Dan Niles

Per @JamesCarville “as the bond market. You can intimidate everybody.” Entire Yld curve from 3m to 30yr hit new 52wk or multi-decade highs. 10yr +15 bps to 5.11%. I continue to advocate not fighting the Fed, bond mkt or seasonality, especially leading up to mid-terms on Nov 3rd.

While oil & equity mkts were flattish last wk, bond ylds hit multi-decade highs in different countries as central banks raised rates. While AI frontier model pacing was initially a concern, that faded by Friday. I expect the upcoming wk to be challenging. I have some simple portfolio rules including: 1) Don’t Fight the Fed: hiking cycle seems to have begun 2) Don’t Fight the Bond Mkt: Multi-decade highs in ylds 3) Don’t Fight Seasonality: Sept is worst month for hit rate & returns with mid-term yrs even worse These rules influence how I think about leverage, the ratio of long positions versus short positions, and individual position sizing. Investing is hard enough without fighting headwinds. As Warren Buffett (a remarkable 61 year run came to an end last week) has said, the market has to keep pitching but you do not need to swing. For AI, the battle lines seem to be drawn. OpenAI and Anthropic are pushing for regulatory capture while $NVDA & $MSFT are pushing for better testing before models are released. I’m in the latter camp. No product from any company should be released before it is safe. AI related companies focused on the potential bullish implications in case there was model innovation pacing including: 1) Focus on security 2) Resources deployed for testing 3) Potential for any training slowdown offset by limited hardware capacity being deployed to the infrastructure layer The SOX index, a representation of the infrastructure layer, rallied 0.8% last week with $IGV (the software ETF) also up 2.8%. Software was led by the security names, the $HACK ETF was up 8.0%, despite the 0.1% decline in the S&P. Longer-term, my view is that companies do not need the most advanced models for 90% of their workflow and usage will increasingly go towards open weight models. I believe that ultimately the LLM layer is likely to become commoditized. As an example, the ASP per token peaked in late May and has gone down ~50% since then while the number of tokens produced has gone up by ~4x. This is Jevons paradox in action. This should also be supportive of the infrastructure layer. As the model layer becomes increasingly commoditized, I believe the winners will be those that have the following attributes: 1) strong open weight models 2) distribution capability 3) training data 4) base business that is highly cash flow generative I believe the following companies have the attributes above to varying degrees: 1) $META (Muse from Meta is the #1 free app on the Apple app store with ChatGPT #2), 2) $GOOGL (I expect a new frontier level model soon that should help the stock) 3) $MSFT (Co-pilot might be the "safe" way ~450M M365 enterprise users choose to access AI. But their 27% ownership of OpenAI does concern me and they do not have the wealth of training data that Meta and Google have.) Earlier this year, the advent of Agentic AI increased token production by 10-100x. This should keep demand strong for the semiconductor portion of the infrastructure layer. Given the escalating geopolitical tensions over the weekend and its potential impact on oil/bond yields, this could prove to be a challenging week: 1) Ukraine fired over 1,000 drones at Russia, including "largest ever" attack on Moscow 2) There was a missile attack on Riyadh for the first time since July On a positive geopolitical note, hopefully something constructive comes from the meeting between Presidents Trump, Xi and the attending business leaders on Wednesday. All the best in the week ahead.

I think this is a combo of the $AMZN deal with $QCOM (+4% intra-day) today which speaks to the issue of $NVDA large customers trying to build custom ASICs. In addition, Nvidia on many days is used as a source of funds when semis are having a strong day and as a result outperforms many days when the Index is down. As an example of this, the correlation of Nvidia to semis was around 0.85 in mid-2025 and has since declined to around 0.61. There is also probably some reversion to the mean trading today given last week Nvidia was up 5.9% with the Semi Index up 2.3% and their biggest thorn in their side on custom ASICs, $AVGO, was down 3.0% last week. Having said all that, it is annoying to not see the stock perform better today.

