I want to be clear that I am not bearish META at all, I am simply taking a few chips off the table while the market is handing me a nice exit.

I’m also pleased to announce that the Rigatoni Capital portfolio hit all-time-highs today, which may be due to some investors sniffing out that the Federal Reserve holding rates in September is back on the table. This is after Fed Governor Chris Waller signaled this morning that he “would be inclined to hold rates steady in September if the incoming inflation data keeps showing progress toward two percent.” It’s been my opinion all along, even after Warsh’s “hawkish” Jackson Hole speech, that the Fed would still hold during the next FOMC meeting September 16th. However, this wouldn’t change my portfolio positioning either way.

The Rigatoni Capital Podcast is now on YouTube, Spotify, Apple Podcasts, Amazon Music, and iHeart Radio.

Broadcom Looks Reasonably Valued after an Earnings Selloff

Broadcom reported fiscal third-quarter results after the close on Wednesday, and the stock sold off slightly due to a disappointing next quarter guide and some questions over gross margin guidance. I would not call the stock outright “cheap” today, but I would argue it’s reasonably valued here.

The quarter itself was again excellent, but investors still seem to want more. Revenue came in at $29.6 billion, up 86 percent year over year, with adjusted earnings of $3.32 per share. The AI semiconductor line grew 221 percent year over year to $16.7 billion, up 54 percent sequentially. The reason the stock fell anyway is that the Q4 revenue guide of $34.8 billion landed just under the $35.03 billion consensus.

There was, however, plenty on the earnings call for the bulls to hang onto, which is why some on the Street quickly called the selloff an overreaction. CEO Hock Tan laid out a target of $115 billion in AI semiconductor revenue for fiscal 2027 and $230 billion for fiscal 2028, with earnings power he expects to exceed $30 per share by fiscal 2028. Those are enormous numbers, and it’s worth being clear that they are AI semiconductor revenue specifically rather than total revenue, and that management is framing them as a trajectory it has secured supply to meet rather than a formal quarterly guide.

Now on the gross margin questions that were asked on the earnings call. On a non-GAAP basis, gross margin was 75 percent, down about 210 basis points from the prior quarter. It was explained that custom AI silicon carries lower gross margins than Broadcom's software and legacy chip business, so as AI becomes a bigger slice of revenue, it drags the blended gross margin lower. Operating margin actually rose to roughly 68 percent and operating income grew 92 percent year over year, because revenue is growing so much faster than the costs required to support it.

I will also be honest about how Broadcom fits in my own portfolio. I already carry large exposure to Nvidia, so a smaller sleeve of Broadcom is in a way a step toward diversifying away from that single-name concentration while staying in the AI semiconductor theme I want to own. It is my view that both can gain share in AI infrastructure over the coming years rather than one winning at the other's expense, and both are effectively saying as much with their guidance. I think context is everything when sharing the stocks I decide to own.

To be clear, I am not saying Broadcom is a buy here, and I am definitely not saying Meta is a sell here. I am just walking you through my thinking and, as always, reporting my portfolio changes to you all as soon as they are made.

Thank you all for reading and don’t forget to subscribe if you enjoyed this post. I leave you below with my top 20 holdings, on a day that I hit all-time-highs in a portfolio that I have built in a decade.

Top 20 Holdings as of 9-3-26

In Case You Missed:

References:

• Broadcom Q3 FY2026 earnings release, September 2, 2026 (Broadcom Inc. Form 8-K, SEC)

• Broadcom Q3 FY2026 earnings call transcript, September 2, 2026 (Investing.com)

• Meta stock reaction to Muse Spark 1.3, September 3, 2026 (The Motley Fool; Yahoo Finance)

Disclaimer: This blog is for informational purposes only and does not constitute financial advice. All opinions are my own, and I am not a financial advisor. The information provided reflects my personal views and is intended to encourage discussion and thought among readers. Investments involve risk, including the loss of principal, and past performance is not indicative of future results. Always conduct your own research or consult with a qualified professional before making any financial decisions.

The Context Explained to each Model: For one, this is not my only portfolio, but it is my largest one. It is a sleeve of my net worth, but not my entire net worth. For example, I own a 401k that is basically just a target-date fund, along with other index funds and fixed-income vehicles like bonds. This portfolio has taken a long-term buy-and-hold approach, with the goal of continuing to compound into the future. The goal of the Rigatoni Capital portfolio is to achieve financial freedom.

What came back was close to unanimous, and the agreement itself is the story. When four models look at the same portfolio in isolation, even with the context of what I am trying to achieve, they give the same criticisms. Every one of them told me that I was too concentrated and needed to spread out my risk, which is a polite way of saying I should sand my portfolio down until it looks more like a total market index fund. In my opinion, all four of them were wrong.

Here is the full portfolio as it stands today: This is the full book, all 37 positions, with the weights and total returns as of 8-31-26. (Also my top 20 holdings are at the bottom of the post - in case that is easier to read).

The Rigatoni Capital Podcast is now on YouTube, Spotify, Apple Podcasts, Amazon Music, and iHeart Radio.

Where all four agreed

The headline was that I own way fewer real bets than it looks like. On paper I hold 37 positions, but once you look through the funds, the independent bets are far fewer.

QQQ: is my largest holding at close to 20 percent of the account, and I then separately own many of its biggest constituents, so my true exposure to a handful of mega-cap tech names runs well above what the long list of tickers suggests.

Nvidia: Every model landed on the same major flag. Once you combine QQQ, VTI, SMH, TQQQ, and my direct shares, my effective Nvidia exposure sits at around 20% of the portfolio.

Semis: All four flagged the semiconductor stacking, since I own Nvidia directly alongside Broadcom, Lam Research, the SMH fund, and another heavy dose of chips buried inside QQQ.

Bitcoin: They flagged my bitcoin exposure showing up twice, through the iShares Bitcoin Trust and again through Strategy, the leveraged bitcoin holding company, and noted that I hold more bitcoin outside this account as well.

Small Positions: They flagged the long tail of positions under one percent that are too small to move the portfolio.

TQQQ: And no surprise, every one of them singled out the three-times leveraged QQQ fund as inconsistent with a long-term buy-and-hold approach.

On the facts, they were correct. But my push back is that they are still missing the broader context or purpose of this portfolio.

My push back

The portfolio I started back in 2016 was never built to win a diversification contest. If broad diversification and being "safe" were the goal, it would therefore be in a broad index fund, which I do agree can achieve excellent returns. This is exactly why I have lately been accumulating the Nasdaq 100, QQQ, on most pullbacks. 10-plus years later, I have accepted that picking stocks is extremely hard, and that any total return I achieved with that style was a combination of luck and being willing to take more risk at a younger age.

The goal of the portfolio is to lean into decade-long market themes (which I often write about here on Substack), to buy great companies below their intrinsic value, and to hold on to them even when they are trading above their intrinsic value. More than that, the goal is to build a compounding machine for the future. I am not looking to beat any benchmark today, this year, or next year. I want the portfolio to beat the benchmark over a long period of time. This strategy comes with something that looks like a risky or over-concentrated portfolio. It will look like I am being sloppy, with many overlaps giving me exposure to the same mega-cap tech names individually and in ETF wrappers. It will also look like I am nuts for holding Nvidia at such a large weight while it is up 48 times.

But this is exactly what I envisioned the portfolio looking like 10 years ago. To let my winners run, to not get scared out of them when they had a big run as long as I still had conviction, and to let winners like Nvidia earn a chance at becoming a large holding.

Again, all four AI models completely overlooked the broader context of this aggressive sleeve, what I think of as my f*ck-you money, which sits deliberately apart from my retirement accounts and target-date funds, all of which, by the way, have a place in your investing journey. It is not my fault that this once-aggressive portfolio has grown into such a large percentage of my net worth. For the last 10 years this portfolio has existed so that I can take concentrated, high-conviction risk on purpose, what some would call a "mad-money" portfolio.

As I often remind many friends and Substack readers, I did take plenty of profit in Nvidia back in 2024. This is not a position that simply got out of hand or that I cannot control. I carry a high-to-medium conviction view that Nvidia grows into a ten-trillion-dollar company in the coming years.

I have also always envisioned that a company like Lam Research or Caterpillar could become a large holding in the portfolio. That said, I admit it felt like time to trim some profits in Lam Research and Palantir this month. I have also trimmed TQQQ, which I still consider a swing trade, and I sold my entire Microsoft position in November 2025, a top holding at the time with a total return of +450 percent. I have sold names like Broadridge Financial Solutions, which I had held since 2017, and Brown and Brown insurance, and I have swing-traded both Meta and Netflix more than once over this 10-year period. So the portfolio can be active when I feel it is necessary.

Another important note of context here. I am not an institutional hedge fund investor, so I have no one to answer to. So although hedging my positions is something I think about constantly, it has not been something I have actually done to answer for an extremely volatile portfolio. It would have ideally helped me sleep a bit better in a year like 2022, when I was down much more than the broader market.

There is also the sense that I do wish I had a larger holding in an index fund like the Nasdaq 100, and that I had started building that position back in 2016, or much sooner. Now I am catching up as I get older, having come to realize the power of a market-cap-weighted index of the best mega-cap tech stocks.

And of course there is international exposure, of which I have almost none, except for my Freedom 100 Emerging Markets ETF. To be fair, I own a nice chunk of this FRDM ETF in another portfolio that I do not disclose in my writing.

And lastly, there are the old-economy names like Waste Management or Chubb that are not bigger positions in the portfolio. I do wish they were, but it is not my fault they have not grown into large positions on their own. This is something I feel I have little control over, though they are always on my watchlist, along with other names that are not tech-driven. In my opinion, high-quality companies like these only go on sale once or twice in an investing journey. You do not get many shots to own this type of company at a great value.

What I am changing

I am not changing shit, which is the point of a long-term investor built on companies I chose years ago. I am by no means married to my positions, and I am happy to let go when the fundamentals of the story change. I am not claiming I can beat the S&P benchmark over the long run. That is nearly impossible. But this is what the portfolio has become, a monster that must be managed and reviewed every so often.

So if you learn anything from this post, it is this: please, PLEASE do not lean on some AI model to critique your portfolio and truly understand the context of what your portfolio is to you and what you are trying to achieve, which for me is financial freedom one day. This is the journey I have chosen. I am not sure whether it will be successful, but this is what it is, for today.

Thank you for reading, and please subscribe for more content on long-term investing.

Top 20 Holdings as of 8-31-26:

Not sure if this is easier to read, versus the graphic above, but here you go. I usually always post my top 20 holdings in each Substack post I send out to subscribers.

In Case You Missed:

Disclaimer: This blog is for informational purposes only and does not constitute financial advice. All opinions are my own, and I am not a financial advisor. The information provided reflects my personal views and is intended to encourage discussion and thought among readers. Investments involve risk, including the loss of principal, and past performance is not indicative of future results. Always conduct your own research or consult with a qualified professional before making any financial decisions.

