Some General Thoughts
Executive Summary
Fed Policy Dilemma: Chairman Warsh’s hawkish Jackson Hole stance has sharply lifted market hike probabilities, despite benign inflation trends and diverging economic conditions between tech capex and interest-sensitive sectors like housing.
Market Breadth & Tech Operating Leverage: Headline indices mask deteriorating market breadth, while Nvidia’s historic operating leverage and 55% to 75% margin profile set it apart from both traditional industrial giants and mega-cap peers.
Structural Gold Bid: Short-term rate squiggles aside, unprecedented and price-insensitive central bank reserve diversification provides durable, multi-year support for precious metals.
The Fed and Chairman Warsh:
While many pundits expected Kevin Warsh to toe President Trump’s line of lowering interest rates following his appointment, things just don’t seem to be setting up for that to be the case. Moreover, following Warsh’s Jackson Hole speech on Friday, it appears as though the new Fed Chair has really boxed himself in with his next policy move. We are at the point where if he doesn’t raise rates at the upcoming FOMC meeting, investors will now be asking why, since his words on Friday single-handedly caused market-based rate hike views to double to around a 60% chance. The question is: what changed since his last talk at the podium in late July, except that economic data have weakened more and inflation data are broadly benign? It begs the question: why didn’t he tighten a month ago? He said he was awaiting new information, and that information caused markets to move from discounting 80% odds of a September hike to 30%, and now back to 60%. David Rosenberg of Rosenberg Research put out the following table in one of his notes last week that does a great job contextualizing the setup on an array of data today versus March 2022, when the Fed embarked on its prior tightening cycle. No variable in this table is in a worse position today (justifying a hike) than it was in 2022. Keep in mind two things: 1.) this is by no means an exhaustive list of inputs used to inform a Fed policy move, and 2.) the Fed Funds rate is +325 basis points higher today.
Look, the Fed is ina tough spot with plenty of evidence to support the case for a hike and hold where they’re at. Put me in the latter camp, and I’ll even go a step further by stating that I think a hike will inevitably be a policy mistake. Based on our work right now, we see inflation plateauing over the next month or two before rolling over and weakening considerably toward the Fed’s 2% target by early Spring next year. As for the economy and growth, parts of the economy are booming (AI capex), while other parts are struggling (housing). It's a complicated issue, but I'd argue that hiking rates enough to even slow the booming area of the economy will - almost by definition - cause the Fed to sacrifice the labor side of their dual mandate.
My view is that the intense focus on inflation stems from the nearly 25% cumulative increase in the cost of living (BLS CPI and Core CPI) since January 2021. The start of this inflationary impulse was due to a heavy dose of fiscal support injected into the economy to pull us out of the COVID-induced recession. Since then, it's been aided and abetted by pro-cyclical deficit spending and rolling supply shock episodes. As a result, lower- and middle-income households with inadequate exposure to asset prices have been caught up in the so-called ‘affordability crisis’. The "man on the street" is not mad because inflation readings are currently running above the Fed’s 2% target. He/she’s mad because a house that costs $250k in his mind is now $600k, and the cost to finance it is ~7%, not 4%.
The world has gone through a robust and abrupt change in the cost of living. Voter anger is rooted in this change. With regard to current Fed policy, by hiking rates, they will be restricting the areas that matter to the general public – housing, autos, appliances, furniture…. If the Fed’s worry is that the economy is running ‘too hot’ as a result of the AI capex buildout, and therefore they have to hike rates to cool things down, I’d argue that to hike enough to slow the AI buildout would require, almost by definition, sacrificing the labor side of the dual mandate. The only thing worse than an inflation shock (which is really behind us now) is an inflation shock followed by a recession/market crash. Not to mention the fact that the rise in oil price is supply-induced and constitutes a tightening (i.e. not something the Fed should respond to). In a nutshell, it’s a tough environment for monetary authorities, but holding here seems wise. Trying to make up for the cumulative increase in inflation over the past five years by hitting growth and labor markets today could result in unintended consequences that create new/worse challenges.
Markets: Equities, bonds, and gold…
So far, early into his tenure, Chairman Warsh’s messages haven’t been welcomed by risk assets – maybe that’s his intention: in his two most recent performances, the S&P 500 closed the day down -0.9% on average, the yield on the 10-year T-note has risen +6 basis points, and gold slipped -1.1%. While the major averages registered modest gains last week, with the S&P 500 index sitting slightly more than 1% away from its August 13th all-time high, there were a few wrinkles to the equity market backdrop worth noting.