Last wk despite WTI +10% & ylds +2-7 bps across the curve, S&P/Nas/R2K +0.1%/+0.4%/+0.1%. $NVDA acquisition of Hugging Face & $Meta release of Muse Spark 1.3 last wk make both names more attractive into year-end. My view is that LLMs increasingly bifurcate into 90%+ usage of open-source/ open-weight models in the future as companies optimize the right models for the right task. Since the focus on controlling AI expenditures, the Silicon Data token cost has fallen over 50% since late May but the weekly usage of tokens in models across OpenRouter has increased by 3.6x over this same time. In addition, enterprises are increasingly focused on making sure their own proprietary data does not leak out when they use third party closed frontier LLMs. Hugging Face is the premier central collaborative platform, repository, and toolkit for open-source and open-weight AI with over 18 million developers. Nvidia has three customers that accounted for 44% of their revenues over the past six months and their largest customers are increasingly designing their own ASICs and in some cases selling them externally. A more diversified customer base that owns their own AI compute stack instead of renting from the big cloud service providers would help Nvidia with both of these issues. With this acquisition, Nvidia is in an even better position to sell enterprises a complete alternative AI stack (from the model to chips) where the customer will own their own data. Valuation is also compelling. Nvidia trades at a 15x CY27 PE versus their own guidance for 70% revenue growth and the Big 3 public cloud service providers at 21-23x for 15-26% total revenue growth. The S&P trades at 19x for 9% revenue growth for comparison. Nvidia is also up “just”24% versus the Semiconductor Index up 66% following underperformance last year at up 39% versus 42%. As for Meta, the stock is down 7% year-to-date after being up just 13% last year driven largely by concerns that 1) they can only monetize their near doubling in AI capex spend through efficiencies in their own business and 2) they were falling behind in the AI model race. The launch of the Muse Spark 1.3 API last week, catapulted Meta back to near frontier status (Top four in the Artificial Analysis Intelligence Index out of 10 models) but with aggressive token pricing (Bottom four in Cost per Task.) Open-weight versions of the Muse Spark lineup are coming soon. This will give the company another way to monetize their aggressive capex plans. This follows Meta's settlement in late August with state AGs on their youth addiction trial which was another overhang on the stock. Trading at 16x CY27 PE for 20% revenue growth is compelling with the settlement and Spark 1.3 launch as catalysts. From a broader market perspective, I recommend caution between now and the US mid-terms for reasons I have fleshed out in prior posts including: 1) Don’t Fight the Fed given I believe a 9/16 hike is likely 2) September -0.5% on avg & up only 48% of the time 3) S&P drawdowns of 10% in lead-up to mid-terms 4) Bipartisan pushback against datacenter expansion 5) Iran dragging out hostilities through US mid-terms

Last wk despite WTI +10% & ylds +2-7 bps across the curve, S&P/Nas/R2K +0.1%/+0.4%/+0.1%. $NVDA acquisition of Hugging Face & $Meta release of Muse Spark 1.3 last wk make both names more attractive into year-end. My view is that LLMs increasingly bifurcate into 90%+ usage of open-source/ open-weight models in the future as companies optimize the right models for the right task. Since the focus on controlling AI expenditures, the Silicon Data token cost has fallen over 50% since late May but the weekly usage of tokens in models across OpenRouter has increased by 3.6x over this same time. In addition, enterprises are increasingly focused on making sure their own proprietary data does not leak out when they use third party closed frontier LLMs. Hugging Face is the premier central collaborative platform, repository, and toolkit for open-source and open-weight AI with over 18 million developers. Nvidia has three customers that accounted for 44% of their revenues over the past six months and their largest customers are increasingly designing their own ASICs and in some cases selling them externally. A more diversified customer base that owns their own AI compute stack instead of renting from the big cloud service providers would help Nvidia with both of these issues. With this acquisition, Nvidia is in an even better position to sell enterprises a complete alternative AI stack (from the model to chips) where the customer will own their own data. Valuation is also compelling. Nvidia trades at a 15x CY27 PE versus their own guidance for 70% revenue growth and the Big 3 public cloud service providers at 21-23x for 15-26% total revenue growth. The S&P trades at 19x for 9% revenue growth for comparison. Nvidia is also up “just”24% versus the Semiconductor Index up 66% following underperformance last year at up 39% versus 42%. As for Meta, the stock is down 7% year-to-date after being up just 13% last year driven largely by concerns that 1) they can only monetize their near doubling in AI capex spend through efficiencies in their own business and 2) they were falling behind in the AI model race. The launch of the Muse Spark 1.3 API last week, catapulted Meta back to near frontier status (Top four in the Artificial Analysis Intelligence Index out of 10 models) but with aggressive token pricing (Bottom four in Cost per Task.) Open-weight versions of the Muse Spark lineup are coming soon. This will give the company another way to monetize their aggressive capex plans. This follows Meta's settlement in late August with state AGs on their youth addiction trial which was another overhang on the stock. Trading at 16x CY27 PE for 20% revenue growth is compelling with the settlement and Spark 1.3 launch as catalysts. From a broader market perspective, I recommend caution between now and the US mid-terms for reasons I have fleshed out in prior posts including: 1) Don’t Fight the Fed given I believe a 9/16 hike is likely 2) September -0.5% on avg & up only 48% of the time 3) S&P drawdowns of 10% in lead-up to mid-terms 4) Bipartisan pushback against datacenter expansion 5) Iran dragging out hostilities through US mid-terms