Kevin writes the Mytwocent$ newsletter on Substack, and if you have not had a chance to subscribe, here is your monthly reminder. While we spent a good amount of time on hard metals today, it's my opinion that the theme of this podcast highlights how Bitcoin sits at the cross-section of both scarcity and networks. These two themes (scarcity and networks) are an investing theme that we both feel will become even more popular in the coming months and years.

Mytwocent$

Casual observations on the passing scene

By Bagholder

This podcast was recorded on Sunday, August 30th.

As a reminder, The Rigatoni Capital Podcast is now on YouTube, Spotify, Apple Podcasts, Amazon Music, and iHeart Radio.

Dr. Copper Is Talking Again

Kevin's starting point was copper, the metal traders' nickname Dr. Copper because it tends to move before the rest of the complex. By his tally copper is up around 24% off its March low. Copper is trading near $6.55/lb as I write this up and is up roughly 45% over the past year, so the move is significant. His read is that when copper leads, gold and silver follow, and he pointed to the mining stocks confirming it. The GDX gold miners and the SILJ silver miners have both run hard over the last month, which matters to him because the miners lever the underlying metal. Money that wants exposure tends to reach for the levered version.

Silver is where he sees the cleanest setup next. He walked through a long list of commodities that made highs in 1980 and have since risen roughly five to six times, which lines up with cumulative inflation over that stretch. Silver is the odd one out. It topped near $50 in 1980 and trades around $66 today, so it has barely kept pace while copper, gold, iron, and the rest multiplied. Pair that with a structural deficit, where the world uses more silver than it mines each year, and his conclusion is that the gap closes through a much higher price. Of course, this is in the context of silver spiking close to $122 an ounce back in January, before selling off hard. So the value being described here is relative to an extreme pullback, and the entry price is always a very important aspect of a commodity like silver.

Inside the System, Outside the System

Kevin has always split assets into those inside the financial system and those outside of it. His portfolio leans toward the outside, and his case rests on debasement. He argued that the bond market is far too large for the Treasury to push around, and recent action supports him.

Treasury Secretary Scott Bessent surprised markets by doubling the ceiling on long-dated buybacks from $2 billion to at least $4 billion per operation, running from September 9 through November 4, and hinted he could tap the roughly $950 billion sitting in the Treasury General Account. Yields dipped on the news and then climbed right back, with the 30-year near levels last seen before 2008.

Kevin’s point is that this is posturing against a market that will price rates where it wants. He extended the same logic to Fed Chair Kevin Warsh, whose hawkish Jackson Hole speech knocked gold, silver, and Bitcoin lower on Friday. Speeches are words, he said, and this is a deeds world, and the deed he expects is more money printing to cover the interest bill.

Products Versus Networks

The most important stretch of our conversation was Kevin’s distinction between companies that invent a product and companies that own a network. He ran through Kodak, Ford, Xerox, Cisco, and GE, all of which brought a breakthrough to market and then watched competitors grind their margins down over the years. Against that he set JPMorgan, Amazon, Netflix, Standard Oil, and AT&T, none of which invented the thing they sell. They control the network the thing runs on, whether that is money, retail, distribution, or connection, and networks prove far more durable than products.

This framing is one of the reasons that Kevin won’t buy Nvidia today. The quarter was every bit as strong as advertised, with revenue of $96.2 billion up 106% from a year ago and gross margins at 75%, and the company guided next quarter above expectations and pointed to roughly 70% revenue growth in FY2028. To Kevin it looks like Cisco in 1999, a product company with extraordinary margins that invite the competition which eventually shows up. He is not calling the top, and he put the timeline at three to five years, but he wants asymmetry and does not see a path where Nvidia rises tenfold from a $5.2 trillion base.

The Intersection

Scarcity plus network is Kevin’s definition of the ideal asset, and Bitcoin is the one thing he sees sitting at that crossing. It is capped in supply and it is a strong network, and the asset is small relative to what could flow into it. He sizes Bitcoin near $1.5 trillion against roughly $34 trillion in gold and a bond market north of $350 trillion. If the long end keeps breaking, some of that fixed-income money goes looking for a home outside the system, and he thinks a scarce network asset is one of the few places it can land. His portfolio matches his mouth, with Bitcoin and silver as the core.

A Contrarian Take on the next FOMC Meeting

Consensus after Warsh's speech shifted toward a rate hike, with prediction markets pricing the odds of a September move near two thirds. Kevin thinks a cut is the real possibility nobody is positioned for, and his reasoning is political rather than economic. Warsh is President Trump's handpicked chair, the midterms are close, and a booming market helps the party in power, so he would not be shocked to see a cut at the September or October meeting.

I personally feel a hold is more likely than a raise, but a cut? That wasn’t even on my radar. I’m sure a lot of “experts” would laugh at this as a real possibility, which is why I am taking it more seriously now.

Thank you all for reading, subscribing and listening to the podcast. And as a second reminder, you can find this podcast with and other daily Rigatoni Capital podcasts on all major platforms like YouTube, Spotify, and Apple.

In Case You Missed:

Disclaimer: This blog is for informational purposes only and does not constitute financial advice. All opinions are my own, and I am not a financial advisor. The information provided reflects my personal views and is intended to encourage discussion and thought among readers. Investments involve risk, including the loss of principal, and past performance is not indicative of future results. Always conduct your own research or consult with a qualified professional before making any financial decisions.

The headline was a clean revenue beat paired with a strong third-quarter guide, which is exactly what a long-term investor like myself was hoping for. Attention moved quickly to a slightly lighter gross margin outlook, so the stock drifted lower through the start of the earnings call. Then the stock started to spike during the call itself, on some exciting and unexpected news along with strong comments:

• Nvidia’s stock spiked during the earnings call primarily because CFO Colette Kress revealed that fiscal 2028 income is expected to grow by 70%, completely blowing past Wall Street’s expected 45% growth projection.

• The 2 Million GPU Commitment: AWS is committing to 2 million additional Nvidia GPUs, on top of the 1 million it originally pledged at GTC 2026, bringing its total commitment to roughly 3 million. The new tranche covers the 2027–2028 Blackwell Ultra, Rubin, and Rubin Ultra product cycles, giving both companies unusually deep multi-year supply visibility.

• U.S. Government AI Factories: As part of this deal, 100,000 of those GPUs are specifically earmarked to support U.S. government AI factories and sovereign infrastructure, driving massive public sector demand.

• Next-Gen Chip Integration: AWS is also buying Nvidia’s brand-new Vera CPUs and integrating its NVLink Fusion technology natively into custom server racks.

• Next-Quarter Blowout Guidance: Nvidia guided for $108 billion in revenue for Q3, soaring past consensus figures and proving that sequential momentum is not plateauing.

So that is the exciting news. Not much to discuss on the circular funding front, except that the company confirmed its Days Sales Outstanding (DSO) surged to 60 days, climbing sharply from 45 days in the previous quarter. Concurrently, the company's accounts receivable ballooned by 55% sequentially, growing from $40.7 billion to $63.1 billion. However, the company specified that these looser credit terms were given to highly creditworthy "investment-grade" customers (primarily the major hyperscalers).

On the numbers, Nvidia reported revenue of $96.2 billion, up 18 percent from the prior quarter and 106 percent from a year ago. Data center did the heavy lifting again at $89.0 billion, up 117 percent year over year. Gross margin held at 75.0 percent on both a GAAP and non-GAAP basis, however, management only guided third quarter gross margin near 74 percent. GAAP operating income climbed to $63.7 billion, up 124 percent from last year. The company returned roughly $26 billion to shareholders through buybacks and dividends during the quarter and still has about $99 billion left under its current repurchase authorization.

By no surprise, Nvidia remains a hold in my portfolio. But don’t worry, I have some real breaking news for you all. Earlier this morning I started a new position in Meta.

Meta Suprise Buy

The moment the news crossed that the company had reached a proposed settlement in the multistate child safety case, I took a stab at the stock. This takes a fairly large headwind off the board. I was roughly flat on the position by the end of the day, and it sits as my 21st largest position, so it does not show up in the top holdings below.

I first bought Meta in early April 2025, right around Liberation Day and near where the stock was bottoming. I sold the position a few months ago at a small gain, so this is my second go-around. As you know, I'm typically a long-term buy and hold investor and rarely swing trade stocks.

Meta did reach a proposed settlement today on the claim that Facebook and Instagram were designed to be addictive to minors and hid the risks. The company agreed to pay the states up to about $17.1 billion over roughly 10 years. Meta did not admit to wrongdoing, and the federal judge overseeing the case still has to approve the deal.

Without getting into the details of what Meta's settlement means for its product changes in the United States, it probably makes a small dent in teen engagement, especially for the Meta advertising engine.

As I write this, I do not know whether this becomes a very long-term investment or a swing trade. I would not recommend trading around a position this way for anyone, but the stake is only about 1 percent of my total portfolio, small enough that I can sit with it and let the decision come to me over the coming weeks.

Obviously, there is a bigger headwind for Meta, and that's their earnings miss last quarter along with a lower Q3 guide, all while they are spending like a drunken sailor on AI data center infrastructure. There must be a clear path to a healthy return on capital. That is still the main hurdle that Meta must overcome.

Thank you for reading, and don’t forget to subscribe for more content on long-term investing.

Quick reminder: The Rigatoni Capital Podcast is now on YouTube, Spotify, Apple Podcasts, Amazon Music, and iHeart Radio.

Top Holdings as of 8-26-26

In Case You Missed:

References:

NVIDIA, Q2 Fiscal 2027 financial results and CFO commentary: https://nvidianews.nvidia.com/
Wccftech, Amazon commits to 3 million Nvidia GPUs in AWS: https://wccftech.com/amazon-commits-to-3-million-nvidia-gpus-in-aws-tripling-a-deal-that-started-at-just-one-million/
Kiplinger, Nvidia earnings live coverage, August 26, 2026: https://www.kiplinger.com/investing/live/nvidia-earnings-live-updates-and-commentary-august-2026
NPR, Meta and states agree to child safety settlement: https://www.npr.org/2026/08/26/nx-s1-5944781/meta-settlement-child-safety-lawsuit
CNN Business, Meta settles state child harm claims: https://www.cnn.com/2026/08/26/tech/meta-states-settle-trial-children
Kentucky Lantern, settlement terms and safety changes: https://kentuckylantern.com/2026/08/26/meta-settles-with-kentucky-other-states-including-changes-for-kids-safety/
Meta Newsroom, statement on the agreement: https://about.fb.com/news/2026/08/agreement-with-state-attorneys-general-supporting-teens/

Disclaimer: This blog is for informational purposes only and does not constitute financial advice. All opinions are my own, and I am not a financial advisor. The information provided reflects my personal views and is intended to encourage discussion and thought among readers. Investments involve risk, including the loss of principal, and past performance is not indicative of future results. Always conduct your own research or consult with a qualified professional before making any financial decisions.

As a long-term investor myself in Netflix and AppLovin, I can tell you that Simeon is my go-to expert on analyzing these types of businesses. Please don't forget to check out his newsletter below:

Accrued Interest

High-conviction stock research from a media executive with a +3,400% track record (2016-2024). Institutional-grade analysis of US media stocks, value investing, and market trends.

