........ from someone who usually gets it right, Dan Niles
The S&P was up 1.7% this past week despite:
1) Russia lowering the bar to use nuclear weapons
2) $NVDA (down 2.7% since reporting on Wednesday) guidance only 1-2% better on revenues and earnings relative to consensus.
This market strength despite Russia’s nuclear stance and Nvidia guidance supports my belief that markets remain strong through year-end & into early 2025 driven by:
1. Trump’s re-election: 1st term S&P CAGR was 14%
a. Expectations for tax cuts & less regulation
b. Deficit will initially go up from an already high 7-8%
(largest for non-war time & non recessionary period)
2. Don’t Fight the Fed
a. Fed started cutting w/ 50 bps w/ more cuts in 2025
b. Other Central banks are also cutting aggressively
3. There is no recession on the near-term horizon
a. GDP growth is ~3%
b. Unemployment is near 4%
4. AI is deflationary (technology substitution for labor)
5. China is finally trying to stimulate their economy
6. Seasonality: Stocks are in best 3 months of the year
I continue to believe that an equal weighted basket of stocks is a great risk adjusted way of capturing this potential upside. In 2016, following the election of President Trump, the S&P gained 5% through year (11/7-12/31/2016) but this was led by the subsectors of Financials, Industrials and Energy (+11% on average) and small caps (Russell 2K +14%.) The information technology sector only gained 2% through year end following the election versus being up 10% (S&P up 4%) year-to-date till the election in 2016.
So far in 2024 following the election, the S&P is up 4% with Financials, Industrials and Energy up 8% along with the Russell 2K. Less regulation and more domestic manufacturing is good for these non-tech sectors. The information technology sector is up 4% but the Magnificent 7 are up only 1% on average since the election (11/4-11/22/2024) if you exclude the 45% surge in $TSLA given the relationship @elonmusk has with @realDonaldTrump.
For Nvidia, while I believe the AI buildout will continue over several more years and that NVDA revenues can double again over that time, I also believe there will be an AI digestion phase in the first half of 2025:
1) Near-term (next 3-6 months) I am trying to be flexible
a. Supply constraints leading to Nvidia customers over-ordering combined with positive seasonality may lead investors to remain bullish
b. Only 1-2% upside to forward revenue & EPS estimates for Nvidia after reporting Q3 results may result in investors being more cautious given industry concerns about “having hit a wall in AI scaling laws”
2) Mid-term I am bearish given supply constraints should end by mid-year 2025 leading to an AI capex digestion phase
3) Long-term (over the next several years) I am bullish given I think revenues will double from current levels as the AI buildout continues
a. $CSCO revenues went up 15.5x over six years from the introduction of the first mass internet web-browser in late 1994, NetScape Navigator, till the peak of the internet infrastructure buildout. Cisco became the most valuable company in the world during this time.
b. Nvidia revenues are up 5.9x over the past two years of the AI infrastructure buildout since the launch of ChatGPT in late 2022. For comparison, Cisco’s revenues were up 3.3x over its first two years of the internet infrastructure buildout. Nvidia has become the most valuable company in the world currently.
My bias is that investors in the near-term want to focus on the positive statements made by Nvidia including:
1) Both Hopper and Blackwell systems having certain supply constraints which are tempering revenues
2) Blackwell shipping in CQ4:24 and demand being ‘staggering
3) Initial Blackwell revenues exceeding previous expectations of several billion dollars
4) The demand for Blackwell to exceed supply for several quarters in calendar 2025
5) Refuting scaling concerns & seeing demand for:
a) pre-training scaling continuing
b) post training scaling which is newer
c) inference time scaling which is newer
My concern is as the $MSFT CEO said at their Ignite conference on AI on 11/19: “Just in the last multiple weeks, there’s a lot of debate of have we hit the wall with scaling laws.” While I am optimistic over the next several years on the AI buildout, I have also been thinking there will be a temporary AI digestion phase for the industry in the first half of 2025 due to ROI concerns. It is worth remembering that Microsoft has guided revenues for the following quarter slightly below consensus when they reported both their June quarter and September quarter. Microsoft and $AMZN are the two biggest spenders on AI infrastructure in the world and are roughly 50% bigger spenders than $GOOGL and $META which are in 3rd and 4th place. The statement above by Microsoft makes me wonder if the digestion phase will be on the earlier side given Microsoft’s close relationship with OpenAI, the creator of ChatGPT.
AI capex for the largest players is up nearly 60% in 2024 and this may slow to the 10-20% range in 2025 unless the ROI on the past 2 years of investment improves. Nvidia revenues were up 102% last year and are estimated to be up 123% this year and the street estimates are for up roughly 50% in 2025. If AI industry capex, does indeed slow to closer to 10-20%, I believe Nvidia will have a hard time achieving those forecasts unless the ROI improves on current investments by their largest customers.
From a macro perspective, one thing I have been watching closely is the surge in 10-year treasury yields. This occurred following 1) the initial 50 bps cut by the Fed on 9/18 and 2) the articulation of inflationary policies by both US Presidential candidates. The 10 year treasury yields surged following the initial 50 bps cut by the Fed from 3.65% on 9/17 to an intra-day high of 4.50% on 11/15 following the election. The stock market and treasury yields surged while credit spreads dropped to a one-year low on 11/12 of 1.05% versus a one-year average of 1.27%. This implies investors believe growth prospects are improving versus fears of inflation picking up.
In summary, with Central banks cutting aggressively and governments stimulating even more, I think stock markets remain strong through the first half of 2025. I continue to believe an equal weighted basket of stocks is the best risk adjusted way to participate in this bullish trend.
As always, the key will be to remain intellectually flexible and data dependent. As Niels Bohr, the Nobel laureate in Physics, said: “Prediction is very difficult, especially if it's about the future!”