NYSE composite weekly breadth was negative for a second straight week. That hasn't happened since March. Additionally, less than 50% of the stocks in the S&P mid- and small-cap indexes are above their 50-day moving averages. Another widely followed breadth metric worth noting is the McClellan Summation Index (a measure of advancing versus declining volume on the NYSE), which has recently begun to falter, with the daily McClellan Oscillator recently turning negative, dragging the longer-term Summation Index down from its recent peaks. A declining NYSI while the headline S&P 500 rises is a textbook bearish divergence. This sudden breadth erosion bears watching.
The hawkish Warsh commentary on Friday weighed heavily on the anti-dollar trade (commodities, precious metals, foreign equities), and Technology stocks were hit hardest, making it the worst-performing S&P 500 sector on the session (-1.3%). Yet five of the seven Mag 7 names finished in the green, supported by company-specific news. The group has shown no net change since mid-May. The dispersion within the group, which accounts for roughly 30% of the S&P 500 Index, has been widening steadily this year, with META and TSLA lagging by a wide margin. The once-tight correlation within the Mag 7 has effectively vanished this year. A declining correlation, from a recent high of 80% to just 10% currently, means stocks are moving independently rather than as a group – as in, being driven more by company-specific factors than macro variables.
While on the topic of the Mag7, I have to include a couple of thoughts on Nvidia’s earnings results from last week. The number that stood out most to me was the FY2028 revenue guidance of nearly $700 billion. I know in today’s world we throw around numbers in the billions and trillions as if they are commonplace. Doing so misrepresents the scale and context of just how large these numbers are:
If you made $1/second it would take you 11.5 days to earn $1 million
If you made $1/second it would take you 31.7 years to earn $1 billion
If you made $1/second it would take you 31,688 years to earn $1 trillion
To put $1 trillion into perspective: if you started making $1 every single second on the day the Roman Empire was founded, you still wouldn’t even be close to one trillion dollars today. Now back to NVDA and contextualizing a $700 billion revenue guide for FY2028. Below is a list of the 20 largest revenue-generating companies in the world over a trailing 12-month period:
Revenue is one thing; earnings, or what the bottom line is from all those revenues, is what really matters. Here is where what Nvidia is doing puts it in a category of its own. Below is the same table of the top 20 revenue generating companies in the world, but this table includes the trailing 12-month earnings and the implied net margin:
Looking at the table above, the companies capable of generating $300B+ in revenue are mostly high-volume, low-margin businesses like retailers (Walmart, Costco), pharmacy distributors (McKesson, Cencora), or legacy energy/auto giants (Sinopec, Volkswagen). Even tech giants like Apple and Amazon operate at roughly 18% to 27%, while Microsoft and Alphabet have some of the highest at just under 40% net margins. As a result, the earnings of these Mega-Cap Tech companies are higher than every other company in the world: Alphabet ($160 bil – excludes $99 bil net gain on equity securities), Amazon ($135 bil), Microsoft ($125 bil), and Apple ($128 bil). NVIDIA, however, has recently been operating with net margins in the 55% to 75% range, which is why it earned $192.88 billion over the prior 12 months on $300 billion in revenue.
If NVIDIA were to actually achieve $690 billion in revenue by FY2028, it would be one of the highest revenue-generating companies in the world. Even more impressive would be if it maintained its current margin profile; it would generate close to $400 billion in net income. To put that into perspective: that single-year profit would be greater than the combined trailing 12-month net income of Alphabet, Microsoft, and Apple.
I don’t point out any of the above as an endorsement to buy, sell, or trade any of the above companies, but rather a feeble attempt at grossly simplifying why Nvidia (at this moment) is the most valuable company (by market cap) in the world at $5.3 trillion. Followed by Apple ($4.6 tril), Alphabet ($4.1 tril), Microsoft ($3.8 tril), and Amazon ($2.8 tril).
Pivoting back to the broader markets, I think a lot of variables are lining up to support a period of chop/consolidation. Earnings and Jackson Hole are behind us, with jobs and CPI reports being the two big pieces of data we’ll get before the Fed’s next meeting on September 16th. After that, we have the mid-term elections, where typically markets are choppy leading up to the election results before resuming the prevailing trend into the year-end push. I’m sure there will be some surprises and noise to risk manage along the way, but I’m finding it difficult to come up with an outlook where any investor's base case shouldn’t be constructive from here into year-end. Volatility is muted, but not excessively low. Bearishness remains stubbornly elevated, and positioning is hardly screaming euphoria. AAII bears have averaged more than 40% over the past three weeks, despite the S&P 500 making fresh records and VIX trading below 16.