By Accrued Interest (Simeon M.)

Netflix (NFLX) and the Watch-Hour Panic

Simeon's core argument is that the market burns too much energy on engagement and watch-hour data that amounts to noise. Netflix reports total global streaming hours north of 90 billion, and that figure has grown at a steady one to three percent a year over the last five years, so the mid-quarter panic about engagement never matched what the numbers were doing.

What he wants investors tracking instead is revenue and revenue per watch hour, where Netflix has separated itself from the field. Subscription video on demand has a natural ceiling, and the next phase of growth comes from podcasts, live sports, and short-form content, which is where the company runs straight into YouTube.

Content spend has grown far slower than revenue, which means Netflix keeps pulling more out of the catalog it already owns. Return on equity climbed from roughly 25 percent in 2022 to 43 percent in 2025, and Netflix managed that without adding leverage, largely through margin expansion.

Operating margins now sit in the mid-30s and look headed toward 40, a level no other global streamer approaches, since most of them are break even at best. Netflix’s excess free cash flow funds the buybacks pushing earnings per share higher, one more line the competition cannot match. This is the setup behind Pershing Square’s Bill Ackman rebuilding his position after exiting in 2022 at a loss of roughly 400 million dollars. Ackman likes the stock now in the low 20s on forward earnings, down from about 41 times, which Simeon called one of the lowest valuations since the company went public.

For more info, check out Simeon’s latest thoughts on Netflix here:

Accrued Interest

How Wall Street Misread Netflix ($NFLX): My Talk with Tim Rowe on the State of Streaming Podcast

I had a great time joining Tim Rowe on the State of Streaming podcast for a wide-ranging conversation about the state of the media business as we head into the back half of 2026. If you love talking about media as much as I do, then State of Streaming…

Read more

13 days ago · 7 likes · Accrued Interest (Simeon M.)

Quick Reminder, The Rigatoni Capital Podcast is now on YouTube, Spotify, Apple Podcasts, Amazon Music, and iHeart Radio.

Check out my podcast with on your platform of choice, or tune in for a daily markets podcast where I go over the headlines and determine what matters for long-term investors.

AppLovin (APP) and the Cost of Getting Comfortable

Simeon and I have another thing in common. We are both bulls on AppLovin as well as Netflix, though we left the latest earnings call a bit disappointed. What rattled the market was the break in a pattern investors had come to count on, the steady quarter-over-quarter targeting gains from AppLovin's AI ad engine. As Simeon put it, "investors, myself included, got too complacent." Once a single quarter of model improvement failed to show up, the market began discounting future improvements too, and management added to the worry by taking down the growth rate and flagging higher training and cost spend ahead.

AppLovin is trading around 15x on estimated 2027 GAAP earnings, which shocked me. Simeon read it as the market saying, “We don’t believe your ‘27 earnings at all.” He treats a low multiple as information rather than a bargain, telling me that “oftentimes a low multiple is a sign that the market is assessing a risk, and it’s up to you to decide if you wanna take that risk or not.”

The e-commerce ramp that carried much of the bull case is running slower than the excitement implied, with Meta and Google still the dominant players in that space. “A lot of the e-commerce growth is coming through higher spend from the participants who are already there,” he said, and bringing on the next wave of advertisers takes years rather than quarters. Add a resurgent Unity competing again on the gaming side, plus the reality that e-commerce carries lower margins than gaming, and the compressed multiple starts to look reasonable rather than mistaken.

For Simeon’s latest newsletter on AppLovin, see here:

Accrued Interest

How Much Lower Can AppLovin Fall?

Background: AppLovin has fallen about 11% ($310 vs. $346.80) since my Q2 update on August 10th, AppLovin Q2-26: I Got This One Wrong. It was one of my most read articles ever, and I explained why the bull run had ended…

Read more

12 days ago · 3 likes · 1 comment · Accrued Interest (Simeon M.)

Duolingo (DUOL)

Simeon moved Duolingo from underperform to market perform, less because his monetization concern eased and more because daily active user growth reaccelerated. He noted the company converts only about 10 percent of free users to paid and leans heavily on subscription revenue, which he put at over 90 percent, with little advertising. Management has framed 2026 as a reset year favoring growth over monetization. His warning still stands: “you always need to monetize eventually.”

For more, check out Simeon’s latest on Duolingo:

Accrued Interest

Duolingo Q2-26: Growing Users, Not Money

Accrued Interest TLDR: Following Duolingo’s second quarter earnings last week, most of my issues with this company continue to be proven correct, and I stand by the underperform call I made throughout the first half of the year. DUOL is down 24% year to date against a rising market…

Read more

19 days ago · 8 likes · Accrued Interest (Simeon M.)

The Warner Bros Puzzle

On the 13F filings, Third Point's Daniel Loeb built a Warner Bros position worth around 533 million dollars, now his largest holding, with Greenlight's David Einhorn and reportedly George Soros buying alongside him.

The bet is likely a merger arbitrage play on the Paramount deal. Simeon has been negative on that merger for more than a year, and his view is that Warner Bros keeps climbing because the odds of completion are rising as Paramount looks likely to concede ground to close it. His caution points at the buyer. Even in a win, he argued, Paramount takes on heavy debt to acquire declining assets while “Netflix and YouTube are ready for war.”

The AI Capex Debate

I closed on something a bit different with Simeon, stepping outside the media space to ask him about the AI bubble narrative. Put simply, he is not bearish on AI and tech stocks. The four largest hyperscalers, Google, Meta, Amazon, and Microsoft, entered this spending cycle with under-leveraged balance sheets and real free cash flow, and he sees clear positive returns on the investment. He called them "four of the strongest companies maybe in the history of capitalism," and his take on Nvidia was that demand does not hinge on any single model lab, since governments and sovereign buyers will absorb whatever OpenAI or Anthropic leave on the table.

The part most relevant to his coverage is that the payoff already shows up where people overlook it. As he put it, “The AI gains are not just going to AI tools. The AI gains are going to ad targeting.” In his framing, that ad targeting benefit is exactly why the bear narrative keeps Meta cheap, and why he still holds it as the cheapest name he follows. It brings the conversation full circle to the Ackman playbook of buying strong blue chips when they slip under 20 times forward earnings.

Thank you again for joining the podcast, it’s always a great time.

And thank you to everyone that has spent time reading, listening, and watching. If you enjoyed this kind of content, don’t forget to subscribe.

In Case You Missed:

Disclaimer: This blog is for informational purposes only and does not constitute financial advice. All opinions are my own, and I am not a financial advisor. The information provided reflects my personal views and is intended to encourage discussion and thought among readers. Investments involve risk, including the loss of principal, and past performance is not indicative of future results. Always conduct your own research or consult with a qualified professional before making any financial decisions.

The idea for this post came to me in the morning while I was recording a podcast. I was moving through the usual pre-market headlines, and somewhere in the middle of it all I said something that I think would make for a cool post.

The best part of being a long-term investor is that the learning never stops. Yes, compounding your wealth is the goal, but with that comes a compounding of knowledge.

Knowledge compounds much like the portfolio; it’s just not a number on a screen.

Warren Buffett has put this about as plainly as anyone. Asked once how a student should prepare for an investing career, he reached for a stack of reports and trade publications and told the room to read several hundred pages a day, as his investment manager Todd Combs has recounted. This is because knowledge accrues the way compound interest does. It's important to spend a little time each day trying to understand something you don't already know: learn a new business, learn a new industry, etc.

In the early years of my investing journey, it was like learning from scratch. It took years to understand many of the intricacies of how the market worked without coming from inside the industry. That said, I am fairly confident this raw point of view carried plenty of advantages. When I pick up an unfamiliar investing topic now, I at least have more than a decade of context, and the learning clicks into place far faster than it did when I started out.

There is a version of this in sports, and golf is the closest fit. Most athletes peak in their twenties, while a golfer can win at an age nobody else in sports gets to. Investing carries that further, because the ceiling here is not physical. It is how long you let your learning compound, so your best years can come later in life than in almost any other field.

The late Charlie Munger believed you should diversify your knowledge by learning the big ideas that live outside your core competency or industry. He read constantly across subjects that had nothing to do with markets, and he treated that breadth as an edge rather than a hobby. He liked to say that in his whole life he had never met a wise person, across any broad subject, who did not read all the time.

Munger noticed that the people who rise over a long career are frequently not the smartest or the hardest working in the room, and what they tend to share is a habit of going to bed a little wiser than they woke up.

The Rigatoni Capital Podcast is now on YouTube, Spotify, Apple Podcasts, Amazon Music, and iHeart Radio.

Artificial Intelligence and Abundance

When ChatGPT first arrived, back in late 2022, I lost sleep for a few nights over the concept. I remember my mind would not turn off, thinking of all the ways this tech could help me learn even more about businesses and markets than I could the week before. I am not sure others reacted the same way, but for me it felt like a door opening onto a kind of abundance of knowledge and wealth creation.

Regardless of how you feel about AI, it still allows me to take an industry, business, or investing topic I know little about and get to a working understanding in an afternoon that might have taken me weeks.

Of course, information has never been more available, but the thing that stays scarce is the desire to use it, or even the desire to learn in the first place. Buffett made a version of this point long ago, noting that everyone can read what he reads, the playing field is level, and most people will not do the work. That was true when the reading meant annual reports and trade magazines, and it holds up now that the reading can mean asking a model to walk you through a company you have never studied.

I have another example here, and I know he may be a controversial figure in the investing world, but hear me out. Martin Shkreli, love or hate him, spends some of his time trying to mentor younger investors. A few weeks ago on a livestream, he mentioned that many people (including himself) are day-trading a name like SanDisk without any grasp of where it fits in the wider AI infrastructure story. His answer was to slow down and learn the basics, and he pointed people to a book he loves, one of many he read through during his years in prison, ‘Code: The Hidden Language of Computer Hardware and Software’ by Charles Petzold. Shkreli went on to say, "The book starts with two kids and a flashlight, and from there it builds up and teaches you how computers work under the hood. Even if you already know computers well, you will come away having learned something new."

AI Built a New ‘Situation Room’ for the Podcast:

A little off subject, but I can give you an example of how AI is pretty incredible for someone like me who is not tech savvy. This past weekend I built a set of live overlays for the podcast broadcast: a scrolling live market ticker, a news wire pulling headlines in real time, a custom economic calendar, and a global situation monitor that maps live conflicts, storms, and other major events. The news wire is something I built myself, so it is not a professional, instant newswire. It does, however, accurately pull live headlines from my Feedly, run them through Cloudflare, and bring them back to the live headline ticker.

The main objective was just to make the podcast a bit more aesthetically pleasing, but how well they all work has been impressing me.

Of course, what might have taken a developer an afternoon took me an entire weekend. But the reason I mention this is that it would never have been possible for me to do a couple of years ago. Many times the code it wrote had to be fixed, and it took a shitload of patience, hitting many token limits along the way.

My entire point here is that the year of agentic AI is no joke. Anyone who is only using AI as a chatbot to help them write an email is nowhere near scratching the surface. It does require imagination and the desire to learn, but this felt like a good place to bring the post to a close.

