the probabilities favored post-NVDA results triggering a rotation rather than 'risk-off'. He elaborated, "pundits said 'as goes NVDA, so goes the market', and this is why stocks, in our view, were so soft on Wed. Investors expected the entire market to follow NVDA in either direction. Post-EPS, the largest stock (NVDA) in S&P 500 fell 6%."

Head of Data Science "Tireless" Ken Xuan concurs. During our weekly research huddle, he said, referring to NVDA, "I think the price action is an over-reaction" and followed up with some color: "the NVDA results are solid; revenue growth is slower than before, but the gross is still pretty strong". Referring to the company's ambitious plans for a $50 billion share buyback, Xuan said, "the only share buyback that I can think of that was larger than Nvidia’s was Apple’s $100B buyback".

Addressing the macro environment, Xuan summarized the main news of the week: namely that the upshot of the Jackson Hole Fed meeting was dovish, as Tom Lee had expected, that the discount window rates lowering on Tuesday was a good indicator that the other FOMC members would be dovish at the next meeting, and that Thursday's GDP number was largely in line with expectations.

Sector-wise, Xuan reminded us that, post Tokyo Black Friday, the market had been led by tech and financials, but that those sectors had since retreated, with last week’s large-cap growth seeing a rotation into small-cap value. Still, he pointed out some bright spots among the former outperformers, such as, for instance, consumer finance, supported by AmEx.

Addressing a point of interest for many investors in another sector that seemed to lose its luster lately, Xuan spoke of a much-discussed stock, saying, "I read the SMCI short-seller report; the delay of the 10K tells the market that something is wrong; people are questioning the overall story of AI, and this is not the first time there are accusations of fraud in the AI sphere." He pointed out that SMCI is very dependent on the overall AI environment, and broadened his comments, noting that, in a sense, companies such as Dell and SMCI are all living in NVDA's world. He concluded by saying that "buying the dip is tricky now".

Head of Technical Strategy Mark Newton was somewhat more sanguine, while riffing on Xuan's commentary: "the market is looking past NVDA and more to the Fed’s rate cuts, and to the future of AI". While acknowledging that, based on cycles and sentiment, we may see consolidation soon, and noting that NVDA cycles move higher into October and then see a pullback, Newton also pointed out that "the VIX in the low teens is going to be a steal between now and November" and reminded us that "Fortune favors the bold". Speaking of NVDA, Newton's rejoinder to Xuan's caution took the form of the advice that "any large pullback would be something you want to own". As for the sector overall, Newton said, "we have seen a little bit of the wind taken out of the sails of technology, and we could see a 5% to 7% – potentially a 10% correction – into November, but I still see reasons to be bullish".

In terms of the market as a whole, Newton acknowledged "a lot of sideways action in the S&P, largely mirroring what technology has been doing" but balanced that observation with the crucial detail that the current breadth levels are constructive for the S&P, and pointed to a rising star, saying, "equal-weighted healthcare is on the verge of breaking out into new all-time highs. That, along with financials, is a noticeable tailwind that’s likely to help out the stock market".

Summarizing the month we left behind, Newton mused that "August has turned out better than expected, although that is normal for election years", pointing to "many positive sectors", meaning that "there is a broadening out of performance". "We’ve seen some very good performance out of the sectors that matter, other than technology" he said, giving a high-five to one sector in particular – real estate, specifically REITs – which had had been a "consistent outperformer".

Looking at a few crucial factors of overall market health, Newton noted that he didn't "see sufficient evidence that we are topping", what with the total put-to-call ratio at the lowest level in three years, both the dollar and yields having slowed their rate of decline, interest rates having cracked meaningfully, and stocks having rebounded. Supported by all the evidence above, Newton ended his commentary on a deservedly positive note, "I would expect that markets push a little bit higher, to the tune of 1 or 2%."

Both Newton and Xuan chimed in during Head of Digital Asset Strategy Sean Farrell's comments in support of his thesis, with Newton saying that consolidation at or near 52-week highs is good news and Xuan sounding a consoling note by pointing out that "August and September are always weak months for crypto". When Farrell had the mic, he pointed to the positives: namely that stablecoin flows were a good sign, that the altcoin core-strategy basket had good news, that profit-taking at higher margins was a good sign in terms of momentum and confidence. He noted that crypto has "mostly stayed the course during these past choppy 5 months", and that "still, we’re at 60K".

Elsewhere

As tech giants get more creative about how they source clean power, Meta, the parent company of Facebook, has made a significant move in the renewable-energy sector by agreeing to buy 150 megawatts of power from an upcoming project by Sage Geosystems, a leading geothermal energy company, slated to start supplying Meta’s data centers in the western United States as soon 2027.

It turns out that the U.S. Defense Department, NASA, and other government agencies have collectively funded research leading to more than 1,000 patents for China-based inventors since 2010. What is potentially concerning is that the patents include those in sensitive fields such as biotechnology and semiconductors. Per data provided to the House of Representatives’ Select Committee on the CCP by the U.S. PTO, HHS funded 356 of these patents, the DoD contributed nearly 100, and NASA played a role, too – despite restrictions on cooperation with China. 

California lawmakers have passed a contentious AI safety bill, known as SB 1047, which mandates costly and rigorous safety testing for large-scale AI models, as well as a kill switch for models that go rogue. The law targets systems that either require substantial computer power or cost more than US$100 million to train – although no current AI models are known to have actually hit that threshold. Tech giants Google and Microsoft have opposed the bill – yet, not surprisingly, it has the support of Silicon Valley’s resident contrarian (Elon Musk). In related news, the U.S. AI Safety Institute has said that OpenAI and Anthropic had signed deals with the U.S. government to research and test AI models.

Give or Take a Trillion. A new pair of analyses from the Penn Wharton Budget Model gave marks to the economic proposals of ex-President Donald Trump and Vice President Harris, finding that Trump’s proposals could add ~ $4-6 trillion to the deficit over a decade, depending on economic effects. For Harris, that sum ranged between $1.2 trillion and $2 trillion. Wharton

“The Business of America is Business.” The Chamber of Commerce outlined its tax priorities as Congress prepares to deal with expiring parts of the GOP tax law in 2025. High on the list of priorities are goals such as keeping the 21% corporate tax rate (some Republicans have toyed with the idea of increasing it), restoring immediate expensing for R&D spending, and preserving the 20% passthrough deduction for certain business income.

Finally, in a non-peer reviewed analysis, a student-professor duo at Pomona College found that only 10 of the 43 partly AI-powered investment funds in the market have done better than the S&P 500 during their lifetimes. What’s more, the average annual return for all 43 funds was about 5% per year worse than the S&P 500. Funds fully powered by AI performed even worse. “The Achilles’ heel of AI systems is that while they are unparalleled at finding statistical patterns, they have no way of judging whether the patterns they find are plausible or pointless,” write the authors of the analysis.