Meanwhile, the market continues to elegantly manage keeping the broader indices propped up while rotation under the surface rings out excessive exuberance. Most recently, this rinsing is taking place in the tech sector, which has spent the summer going nowhere, albeit while compressing into an increasingly tight range. Despite all the bullish (AI is going to take over the world – earnings and margins have never been stronger) and bearish (circular financing and negative free cash flow) narratives, the Nasdaq is at mid-May levels. What we have is an index that is increasingly squeezed between the longer-term trend line, the 100-day moving average, and the shorter-term downtrend. The range is tightening, and psychology has compressed with it. When this finally breaks, the move could get violent. I lean in the camp of a bullish breakout, and will be open to changing my mind and adjusting if it turns out differently.
Which brings me to the following post on X by Gavin Baker (a highly respected investor and thought leader in the Tech space) on data centers that I think is worth sharing. While I agree with much of what he says, that isn’t why I'm sharing his post. We live in a world where information has become both a tool and a weapon used for propaganda, education, and misdirection. I appreciate when friends, clients, and other professionals share thoughtful, accurate, and informative content no matter its origin or who it's from. Whether we like it or not, we are all going to live through this AI revolution, and there will be things we support or oppose. No matter what, we are all better off being more educated and informed on the evolution along the way. The current hot-button issue is data centers (call it this era’s fracking rigs). A month from now, it will likely be something different. My only advice is to be open-minded in understanding all sides of the debate while forming your opinion.
Let’s end with a couple of thoughts on gold, which had a rough day on Friday with the hawkish messaging from Warsh, but don’t lose sight of the bigger picture of what’s really been moving gold over the past four years – central bank purchases. Sure, gold will get caught up in short-term squiggles based on moves in the U.S. dollar or interest rates, but make no mistake that until we see a pivot to more fiscally responsible policy globally or central banks walking back from their relentless pursuit of diversifying their foreign exchange reserves, gold will remain in a structural bull market. Central bank purchases now absorb 32% of annual mine production with recent surveys showing no signs of this letting up.
Fully 83% of central banks expect to increase gold holdings over the next five years, and 74% expect the dollar's share of their reserves to decline. The World Gold Council's 2025 survey found zero central banks indicating reduction plans, and the June 2026 survey found a record 45% planning increases in the coming year. That's not positioning. It's a stated multi-year policy commitment across a buyer base spanning more than forty central banks.
Emerging market central banks, in particular. For example, Brazil divested $61 billion in Treasuries last year while doubling its gold reserves, making gold the second-largest component of its portfolio. China extended its buying streak to 21 consecutive months. Poland added 102 tons in 2025; India bought 76 tons in 2024-2025. The causal mechanism is specific and irreversible: it was the freezing of roughly $300 billion in Russian reserves in 2022, which demonstrated that foreign currency holdings are a potential liability rather than a pure asset for any central bank in a jurisdiction subject to sanctions. The custody dimension is live too – France moved 129 tons out of the New York Fed between July 2025 and January 2026, and the 2026 survey shows 9% of central banks increasing domestic storage and 10% diversifying overseas vaulting locations.
The official-sector rotation into gold is the most persuasive single piece of evidence for the structural dollar bearish narrative. Since it is policy-driven and running through a buyer who is price insensitive but rather strategic in nature, the gold bull market and dollar bear market will likely be joined at the hip for years to come.
There are two market-moving developments on the docket for this week. The first is the quarterly results from semiconductor company Broadcom on Wednesday, which will test the market’s tech-centric resolve (the share price is off -23% from the June peak); the second is the nonfarm payroll report for August that comes out on Friday. The consensus is +55k on headline payrolls (which would match the six-month average), a trend-line +0.3% reading on average hourly earnings, a flat workweek at 34.3 hours, and the jobless rate remaining at 4.1%. If we get another weak report, we’ll see if we’re still operating in a ‘bad news is good news’ backdrop.
I won’t be penning a missive next week as I’m going to take the long weekend as an opportunity to go visit Mom and Dad. Hope you all have a great Labor Day holiday. Safe travels and best wishes.
The articles and opinions in "Capital Market Musings and Commentary" are for general information only, and not intended to provide specific investment advice. Performance, dividends and other figures have been obtained from sources believed reliable but have not been audited and cannot be guaranteed. Past performance does not ensure future results. Investing inherently contains risk including loss of principle. Advisory services offered through Casilio Leitch Investments, an SEC registered investment advisor. Copyright © 2026 Casilio Leitch Investments. All Rights Reserved.