Thank you all for reading, and please subscribe if you enjoy this kind of content on long-term investing!

In Case You Missed:

References:

Warren Buffett on reading several hundred pages a day and knowledge compounding like interest (Todd Combs recollection): Quote Investigator, quoteinvestigator.com/2018/10/20/pages/ and CNBC, cnbc.com/2018/03/27/warren-buffetts-key-tip-for-success-read-500-pages-a-day.html

Charlie Munger on learning machines and growing a little wiser each day (USC remarks, collected in Poor Charlie’s Almanack): fs.blog/learning-machine/

Charlie Munger on never knowing a wise person who did not read constantly (Poor Charlie’s Almanack): goodreads.com/author/quotes/236437.Charles_T_Munger

Disclaimer: This blog is for informational purposes only and does not constitute financial advice. All opinions are my own, and I am not a financial advisor. The information provided reflects my personal views and is intended to encourage discussion and thought among readers. Investments involve risk, including the loss of principal, and past performance is not indicative of future results. Always conduct your own research or consult with a qualified professional before making any financial decisions.

I made a rare set of moves in the Rigatoni Capital portfolio over the last two trading days. As promised, I always journal to my subscribers first before I tell anyone.
Before we get started, here are my updated top 20 holdings as of end-of-day, Monday, August 17th.

Let's start with the first of three names I trimmed, since this was on Friday, August 14th. The 3x leveraged QQQ ETF, ticker TQQQ, had a nice bounce after I entered into a position on July 29th, which I discussed with subscribers in the post below. Yes, this was an incredible V-shaped recovery on the day that we had leverage blow-ups in South Korea, to Situational Awareness, to a disconcerting first Kevin Warsh presser following the Fed's first rate decision under his chairmanship.

"Of course, on July 29th, I explained to you that my bigger purchase that day was adding to my Nasdaq 100 QQQ position, and therefore I was 'sprinkling' a little TQQQ that day for some spice. Since that entry, I was up 34% on TQQQ, so I took my entire profits off the table, while still leaving my initial position, which is now my 21st position, roughly 1.06% of the portfolio.

This morning, I was a bit more trim-happy. I suppose the good feeling of locking in those profits carried over the weekend. So I decided to take my current profits in Palantir, where I was up around 10% this morning. I also trimmed a bit of Lam Research, which has been long overdue. Both of these remain large core holdings for me and sit comfortably in my top 20. Nothing about these trims changes the fact that I want to own these businesses for the long term.

If you are wondering why I am suddenly trimming profits, which is rare for me, the real answer is that I was not reacting emotionally. I wanted a certain amount of cash on the sideline while stocks have been ripping over the last few weeks since this V-shaped recovery on July 29th. The summer dog days will be coming to an end soon, and we have a busy fourth quarter ahead of us leading up to the midterms. U.S. politics is going to get loud, and loud politics tends to create extra jitters in a market that is always jittery.

On top of that we have the usual round of AI infrastructure jitters. The Wall Street Journal ran a piece last night that argues big tech's AI spending is roughly $3 trillion higher than it looks, because a large swath of those future obligations sit off the balance sheet. This would be in the form of uncommenced data-center leases and long-term purchase commitments for chips and power. I have lost count of how many times I have seen the word 'backstop' at this point. This all ties into the circular funding debate and those pretty diagrams we have all been staring at, showing the same pool of capital circling between the chipmakers, the cloud providers, and the AI labs.

The pull and tug continues between investors who trust the buildout and investors who are wary that it could disappoint. My own concern is that the whole structure leans too heavily on OpenAI and Anthropic delivering on very high expectations. There is real risk in having only two frontier labs anchor the entire buildout, and I don't know whether those two become great, durable business models or whether their models get commoditized and the disruption they are selling ends up eating them. As a long-term, somewhat passive investor who just wants to stay long the broad market, I do not think I need to pick a winner between the open-weight labs and the closed-weight labs right now. I just don't know the answer, and we must admit that no one knows.

What I do know is my comfort level. I already have enough exposure to the AI theme to sleep ok at night, and much more than that would start to make me feel uncomfortable. That is possibly the reason I trimmed the last 2 days.

Circling back to Lam Research

The last time I gave a portfolio update (on July 29th), I walked through Lam Research earnings, and I was pleased with what I saw. The numbers were strong and the geographic mix was a positive, because I want to see some diversification away from a heavy customer in China with the current export controls in place.

Lam ran up to its highs near $440 through the end of June before semiconductors pulled back through July, and even after rebounding since late July, it sits at $343 today. It had traded as low as around $251 on July 29th. At $343 now, the stock trades at about 36 times next-twelve-month earnings estimates. That multiple has actually compressed over the past few weeks, partly because the September guide stepped revenue up from $6.72 billion to $8.10 billion and lifted forward earnings, and partly because the stock itself pulled back. Yes, this is still an expensive stock, but the multiple was not the main reason I trimmed profits. If anything, the valuation has come down to a much more realistic level. This was simply a chance to take a tiny bit of profit heading into the fall, and I would not be surprised at all if the stock puts together another run from here.

Final Thoughts:

So as usual, I am not bearish on AI, AI stocks, the AI infrastructure buildout, semiconductors, memory chips, or Palantir's moat. Nothing about my long-term thinking has changed. This was just a position-sizing decision that may allow me to sleep a bit better tonight. I still want to own these two businesses, and I plan to hold them for years to come.

Thank you all for reading, and don’t forget to subscribe for more content on long-term investing.

The Rigatoni Capital Podcast is now on YouTube, Spotify, Apple Podcasts, Amazon Music, and iHeart Radio.

In Case You Missed:

Disclaimer: This blog is for informational purposes only and does not constitute financial advice. All opinions are my own, and I am not a financial advisor. The information provided reflects my personal views and is intended to encourage discussion and thought among readers. Investments involve risk, including the loss of principal, and past performance is not indicative of future results. Always conduct your own research or consult with a qualified professional before making any financial decisions.

As most of you know by now, Bill Ackman’s Pershing Square purchased Netflix, a stock I already own. I am assuming you have read a few opinions on this already, but how could Rigatoni Capital not give an opinion on this epic news?

Netflix was one of six new positions Ackman disclosed in the shareholder letter, alongside Visa, Mastercard, S&P Global, Intercontinental Exchange, and Alcon.

Before we dive in, we must agree that there is a special feeling, possibly a dopamine-like hit, that you get when a well-known superinvestor buys the same name you purchased before them. The one I recall most clearly was when Berkshire Hathaway revealed a healthy-sized position in Chubb in May of 2024, a stake its team had quietly built for months, and one I had owned for many years. Just like Netflix’s stock today, Chubb jumped on the news. There is a quick hit of validation, because as a long-term investor you have already done the work on the business and put your hard-earned money into the name.

The Rigatoni Capital Podcast is now on YouTube, Spotify, Apple Podcasts, Amazon Music, and iHeart Radio.

What Ackman Put in the Letter

The disclosure came in Pershing Square's second quarter 2026 shareholder letter, and since I own the stock, I read the whole thing closely. The letter lays out a long thesis on Netflix that I have been making on Substack and on the podcast for a while now. The short version is that the stock is trading at a steep discount, down roughly 50% from its June 2025 high, and that the market has priced the business wrong.

An important reminder: Ackman sold out of Netflix entirely on April 20, 2022, right near the bottom of that cycle. If he had simply held, he would be up roughly 230% from that date, even after the stock has given back close to half its value from June 2025.

Pershing Square describes Netflix as the dominant global streaming platform with more than 325 million subscribers, close to double the combined base of its two closest competitors. The firm points out that content spend has grown at only about 2% a year since 2021 while operating margins have expanded from 21% to around 31.5%, that the company converts roughly 90% of earnings into free cash flow, and that advertising is on a fast ramp toward roughly $3 billion in revenue this year. Over the last five years Netflix grew revenue at 12%, operating profit at 21%, and earnings per share at 27% a year, and the stock still de-rated from more than 40 times forward earnings down to about 21 times.

Netflix Bull and Bear thoughts

I have written about my Netflix position before, so I will not repeat all of it here, but I want to be honest about the part of the current strategy that I am not fully sold on. I put a lot of weight on Netflix continuing to pump out high quality original content, and lately that has felt like less of a priority than it used to be. I also know that this is a bias I carry as a consumer, because who doesn't want a great new series to watch? But that preference is not the same as an investment thesis. So I am trying to keep the bias out of the analysis for now, and it is quite possible management is working on correcting this. The issue could be cyclical, and management may be feeling some of the same pressure and frustration I am, which would push them to lean back into the high quality content that built the brand in the first place.

Outside of that one reservation, the business is firing on many cylinders. It has held onto subscribers through back-to-back price increases, which tells you a lot about how sticky the product is, and both cash flow and earnings look healthy even once you strip out the one-time $2.8 billion Warner Bros. termination fee. On capital return, Netflix pulled back on buybacks while it was pursuing Warner Bros. and then leaned back in once the deal died, authorizing a fresh $25 billion repurchase program in April.

While Netflix still looks like a growth stock on paper, it is also meeting the reality of a more mature business, since there are only so many subscribers left to add globally. That transition from a high-growth name toward a steadier and more predictable one may be a small part of why the multiple fell from around 40 times to the low 20s, even after the Warner Bros. deal fell through. Ackman does frame that same maturity as a positive, since a company that has already won the streaming wars does not need to keep spending like drunken sailors.

Ackman does have a history of knowing when a situation has gone too far in either direction. He took a swing at Chipotle during the 2015 to 2016 E. coli scare that cut the stock roughly in half. He then rode Chipotle for almost 10 years before fully exiting in early 2026, which capped an almost decade-long run.

As a reminder, Ackman was not so fortunate his first go around with Netflix. He first bought the stock in January of 2022, putting more than a billion dollars to work, and then sold only a few months later at a loss of more than $400 million after Netflix reported its first subscriber decline in a decade.

This go around, he feels enough is enough in the Netflix decline. The selloff kept going after Netflix's bid for Warner Bros., and then the stock stayed in the penalty box over softer watch time data from Nielsen. Putting a business like Netflix in a timeout over one engagement metric strikes me as short minded rather than long-term minded, and it is hopefully the kind of temporary dislocation that Ackman has built his career on buying.

The Investing Lesson Is to Look Where the Crowd Is Not

The most important aspect of this entire story is a theme that I have been discussing here on Rigatoni Capital since I launched the newsletter. Pershing Square notes that the S&P 500 rose about 10% through the first six months of the year, and that nearly all of those gains came from just two sectors tied to AI infrastructure, semiconductors and tech hardware and equipment, which together make up a small slice of the index. More than 90% of the companies in the S&P 500 contributed less than 2% of the return, and close to 40% of the index was down on the year. That concentration is exactly why a Visa or a Netflix can get forgotten while all of the capital and attention flows toward the AI buildout.