Last wk, SPX/Nas/R2K +0.5%/+0.8%/-1.5% w/ oil -4%. But a hawkish Warsh on Friday led to a bear flattening of the yield curve. Despite $NVDA guide of 70% CY27 rev growth vs consensus of 47%, SOX Index -2.3% while software $IGV +5.9% on solid earnings. In general, many AI investors have been bullish on semiconductors and bearish on software on the belief that AI will displace many point solution software companies. This is why the SOX index is up 62% YTD and IGV is still only up 4% YTD versus the S&P +13%. Situational Awareness was the poster child for this type of positioning. But since the unwinding of the Momentum trade which started on 6/22 (I wrote about these concerns on 6/20), IGV has rallied 25% while the SOX Index has declined 22% through 8/28. For perspective, the Morgan Stanley Momentum index (momentum long performance minus momentum short performance) from 6/22-8/28 is down 36% while their more concentrated TMT index is down 54%. But a bullish twist on AI for the software sector introduced recently is that AI agents will access software tools ~10-100x more often than humans. On 8/6, $TEAM, which was in the bucket of software names widely considered at risk of being replaced by AI, rallied 35% the next day in reaction to solid earnings & outlook. Then on 8/13, $WDAY rallied 18% on the news that private equity firm Silverlake might be pursuing an acquisition which I wrote probably put a floor underneath software. Workday was also supposed to be in the AI crosshairs and private equity has higher bars to clear given their use of leverage and holding period than a typical investor. Then on 8/26, $CRM reported solid results, guidance and a deal with Anthropic (in which they also first invested in May of 2023.) The stock was up 23% in reaction the next day. This seemed to be a strong counterpoint to the SaaS-pocalypse worries. This strategic alliance allows users to execute actions natively inside Claude without needing to open traditional software screens. Salesforce also seems to be changing how they charge customers with fees more related to customer use and benefits to their business. Then on 8/27, Workday reported results which were good enough but arguably acquisition prospects drove more of the stock reaction of +6% the next day from the slightly down opening price. Historically, system of record, security and gaming software have been the only three areas I have liked within software. I now wonder whether the fundamental implications of Atlassian, Workday and Salesforce are supportive of the technical reactions in the software stocks as a group as agentic AI continues to ramp. So how do I square this with my concerns that the rapidly escalating amounts spent on AI by corporations has to come from somewhere? Annualized revenue run-rates for Anthropic and OpenAI have ramped from $29B to start the year to $105B just 7 months later. Software spending globally excluding AI was roughly $1 trillion in 2025. But IT services at $1.7 trillion is a bigger category which I believe still has risk. And finally, knowledge worker compensation is an even bigger category where disruption would be even less noticeable at an estimated $35-50 trillion in 2025 or roughly 30% of the global workforce. Looking forward, the deal on Friday for Venezuelan oil fields that hold the largest crude reserves in the world at 17-18% should get us off to a positive start to the week with declining oil prices. But a bit further out: 1) “Don’t Fight the Fed” given I believe a hike is likely on 9/16 because the 10/28 mtg is right before mid-terms, 2) September has the poorest seasonality of all months, 3) there is even worse seasonality than normal during mid-term election years (see prior posts for more detail) and 4) recent bipartisan pushback against datacenter expansion (one of the few things both sides seem to agree on though I believe this is wrong and hope it will change with more education) puts pressure on the AI infrastructure names. As Warren Buffett says, the market has to keep pitching but you do not need to swing.