Of course I am not down on semis. I have very large exposure to the group, and I even entered into Broadcom this year for the first time. The SMH is now back above its 50-day, and the group has found its footing again after what some may call a hard "reset" from deleveraging last month. However, when one theme gets this crowded, the value can show up in the names that theme leaves behind, in high quality businesses that are not heavily exposed to AI or at serious risk of being disrupted by it.

As you may know, this is where I have been spending my time with research, and I believe if there is any value left in this market, it would likely be in more old economy stocks, or even blue chip names that still have many years of compounding in front of them. It is obvious that most of the capital has flowed toward a narrow part of this market, and the names left behind include great companies like Netflix along with plenty of stocks that do not touch the AI infrastructure buildout directly. Ackman is reminding investors to keep a running list of quality companies outside the crowded trade for the time being.

The Ackman Versus the Index Debate Is a Separate Conversation

On days like this, we get a lot of noise about Ackman’s long-term track record against the S&P 500, and this is a separate topic that I find to be a distraction. Also, I would never assume that because a well-known investor is stock picking, he knows something the rest of us cannot access in this day and age. Most of this information is transparent for long-term investors, and retail investors today can build as much conviction and knowledge as anyone, with the main exception being the ability to sit across the table from a management team in person.

I am not claiming Pershing Square will beat the index going forward by picking public stocks, and I would not be surprised if it turns out that Ackman has not beaten the indices over a long stretch, because that outcome takes a lot of luck. The question is a distraction, and it takes away from the main things we can all learn from professional investors like Ackman.

Thank you all for reading, and please don’t forget to subscribe for more content on long-term investing. And now I leave you with my top 20 holdings as of August 13th.

Top 20 Holdings as of August 13th:

In Case You Missed:

References

• Pershing Square Inc., Second Quarter 2026 Letter to Shareholders (August 12, 2026)

• Benzinga, “Bill Ackman’s Pershing Square Re-Enters Netflix Four Years After Losing More Than $400 Million” (August 2026)

• CNN Business, "Warren Buffett finally reveals the mysterious company he's invested billions of dollars in" (May 2024)

• Institutional Investor, "Facing Losses, Ackman Exits Chipotle and Nike" (November 2025), https://www.institutionalinvestor.com/article/facing-losses-ackman-exits-chipotle-and-nike

Disclaimer: This blog is for informational purposes only and does not constitute financial advice. All opinions are my own, and I am not a financial advisor. The information provided reflects my personal views and is intended to encourage discussion and thought among readers. Investments involve risk, including the loss of principal, and past performance is not indicative of future results. Always conduct your own research or consult with a qualified professional before making any financial decisions.

Disclosure: Long Berkshire Hathaway, $BRK.B. 1.85% of the portfolio, total gain +162%.

Berkshire Hathaway reported its second quarter this morning, and as always, it gives us a true read on the real economy that the mainstream headlines rarely deliver. The financial press will spend today, rightly, on the portfolio holdings story, as it’s the most entertaining aspect of Berkshire.

New CEO Greg Abel turned Berkshire into a net buyer of stocks for the first time in years, kept the share buybacks going, closed the splashy $6.8 billion acquisition of homebuilder Taylor Morrison, and put $10 billion directly into Alphabet through a private placement tied to the company’s AI infrastructure raise. We will get more details on the exact share counts and sizes of their holdings in the 13-F that lands in a week or so.

The purpose of this post is reviw the operating businesses that Berkshire owns and runs, and read what they are telling us about the economy, and more specifically the old economy, meaning the industrials, transportation, and energy-utilities that rarely get discussed.

Operating earnings rose about 16 percent to $12.98 billion, one of the highest quarterly operating totals the company has posted in recent years. However, I discovered that a good chunk of that 16 percent was not due to the operating businesses. Berkshire carries yen and euro denominated debt, and as those currencies weakened the reported value of that debt fell, which books as a gain of roughly $1.2 billion, a large year-over-year increase. If we strip that currency swing out, the underlying businesses grew closer to 5 percent.

The Rigatoni Capital Podcast is now on YouTube, Spotify, Apple Podcasts, Amazon Music, and iHeart Radio.

Insurance Segment Still Dragging

GEICO's loss ratio climbed to 76.6 percent in the quarter, an increase of nearly five points, driven by higher claims frequency and severity, with bodily injury severity running up ten to twelve percent. GEICO pre-tax underwriting earnings fell to $994 million from $1,821 million a year ago as a result, which pulled the whole underwriting line down for the quarter even though the reinsurance and primary businesses improved. The comparison also flatters last year in one respect, since the first half of 2025 carried the Southern California wildfire losses while the first half of 2026 saw no significant catastrophe events at all. Insurance investment income declined about 8 percent as lower short-term rates trimmed the yield on that giant Treasury bill pile, and float grew modestly to roughly $177.5 billion.

BNSF and the freight economy

Total carloads rose 6.5 percent in the quarter, led by consumer products and intermodal, which climbed 9.3 percent on higher West Coast imports, market share gains and tightening truck capacity. Agricultural and energy products rose 11.5 percent on grain exports and petroleum, while coal fell again as plant retirements and low natural gas prices cut into demand.

Fuel expense jumped 68 percent in the quarter on higher fuel prices, which is worth watching, but improved operating efficiency and the volume gains still pushed BNSF net earnings up 6.3 percent. Imports moving, trucking tight, coal fading, and that is a fair picture of where goods are flowing right now.

BHE, utility demand and the wildfire overhang

Berkshire Hathaway Energy posted a strong quarter with net earnings up 26.9 percent, helped by the U.S. utilities and by natural gas pipelines, where first-half earnings climbed after regulators approved higher rates and a colder winter pushed more gas through the system.

Retail electricity volumes grew 3.1 percent overall, with MidAmerican up 6.0 percent and NV Energy up 4.3 percent, reflecting higher customer usage and a rising customer count.

The overhang that has weighed on this segment also eased. PacifiCorp, the electric utility BHE runs across Utah, Oregon, Wyoming and other Western states, has been held liable for a series of 2020 and 2022 wildfires in its territory, and in April the Oregon Court of Appeals reversed the Phase I liability verdict against it, with the Oregon Supreme Court now set to hear argument on November 3. PacifiCorp has accrued roughly $2.85 billion in cumulative probable losses and paid about $2.3 billion in settlements, leaving unpaid liabilities of about $572 million at quarter end. The case is far from settled, though the appellate reversal is a meaningful step in the company's favor.

Industrials and the split in the consumer

The industrial side is strong, with Precision Castparts posting a 34 percent earnings gain on aerospace and industrial gas turbine demand, and metalworking business IMC growing revenue 26.5 percent, though some of that reflects customers accelerating orders ahead of price increases. OxyChem, the chemicals business bought from Occidental in January, added about $1.4 billion of revenue and $149 million of pre-tax earnings this quarter.

The consumer and housing side is where there is softness. Clayton Homes earnings declined on low home construction activity, and the HomeServices brokerage stayed weak on high prices and thin inventory. Auto unit sales at the dealership group slipped too, and the filing points to sluggish demand at the furniture and jewelry businesses, which ties to economic uncertainty and softer consumer confidence. McLane, the wholesale distributor that supplies groceries and consumer goods to convenience stores, restaurants and retailers, saw sales into its retail-store customers fall about 9 percent for the half on customer losses, though its restaurant business grew.

Berkshire Remains a Hold in the Portfolio

Berkshire’s operating businesses overall show decent strength, with certain segments showing softness. Anything tied to industrial production, aerospace, power generation and imported goods is running warm, while the consumer and housing markets are running cool. Lower short-term rates are also slowly draining investment income.

Berkshire will always remain a firm hold in my portfolio. That said, with the stock this hot, I would argue it is trading mostly on its large portfolio holdings like Apple and Coca-Cola rather than on the operating results. Nothing is broken in the underlying operations, but before getting too bulled up on the stock here, keep in mind the year-to-date move has pushed the B shares through strong resistance around $515 and toward all-time highs.

Of course, buybacks are worth watching in the coming quarters. Abel restarted the program earlier this year after saying the stock was trading below its intrinsic value, and while the roughly $4.8 billion repurchased this half is modest by Berkshire's standards, the tell will be whether he keeps buying or lets it taper with the B shares now at $521.

Thank you for reading, and please subscribe for more content on long-term investing.

Top 20 Holdings as of 8-7-26

References

• Berkshire Hathaway Inc., Form 10-Q for the quarterly period ended June 30, 2026 (filed August 8, 2026)

In Case You Missed:

Disclaimer: This blog is for informational purposes only and does not constitute financial advice. All opinions are my own, and I am not a financial advisor. The information provided reflects my personal views and is intended to encourage discussion and thought among readers. Investments involve risk, including the loss of principal, and past performance is not indicative of future results. Always conduct your own research or consult with a qualified professional before making any financial decisions.

One of the more important earnings reports of the season for me was AppLovin, which reported tonight. This is due to the many questions investors have as it officially opens its self-serve advertising platform (rebranded from AXON to AppLovin Ads) to all global businesses.

The last few days have been great for the Rigatoni Capital portfolio, pushing toward all-time highs on the back of Amazon, Palantir, Caterpillar, and Nvidia. AppLovin was one of my top holdings I was potentially the most concerned about, but more so just eager to hear what was happening this quarter during the rollout of AppLovin Ads.

I captured my updated portfolio after hours tonight so you get the most realistic picture following the AppLovin selloff. The stock was down almost 27 percent after hours ahead of the call, then bounced a little during opening remarks, and ended the session down about 16 percent at $350 per share, right back to an important support level that I lay out in the chart below.

Top 20 Holdings as of 8-5-26

AppLovin put up another quarter most companies would envy, with revenue up 53 percent year-over-year to $1.92 billion, net income up 55 percent to $1.27 billion, and adjusted EBITDA of $1.61 billion at an 84 percent margin.

The problem is that most companies are not priced the way AppLovin is priced (almost to perfection). The quarter itself was solid, but the slip was on the top line, where revenue of $1.92 billion came in about $16 million shy of the $1.94 billion consensus, alongside a third-quarter guide whose midpoint sat a touch under expectations.

Co-founder and CEO Adam Foroughi said the pace of meaningful model improvement was lighter than normal in the quarter, and the next step up in performance landed just after quarter end. Management was equally clear that it saw nothing pointing to weakening advertiser demand or a shift in the competitive environment. However, this does not address the concern that the consumer rollout is off to a slow start.

On a positive note, free cash flow guidance moved in the right direction. AppLovin is still a cash cow, generating far more than its operating business consumes, and I don't think this quarter changed that fact. The trouble is that the business must fire on all cylinders, with expectations being fairly high.

Going into the print, the company was roughly a $140 billion market cap, with a forward multiple around 25x and an NTM price-to-sales near 15x. After the selloff to $350 per share, my quick math shows the multiple compressed to 20x with an NTM price-to-sales near 12.5x. These are just quick math assumptions, so please don’t hold me to these numbers.

The Miss Was Blamed on Timing

Gaming is still the majority of revenue, and when the models get better, advertisers can profitably spend more at their target return, so budgets step up. When the models sit flat for a quarter, the whole machine idles. Foroughi was candid that this is the nature of research, telling analysts there is “no guarantee that we’re always going to have lifts in every single period of three months.” The team relies on compounding many small A/B-test wins across a quarter, and in Q2 those wins were smaller than usual, with a more material uplift landing in the early part of Q3. That is essentially their explanation for why the quarter was soft and why the company guided the next one up with confidence.

On margins, or the cost side of the business, CFO Matt Stumpf reiterated that the company spends roughly ten cents of compute for every incremental dollar of revenue. The higher training and inference costs tied to the models that went live in Q3 are already baked into guidance. Foroughi framed the tradeoff plainly, saying “every dollar of it is dependent on return,” and that when more compute produces more revenue through better model performance, they will make that trade every time.

Consumer Advertising Business

The consumer advertising business, which most bulls like myself treat as the real prize, had a good quarter on the surface. Advertiser spend set another record, finishing 28 percent above Q4 2025 levels, and Q4 is the seasonal peak for these advertisers, so growing past it in a slow quarter shows a steep curve. The catch is that this growth came from advertisers already on the platform, not from the public launch. AppLovin opened the platform to everyone in late June under the name AppLovin Ads Manager, but management was careful to say this launch would not change the business overnight, and that is where my slow-start concern sits.

Foroughi said "mid-market is a sweet spot right now," because those brands understand there is a learning cost to a new channel and will invest enough to reach their return goals and scale. The very small shops that sign up directly often cannot spend enough to get there yet, and the reason is creative, which he called the biggest hurdle in the system. AppLovin can auto-generate the interactive end card, but it still cannot reliably hand a small advertiser a finished 30 to 60 second video.

Final Thoughts

I will say again, I still believe the business is a cash cow, generating far more than its operating business consumes. The valuation is always going to be expensive, because this is a very attractive and exciting growth business. The company ended the quarter with $3.05 billion in cash against $3.7 billion of debt. Free cash flow was light at around $863 million, but Stumpf tied that to the timing of international cash taxes and interest payments, which he framed as a timing issue rather than a change in earnings power.

If demand is indeed healthy and the consumer side begins to scale faster, this is obviously still an exciting company. But AppLovin will remain priced for flawless execution, and that is why I would imagine the stock stays in the penalty box until we get more insight next quarter. Of course, I could always be wrong. We could see a strong bounce, but it would not change my positioning, as it remains only a hold for me in the portfolio.

Thank you for reading, and please subscribe for more content on long-term investing.

The Rigatoni Capital Podcast is now on YouTube, Spotify, Apple Podcasts, Amazon Music, and iHeart Radio.

In Case You Missed:

References:

AppLovin Corporation, Second Quarter 2026 Financial Results and earnings call, August 5, 2026. https://investors.applovin.com

Disclaimer: This blog is for informational purposes only and does not constitute financial advice. All opinions are my own, and I am not a financial advisor. The information provided reflects my personal views and is intended to encourage discussion and thought among readers. Investments involve risk, including the loss of principal, and past performance is not indicative of future results. Always conduct your own research or consult with a qualified professional before making any financial decisions.

The only time I can remember feeling like I truly won "the game" as a long-term investor was a single stretch built around one stock. I had held it with high conviction for at least two years, probably longer. I held through the drops because the conviction was real, and I added on the way down, and then I added again. It sat next to Nvidia as one of my largest positions. Some of you who know me personally can probably guess the name, and for everyone else it does not change the point, so I am going to leave it out.

Anyway, the stock was a dog for 2 years. But then the company reported earnings and I was extremely pleased with the quarter. The stock ripped, up double digits after hours. What I felt was relief more than anything, because I had spent years telling anyone who would listen why I was bullish, and it looked like the research had finally paid off. Then, in the following weeks, it did what this particular stock always did and began to fade.

This time the fade surprised me. I moved through confusion and denial and told myself the market simply did not understand the story. I blamed the insider selling that was in fact happening, I blamed the CEO, and I blamed a company that seemed a little too proud of its adjusted EBITDA. Inside of a few weeks or a few months I went from joy to relief to shock to anger. I was not journaling at the time, so the exact timeline is fuzzy to me now.

The Rigatoni Capital Podcast is now on YouTube, Spotify, Apple Podcasts, Amazon Music, and iHeart Radio.

The honest truth is that my weakness as a long-term investor has always been letting go and knowing when to sell. Buying into fear is something I tend to do fairly well. Selling is the part I rarely get right. Even after learning that lesson, I still do not agree with the narrative that "pigs get slaughtered" or "you never go broke taking a profit." My goal is not to book a profit. My goal is financial freedom and owning assets that compound over time.

Somewhere in the range of six to twelve months later I gave up on that business, which is one of the rare times I have done that in more than a decade of investing. I sold in chunks, and once selling started to feel good, I sold more chunks, and then more, until a few months had passed and I was fully out of one of the highest-conviction positions I have ever held. The good news, which I cannot prove to you, is that I did not lose money. I sold at a profit, with the later shares landing close to breakeven. Call it luck, even though the position was dead money for two or three years.

I bring all of that up because of how today felt. Heading into late July, and again at the short-term bottom last Wednesday, plenty of people panic sold or got margin called. The case for staying invested is that a small number of days tends to produce the bulk of a lifetime of gains, and today “may” have been one of those days. I was up 5.81%, which is one of the biggest single days the Rigatoni Capital Portfolio has ever had. On a day like that you might expect me to go out for a steak.

Top 20 Holdings As of 8-4-26

Instead I have the same eerie feeling I had when that old high-conviction name spiked after earnings. Palantir closed up 29.45% today, with some short covering in the mix and one of the best earnings reports I have ever read. It came right after Amazon’s strong report last Thursday and Caterpillar’s strong report this morning. Those are three high conviction names I write about here every day.

I make the unpopular case for the ROI behind the hyperscaler AI capex, the AI infrastructure and reshoring theme through Caterpillar, and a Palantir that, in my view, is not getting its lunch eaten by Anthropic.

I have recorded podcasts explaining why Palantir stays a hold in my portfolio even though I bought it at an expensive price. One of my first Substack posts, back in late 2024, pitched Caterpillar on cyclical tailwinds and the segment exposure I thought was about to work. I have written plenty on why Amazon looked undervalued to me and was either a buy or a hold. I do not give stock picks and I cannot predict the future, but today I feel vindicated, and I would be lying if I said otherwise.

That feeling is exactly the one I have learned to distrust. In the old story I felt relieved that the market had come around to my idea, and that relief arrived just before the stock rolled over. My Palantir position is now flat. That is not a reason to celebrate, and it is not a reason to feel relieved either.

The job now is to keep my eyes on the ball. I cannot predict what Palantir does over the next few weeks, and the long run is not a guarantee either. What I can control is where my attention goes, and my attention stays on the 2026 watchlist. That list includes individual names like TransDigm Group and American Express, alongside index exposure through the QQQ. The plan is to add or enter positions if they trade below what I believe is their intrinsic value.

The job is to also review earnings on everything I own, even the smaller positions, which means I revisit each business four times a year. The approach is called buy and homework, not buy and hold. I will still live and breathe the market, still try to understand geopolitics, macro, and market trends and the tailwinds and headwinds around them. But most of the time it’s best to take little action, because investors who are “too online” tend to make more mistakes than passive investors. The point is to avoid the major mistakes that could cause major hits to the portfolio.

So, no steak for you!

Palantir Q2 Earnings Recap

Palantir had one of the best earnings reports I have ever seen, which is scary because I have been saying that each quarter while the stock still has to grow into its valuation.

The Highlights

• Revenue Growth: Total revenue reached $1.935 billion, representing a 93% year-over-year increase [1].

• U.S. Commercial Strength: Revenue in this segment grew 149% year-over-year to $764 million [1].

• Profitability: GAAP net income was $1.062 billion, achieving a 55% margin [1].

• Operational Efficiency: Adjusted income from operations was $1.194 billion, reflecting a 62% margin [1].

• Cash Position: The company holds $9.2 billion in cash, cash equivalents, and short-term U.S. Treasury securities [1].

• Contract Momentum: U.S. commercial remaining deal value (RDV) grew 124% year-over-year to $6.238 billion [1].

Key Takeaways

• Sovereign AI Strategy: Management emphasizes that customers are increasingly prioritizing “AI sovereignty,” seeking platforms that provide control without forcing them to surrender their data as training material for future models [1].

• Deal Velocity: The company closed 73 deals of at least $10 million, signaling strong enterprise-level adoption and the ability to land and expand within large organizations [1].

• Rule of 40 Outperformance: A Rule of 40 score of 155% is significantly above industry benchmarks, underscoring the company’s rare combination of hyper-growth and high profitability [1].

• Non-GAAP Adjustments: The company continues to exclude stock-based compensation and related payroll taxes from its adjusted metrics, arguing these are non-cash or uncontrollable expenses that do not reflect core operational performance [2].

Future Guidance

The company now expects full-year revenue to fall between $8.150 billion and $8.158 billion, with U.S. commercial revenue projected to exceed $3.424 billion, representing growth of at least 134% [3]. Additionally, the company raised its adjusted income from operations guidance to a range of $4.889 billion to $4.897 billion and adjusted free cash flow guidance to between $4.5 billion and $4.7 billion [3].

Palantir’s commercial revenue ALMOST took the lead over government revenue for the first time. I suppose we must wait one more quarter for this historic moment to occur.

Caterpillar Q2 Earnings Recap

Lots of history-making in CAT’s earnings this morning. Record quarterly revenue surpassing $20 billion and a record backlog of $72.1 billion.

The Highlights

• Sales and Revenues: Increased 24% to $20.5 billion ^[1].

• Profit Per Share: Reported at $7.77, with an adjusted profit per share of $8.17 ^[1].

• Operating Profit: Rose to $4.295 billion, up from $2.860 billion in 2Q 2025 ^[2].

• Operating Margin: Improved to 20.9% (GAAP) and 21.9% (Adjusted) ^[1].

• Cash Deployment: Returned $2.2 billion to shareholders through share repurchases and dividends ^[1].

Key Takeaways

• Segment Performance: Growth was widespread, with Construction Industries leading the way with a 35% increase in sales, followed by Resource Industries at 20% and Power & Energy at 17% ^[3].

• Strategic Tariff Recoveries: The second-quarter operating profit included $392 million in expected International Emergency Economic Power Act (IEEPA) tariff recoveries, which contributed to the improved bottom line ^[2][4].

• Operational Efficiency: Despite inflationary pressures, the company successfully leveraged higher volumes to expand margins, with the Construction Industries segment profit margin increasing by 3.2 percentage points to 23.3% ^[5].

• Divestiture Impact: The company recorded restructuring costs related to the divestiture of certain non-U.S. entities, which were excluded from adjusted financial measures to provide a clearer view of ongoing business performance ^[6][7].

Future Guidance and Outlook

While the provided documents do not explicitly detail specific numerical guidance for future quarters or the full fiscal year, management emphasized that the current “strong order rates and a growing backlog” indicate sustained momentum across its primary segments ^[1].