Last wk, SPX/Nas/R2K +0.5%/+0.8%/-1.5% w/ oil -4%. But a hawkish Warsh on Friday led to a bear flattening of the yield curve. Despite $NVDA guide of 70% CY27 rev growth vs consensus of 47%, SOX Index -2.3% while software $IGV +5.9% on solid earnings. In general, many AI investors have been bullish on semiconductors and bearish on software on the belief that AI will displace many point solution software companies. This is why the SOX index is up 62% YTD and IGV is still only up 4% YTD versus the S&P +13%. Situational Awareness was the poster child for this type of positioning. But since the unwinding of the Momentum trade which started on 6/22 (I wrote about these concerns on 6/20), IGV has rallied 25% while the SOX Index has declined 22% through 8/28. For perspective, the Morgan Stanley Momentum index (momentum long performance minus momentum short performance) from 6/22-8/28 is down 36% while their more concentrated TMT index is down 54%. But a bullish twist on AI for the software sector introduced recently is that AI agents will access software tools ~10-100x more often than humans. On 8/6, $TEAM, which was in the bucket of software names widely considered at risk of being replaced by AI, rallied 35% the next day in reaction to solid earnings & outlook. Then on 8/13, $WDAY rallied 18% on the news that private equity firm Silverlake might be pursuing an acquisition which I wrote probably put a floor underneath software. Workday was also supposed to be in the AI crosshairs and private equity has higher bars to clear given their use of leverage and holding period than a typical investor. Then on 8/26, $CRM reported solid results, guidance and a deal with Anthropic (in which they also first invested in May of 2023.) The stock was up 23% in reaction the next day. This seemed to be a strong counterpoint to the SaaS-pocalypse worries. This strategic alliance allows users to execute actions natively inside Claude without needing to open traditional software screens. Salesforce also seems to be changing how they charge customers with fees more related to customer use and benefits to their business. Then on 8/27, Workday reported results which were good enough but arguably acquisition prospects drove more of the stock reaction of +6% the next day from the slightly down opening price. Historically, system of record, security and gaming software have been the only three areas I have liked within software. I now wonder whether the fundamental implications of Atlassian, Workday and Salesforce are supportive of the technical reactions in the software stocks as a group as agentic AI continues to ramp. So how do I square this with my concerns that the rapidly escalating amounts spent on AI by corporations has to come from somewhere? Annualized revenue run-rates for Anthropic and OpenAI have ramped from $29B to start the year to $105B just 7 months later. Software spending globally excluding AI was roughly $1 trillion in 2025. But IT services at $1.7 trillion is a bigger category which I believe still has risk. And finally, knowledge worker compensation is an even bigger category where disruption would be even less noticeable at an estimated $35-50 trillion in 2025 or roughly 30% of the global workforce. Looking forward, the deal on Friday for Venezuelan oil fields that hold the largest crude reserves in the world at 17-18% should get us off to a positive start to the week with declining oil prices. But a bit further out: 1) “Don’t Fight the Fed” given I believe a hike is likely on 9/16 because the 10/28 mtg is right before mid-terms, 2) September has the poorest seasonality of all months, 3) there is even worse seasonality than normal during mid-term election years (see prior posts for more detail) and 4) recent bipartisan pushback against datacenter expansion (one of the few things both sides seem to agree on though I believe this is wrong and hope it will change with more education) puts pressure on the AI infrastructure names. As Warren Buffett says, the market has to keep pitching but you do not need to swing.