Subscribe now

In Case You Missed

References:

• Palantir Technologies, Q2 2026 results press release, August 3, 2026. https://www.sec.gov/Archives/edgar/data/1321655/000132165526000039/a2026q2ex991pressrelease.htm

• Caterpillar Inc., second-quarter 2026 results, August 4, 2026. https://www.caterpillar.com/en/news/corporate-press-releases/h/2q26-results-caterpillar-inc.html

Disclaimer: This blog is for informational purposes only and does not constitute financial advice. All opinions are my own, and I am not a financial advisor. The information provided reflects my personal views and is intended to encourage discussion and thought among readers. Investments involve risk, including the loss of principal, and past performance is not indicative of future results. Always conduct your own research or consult with a qualified professional before making any financial decisions.

CompoundingAlpha

Actionable equity research, rigorous valuation models, and a transparent tracking portfolio.

By Nick | CompoundingAlpha

This podcast was recorded on Monday evening, August 3rd.

As a reminder, The Rigatoni Capital Podcast is now on YouTube, Spotify, Apple Podcasts, Amazon Music, and iHeart Radio.

Is the AI Infrastructure Boom Already Over?

Nick's latest post asks whether the AI infrastructure boom is already over, and it is a fair question after the month we just had. July ended with the blowup of Situational Awareness, which was forced to sell its entire public equity book to Ken Griffin's Citadel after margin calls from its prime brokers unwound a portfolio levered roughly four times over. The SOXX fell close to 29% from its June peak during the selloff, and the memory and chip names at the center of the trade, Micron and SK Hynix and Sandisk among them, took heavy damage before finding some footing near month end.

Nick's read is that the boom itself is not finished. The hyperscalers are still raising capex guidance for the coming year, and the projections he is watching have that spending running into 2028 and beyond. What he is less certain about is whether the easy money is still on the table. The first leg of this trade was piling into memory and chip names, and after the valuation reset of the past few weeks he would not call those easy anymore, even though he still sees upside left in some of them for investors who are bullish on the capex numbers. He pointed to Micron as one that could look reasonable if you believe the spending holds through 2028.

Where he sees the next phase is in the less obvious beneficiaries, the cooling and water treatment side and the broader energy build-out that has to happen to support all of this compute. On the demand side, he made the point that the spending is not slowing because the hyperscalers are already seeing the payoff.

CompoundingAlpha

Is the AI Infrastructure Boom Already Over?

If you’ve been anywhere near the stock market for the last three years, you’ve witnessed the gravitational pull of the “AI trade.” What started as a frenzy over generative AI software like ChatGPT quickly morphed into a massive infrastructure buildout. The logic was sound: if AI is the new gold rush, you want to be selling the picks and shovels…

Read more

a month ago · 7 likes · Nick | CompoundingAlpha

Palantir: The AI Gold Standard (PLTR)

Palantir reported its second quarter after the close while we were recording, and the headline numbers were about as strong as they come. Revenue grew 93% year over year to roughly $1.94 billion, U.S. commercial revenue grew 149%, and the company's Rule of 40 score climbed to 155%, a figure I find almost funny to look at.

Management raised full-year revenue guidance to roughly $8.15 billion, about 82% growth, and lifted the U.S. commercial and free cash flow outlook alongside it. The stock jumped around 10% in the initial reaction, a welcome change after a year in which it had been down close to 29% on fears that AI software disruption could hurt the moats of a company like Palantir.

The gold standard label comes from the product rather than the growth rate. Nick's argument is that Palantir's AI tools do not hallucinate or feed false data into enterprise workflows the way other systems can, and for the large, safety-conscious customers Palantir sells to, that reliability is the whole ballgame. We both credited the ontology, the data spine underneath the platform, as the piece that makes it work, and Nick expects Palantir to be one of the main beneficiaries of the agentic AI build-out. His one real worry is competitive, the chance that a model provider like Anthropic pushes into the same enterprise space, though he is not convinced that happens.

The valuation is rich and I will not pretend otherwise, but if the company keeps growing at these rates and keeps replicating in the commercial market what it built with government, it grows into the multiple faster than the skeptics expect.

CompoundingAlpha

The AI Gold Standard: A Deep Dive into Palantir ($PLTR)

If you want to understand Palantir Technologies ($PLTR), you have to understand its contradictions. For over a decade, it was Silicon Valley’s most secretive outcast—a company that eschewed consumer advertising, embraced the U.S. defense apparatus, and was run by a self-described socialist philosopher, Alex Karp, backed by libertarian billionaire Peter Thiel…

Read more

a month ago · 24 likes · Nick | CompoundingAlpha

Axon Enterprise (AXON)

Axon Enterprise is the company most people still remember as Taser, and it is another IGV component, a business-to-government software story now, that sold off hard on the same AI disruption fears. I have always thought of Axon as almost untouchably expensive on a price-to-sales basis, at times even richer than Palantir, so I wanted Nick's read on whether the business justifies it.

Nick likes both the stock and the company, and the growth story he is most interested in is the software side, specifically a product called Draft One that uses body camera footage to help officers write their reports. Axon has spent a decade collecting body cam data and storing evidence on its evidence.com platform, and Draft One puts that data to work by filling in the routine parts of report writing. The logic is simple, the average officer spends something like a quarter to 40% of the week on paperwork instead of being in the field, and freeing up that time carries real pricing power for Axon, especially at small departments running only a few officers per shift.

On my worry about how much market is left, Nick sees room in adjacent software, in international expansion, and in newer lines like a drone-as-first-responder program that lets an officer launch a drone from the car to get eyes on a scene. He also reminded me that the legacy Taser business is still large and runs on a razor-and-blades model, where recurring cartridge sales matter more than the hardware. The stock trades around 15 times sales by his estimate, which is expensive, so he would be hesitant to buy here, but he called it one of the premier businesses he has studied and a name he keeps watching for an entry point. Motorola is the competitor to keep in view, since its existing police radio relationships give it a way into the body cam market.

Toast (TOST)

Nick’s view is that Toast has effectively won the restaurant vertical, and the reason is that it does not sell a payment rail alone. It sells hardware plus a backend that handles inventory, payroll, loyalty, marketing, and even happy hour specials, and once a restaurant runs its operation on that stack it becomes very hard to leave. He sees Toast in mom-and-pop diners around where he lives as often as in chains, and those small operators almost never switch without a dramatic reason. The product gets stickier the more of it a customer uses.

Toast is also pushing up-market and sideways, into larger chains like the Darden brands, and into hotel management systems with partners like Hilton and Marriott, where the pitch is a unified guest experience, charging a poolside drink straight to a room number through the same system. Nick thinks the valuation is reasonable rather than stretched, and by his read close to 20% of the U.S. restaurant market now runs on Toast.

FTAI Aviation (FTAI)

FTAI Aviation started out as Fortress Transportation and Infrastructure Investors, which is where the ticker comes from. It later spun off its rail and infrastructure assets into a separate company, FTAI Infrastructure, leaving FTAI Aviation as a pure play in jet engines.

What separates FTAI from a traditional engine lessor like AerCap is its Module Factory. A conventional engine repair can keep an airline waiting three to four months, an eternity for a low-margin business that needs its aircraft in the air. FTAI keeps a large inventory of engines and swaps a working one in right away, then takes its time repairing the engine it pulled, the way you exchange an empty propane tank for a full one at the store rather than waiting for your own to be refilled. To stay asset-light, the company runs what it calls Strategic Capital Initiatives, private investment partnerships in which it holds roughly a 19% stake and which feed a steady stream of older jets and engines back to the factory while keeping those assets off FTAI's own balance sheet.

The piece Nick is most excited about is the newest one, converting retired CFM56 engines, the most widely produced commercial jet engine in history, into roughly 25-megawatt turbines that can power data centers, through a division called FTAI Power. He noted a data center engine order in the neighborhood of $1.8 billion against trailing revenue of about $3.1 billion, a large number for a company this size, with management expecting to be sold out of that capacity by around mid-2027.

CompoundingAlpha

Jet Engines and AI Data Centers: The FTAI Aviation ($FTAI) Deep Dive

When most investors look at the commercial aviation sector, they see a graveyard of capital. It is an industry infamous for cyclicality, massive capital expenditure requirements, and razor-thin margins. The traditional playbook—buying aircraft and leasing them out for decades—yields predictable but structurally low returns…

Read more

2 months ago · 20 likes · Nick | CompoundingAlpha

Thank you again, Nick, for joining me this evening, especially while the entertaining Alex Karp is probably going on one of his historic rants during the earnings call. Again, don’t forget to check out on Substack!

Thank you all for watching and reading. Don’t forget to subscribe for more content on long-term investing!

In Case You Missed:

Disclaimer: This blog is for informational purposes only and does not constitute financial advice. All opinions are my own, and I am not a financial advisor. The information provided reflects my personal views and is intended to encourage discussion and thought among readers. Investments involve risk, including the loss of principal, and past performance is not indicative of future results. Always conduct your own research or consult with a qualified professional before making any financial decisions.

Suddenly, just like that, the hyperscalers look like strong businesses again when you revise the story a bit, hold the capex estimates tight (kind of), and keep diversifying the book. Microsoft is diversifying away from leaning only on the frontier labs like OpenAI and Anthropic, and as a side note, Lam Research is doing the same away from just China. Suddenly the valuations in the hyperscalers look attractive again.

Sure, the huge margin calls from Situational Awareness, the AI hedge fund run by Leopold Aschenbrenner, could have been the reason for the significant sell off late yesterday when I bought the dip at 3:58pm exactly. But that is a short-term action story. And although it's one of the craziest stories of 2026, this is technically a distraction for long-term investors like myself.

As I was writing my post last night, I had CNBC on in the background, and most of the panel was extremely bearish after Meta's earnings. The way I remember it, more than one panel member basically said this is not a great environment for the hyperscalers and they were staying away. Microsoft answered that same night, and Amazon followed a night later.

The Rigatoni Capital Podcast is now on YouTube, Spotify, Apple Podcasts, Amazon Music, and iHeart Radio.

Microsoft

Microsoft may have been the story that read the room properly. The stock flew, up as much as 15% after the report, and I think it comes down to two things. First, the capex guide wasn’t increased, which matters a lot right now when the market is punishing capex hikes that don’t come with a clear payoff. Second, there was some news on the diversification front, with the commercial backlog jumping 84% to $678 billion and the sequential growth coming from customers other than the frontier AI labs. So the demand is broadening beyond a heavy dependence on compute from Anthropic and OpenAI. That is why the stock flew, at least in my opinion.

If you recall, I sold Microsoft late in 2025 at a total gain of +450%, because I was concerned about its heavy reliance and partnership with OpenAI. While I still think the company is probably a good long-term investment, my strategy from here would be to just buy the Nasdaq 100 instead of taking on more direct exposure to Microsoft.

Amazon

Full Disclosure: AMZN is 4.67% of my portfolio, total gain +122%.