Here is your $CSCO comparison. Cisco was forecasting sustainable revenue growth of 30-50% in 2000. This is an excerpt from their earning release on May 8th 2001: "The first four months of 2001 were extremely challenging as we went from year-over-year bookings in excess of 70% in November, to 30% negative growth within a span of several months. This may be the fastest deceleration any company of our size has ever experienced," said John Chambers, president and CEO of Cisco Systems. To be clear, I am not saying this is where we are today on a fundamental basis on the AI trade. I think this is a problem sometime in 2027 at the earliest. But my point is to not rely on backlog or bookings that can change very rapidly especially when you have large customer concentration. $NVDA had two customers last quarter than combined were 39% of revenues.

“Don’t Fight the Fed.” Warsh at Jackson Hole: “Price stability is not self-executing… 65 months of sustained, elevated inflation sits squarely with the Central Bank.” Probability of rate hikes just went up going into mid-terms which historically see more than avg downside risk.

James Carville: “as the bond market. You can intimidate everybody.” Despite Bessent’s efforts, bond ylds increased last wk 1-7 bps while S&P/Nas/SOX (AI proxy) -1.4%/-2.1%/-5.4%. Next wk has Iran sanctions (Mon), $NVDA earnings (Wed) & Warsh at Jackson Hole (Fri.) On Monday, the US is likely to put financial sanctions on Iran and any country supporting them in lieu of further military action going forward. The polling numbers in the US for support of the war are poor. But given China, who buys the most Iranian oil, will most likely be exempt from sanctions, I doubt this will have much impact other than what Iran does to retaliate. I believe Iran will try to keep the Straits hostage through at least the mid-terms. The US hostages in Iran were held for 444 days despite financial sanctions. They were released just hours into the inauguration of President Reagan in 1980. The “Gipper” (Knute Rockne, All American is a good nostalgic sports movie but I am a sucker for most of them) crushed the re-election attempt by President Carter by winning 489 electoral votes to 49 due to high inflation, interest rates and gas prices compounded with the foreign policy issues. On Wednesday after the market close, Nvidia reports results. Despite solid results, the stock has been down the next day in reaction for the past four quarters and declined from open to close the past eight quarters (and been down for the full day for six of them.) But trading at just a PE of 17x CY27 vs the S&P at 19x and with the stock down 4% over the past two weeks versus just a 1% decline in the S&P, I think the risk vs reward is good. However, the public (& therefore political) sentiment against datacenters continues to grow going into mid-terms which is likely to restrain multiple expansion. From a fundamental perspective for Nvidia, hyperscaler capex grew 92% y/y and 29% q/q in Q2 which is the fastest ever since launch of ChatGPT in late 2022. Public cloud revenue growth at the Big 3 vendors accelerated to the fastest pace ever at 43% y/y and 15% q/q in Q2 and arguably more importantly their operating margins expanded by 2% to a record 39% in aggregate. This compares to consensus estimates for Nvidia revenue growth decelerating to 13% q/q growth in Q2 from 20% or greater in each of the past 3 quarters. With regards to the AI infrastructure names, the roughly 50% decline in token costs driven by open-weight models since late May is being more than offset by the ~2.5x increase in token production during that time and increase in operating margins. But datacenters need to be put somewhere. When Gallup polling numbers for those opposing their construction locally are worse at 71% than for nuclear reactors at 53%, that is not good. Politicians want to keep their jobs also. On Friday at 10am, Fed Chair Kevin Warsh will speak at the central bankers meeting at Jackson Hole. The S&P declined 1.2% and 1.5% during his last two press conferences following FOMC meetings. The financial picture is even more complicated today: 1) the Iran situation looks likely to drag on for longer, 2) global bond yields are higher, 3) there is intervention in Japanese currency markets, and 4) the actions by the US Treasury to try and lower bond yields is putting downward pressure on the US dollar and upward pressure on inflation. In summary, despite strong S&P earnings growth, I am wary of further declines given the typical season drawdown of 10% from peak to trough during mid-term years since 1990 from 7/31-11/9. I gave more detailed stats last Sunday. Best of luck in the week ahead.