I am extremely pleased with Amazon’s print. Net sales grew 20% to $200.6 billion, operating income was up 43% to $27.5 billion, and AWS was the star, growing 37% YoY to a $42.2 billion quarter and a $169 billion annualized run rate. CEO Andy Jassy called it the fastest AWS growth in 18 quarters. AWS operating income jumped to $16.6 billion from $10.2 billion a year ago, so you got faster growth and margin expansion at the same time. Advertising was strong again at 26% growth to $19.8 billion, one of the highest-margin lines in the whole company.

On guidance, Amazon lifted its 2026 capex guide to approximately $220 billion, up from $200 billion previously, and there could be a 2027 capex update coming as well on the earnings call. Normally a capex hike like that gets punished right now, and yet the stock still ran hard into the earnings call as I write this. So it works when there is a clear ROI or monetizing story, and Amazon has one, with AWS reaccelerating from 28% growth last quarter to 37% this one, the AI and chips businesses each past a $25 billion run rate, and Trainium pulling multi-year, multi-gigawatt commitments from the two leading labs. For the third quarter, Amazon guided revenue of $197 to $202 billion and operating income of $22.5 to $26.5 billion. Amazon remains a hold in the Rigatoni Capital portfolio (surprise, surprise).

Subscribe now

Apple

Full Disclosure: AAPL is 3.86% of my portfolio, total gain +637%.

Apple estimates beat, but the stock is selling off a bit, which I’m not surprised knowing how hard this stock has run into the print. It’s clearly the winner out of the Mag7 YTD. Revenue was up 16% to $109.4 billion, diluted EPS was $2.02, up 29%, and iPhone, Mac, and Services all set June quarter records, with every geographic segment growing double digits, including Greater China up about 22% to $18.8 billion. Gross margin came in strong at 50.1%.

A couple of things to keep in mind. The quarter included a favorable tariff-refund impact of about two points on gross margin and $0.11 on EPS, so the underlying beat is a bit smaller than the headline. And Services look a bit light at around $30.7 billion, growing about 12%, which isn’t great for their multiple since that is the high-margin recurring line the market leans on most. The stock touched roughly $5 trillion and an all-time high just two days ago, so a small pullback on a quarter this clean doesn’t worry me. This was also Tim Cook’s last earnings report as CEO before the handoff to hardware chief John Ternus.

Final Thoughts

Not surprisingly for the Rigatoni Capital portfolio, both Apple and Amazon remain a hold. As I have stated numerous times in my Substack posts and daily market podcasts, one day investors will suddenly look back at how cheap the hyperscalers were trading and think "damn, that was my chance at a wonderful business that was being punished for investing in real compute demand. I mean hell, the RPOs were right there in our face and we just mocked them and didn't think there could be any ROI!” Surely nothing comes risk free. There shall be volatility. The ROI has to be there, and the share buybacks are simply on hold for the next few years or so.

And of course, on the evening of the Situational Awareness blowup, it’s a good reminder that leverage always turns into disaster in drawdowns. That leaves me with the question I’m already asking 24 hours later, which is what am I going to do about the TQQQ position I now have? As you can see in the top holdings below, it’s now my 20th largest holding, taking the spot of XLF. As a reminder, this is not a great investment for long-term, buy-and-hold, smart investors.

Thank you all for reading, and please subscribe for more content on long-term investing!

Top Holdings as of 7-30-26

In Case You Missed:

References:

1. Amazon.com, Inc., “Amazon.com Announces Second Quarter Results,” July 30, 2026.

2. Apple Inc., “Apple reports third quarter results,” July 30, 2026.

3. Microsoft Corporation, “FY26 Q4 Press Release,” July 29, 2026.

4. CNBC, “Leopold Aschenbrenner’s hedge fund is facing steep AI losses,” July 30, 2026.

Disclaimer: This blog is for informational purposes only and does not constitute financial advice. All opinions are my own, and I am not a financial advisor. The information provided reflects my personal views and is intended to encourage discussion and thought among readers. Investments involve risk, including the loss of principal, and past performance is not indicative of future results. Always conduct your own research or consult with a qualified professional before making any financial decisions.

On arguably one of the busiest days in financial markets this summer, it was my last-minute decision to go ahead and buy the dip at 3:58pm EST. To be a professional buy-the-dipper, you have to be somewhat psychotic, and that is not something I trust you are, so please don't follow in my footsteps! Today I added to my QQQ position and even entered into a small TQQQ trade, small relative to portfolio sizing.

I don't recommend following my moves, let alone ever buying a 3x leveraged ETF like TQQQ if you're a long-term investor. But I would be a hack if I didn't tell you what I was doing.

While writing this, I even heard someone on the CNBC Fast Money panel sternly say, “I really don’t like this environment for the hyperscalers.” Naughty naughty if you think buying the dip in these names is at all a good idea. That is the kind of reasoning behind why I keep CNBC on in the background, the ultimate contrarian signal for me.

But of course, I don’t want to be a contrarian just to be a contrarian. The market can certainly test lower lows, and in this kind of bearish environment I half expect it to. Let me walk you through my rationale as a professional buy-the-dipper.

Subscribe now

The Rigatoni Capital Podcast is now on YouTube, Spotify, Apple Podcasts, Amazon Music, and iHeart Radio.

The macro backdrop

I'm writing this after the close on Wednesday, 7/29, after the Federal Reserve roughed up the markets, specifically the 10-year and other treasury yields. But again, this is fog of war in my opinion. It's the war of not knowing what Kevin Warsh is going to do, and the uncertainty around a very different Federal Reserve chair, on a night when we may see heavy retaliation from the U.S. after Iran struck U.S. forces in Jordan the night before.

Okay, so that is the macro outlook. Overall, we've had great earnings in financials, industrials, and some consumer staples, while the mega-cap tech names and hyperscalers have delivered more mixed headline numbers. Meta beat on revenue while EPS came in light, and the somewhat soft third-quarter revenue guide is interesting, since it pressures the ROIC on all that AI infrastructure capex. The move in yields is also weighing on tech stocks and longer-duration assets in general.

I will never time the exact bottom in the markets, let alone the Nasdaq 100. But the expected value, the EV, of owning a fairly diversified, market-cap-weighted index like the Nasdaq 100 is a great bet over the long term, provided you can control your emotions on paper and have a horizon of more than 10 years.

On the macro stories, I’m hopeful the inflation pressures prove transitory. The U.S. is stuck in the overseas drama, which is a shame and a stain on President Trump’s second term. But that doesn’t mean it lasts forever. If you have a long-term horizon, the same themes stay intact here at Rigatoni Capital.

The four bullish themes remain the same

Theme one is that regardless of which AI models get commoditized, the demand for AI infrastructure and compute is still there. The technology is disruptive and will change the way we live for the better. This benefits all the hyperscalers carrying global compute demand, and of course the bigger tech names get bigger in our story. The odds tilt in Nvidia's favor to double its market cap over the next few years.

Theme two is what we call the industrial revolution, western reshoring, defense manufacturing, refined rare earth dependence, and the broader industrial buildout. It can be backed by AI themes like data centers, but ultimately it comes down to building physical things like robotics, drones, and aerospace. A cyclical tailwind will come for most industrials, and it runs down the line from defense manufacturers to transports to raw materials to construction to energy to the moving of large things.

Theme three is the goldilocks in a growing, tech-driven economy, which sets up well for financials. As noted above, the yield-curve spread has to work in their favor, and a macro backdrop of deregulation and growth helps the financiers working behind the desks in the high rises of large corporate centers like NYC and Miami. In other words, a lot has to go right for financials to lead the market, but it's my opinion that it will happen in my lifetime.

Theme four is that there is still no alternative, TINA. You can earn some yield on U.S. treasuries, but they are still not the best store of value against inflation, especially in a world of fiscal dominance, where the system we all live in, at least here in the U.S., depends on the broader stock market, real estate, and even hard assets like gold and Bitcoin going from the bottom left to the upper right. This is the money-printer thesis, and the thesis that the Federal Reserve sits in a lose-lose situation as the U.S. debt nears $40 trillion any day now. Owning assets is still the only way to achieve wealth, and you can do it passively, as long as you can stick through the incredible amount of volatility.

The TQQQ trade

As stated above, I added a tiny bit to my largest position in the Rigatoni Capital portfolio, the QQQ. I also sprinkled in some TQQQ, which of course is not a wise choice for most investors and something I would never recommend to anyone. This is a short-term trade given it’s 3x levered, and while I know many have used this ETF as their personal ATM machine over the years, it’s not something I plan to hold in the Rigatoni Capital portfolio. As you can see below, it’s not in the top 20 holdings, so it doesn’t get presented here, but to be upfront, it’s my 21st holding for the time being, a total of 1.16%.

I want to use this post to communicate with my subscribers who follow my type of long-term thinking. I told you that if you’re subscribed, I would always let you know first whenever I make a buy or a sell in the portfolio.

Lam Research earnings

Lam Research is a top holding in the portfolio, currently up +720 percent on the position at roughly 2.66% of holdings. I usually don’t use a single quarter’s earnings report to decide whether to sell a position, but I do read the report, and as boring as it sounds, I normally just let you know it remains a hold. That is exactly what I would say today. The stock beat on the bottom line, with revenue essentially in line at $6.72 billion, up 30% year over year.

A very important takeaway is the geographic revenue mix this quarter. Taiwan came in first at 27%, just ahead of China at 26%, and together with Korea those three markets made up 73% of total revenue. While China sales decreased sharply on the new export rules, demand across the rest of Asia was strong enough that Taiwan and Korea offset those losses and then some. This is a big victory for a long-term investor like me who has been fearful of the heavy reliance on the Chinese market. It’s all about the continuation of diversification for this company.

For the quarter ending September 27, 2026, Lam Research guided to revenue of $8.10 billion, plus or minus $400 million, with a gross margin of 52.0% (+/- 1%), an operating margin of 39.5% (+/- 1%), and net income of $2.15 per diluted share (+/- $0.15) on a diluted share count of 1.255 billion. I haven’t had a chance to listen to the earnings call yet, but the usual questions will matter, including margin sustainability, the revenue guide jumping from $6.7 billion to $8.1 billion and where that jump is coming from, and any commentary on China under the new export controls. The photo below does not include this latest quarter’s guidance of $8.1 billion plus or minus $400 million.

Thank you for reading, and please subscribe for more content on long-term investing.

In Case You Missed:

References

1. CNBC, “Dow drops 1,100 points for worst day since April 2025,” July 29, 2026.

2. CNBC, “Fed meeting recap: July 2026,” July 29, 2026.

3. Al Jazeera, “Iran attacks US forces in Jordan as Trump meets Netanyahu,” July 29, 2026.

4. Lam Research Corporation, “Reports Financial Results for the Quarter Ended June 28, 2026,” July 29, 2026.

5. Investing.com, “Lam Research posts record quarterly revenue, boosts outlook on AI chip demand,” July 29, 2026.

Disclaimer: This blog is for informational purposes only and does not constitute financial advice. All opinions are my own, and I am not a financial advisor. The information provided reflects my personal views and is intended to encourage discussion and thought among readers. Investments involve risk, including the loss of principal, and past performance is not indicative of future results. Always conduct your own research or consult with a qualified professional before making any financial decisions.

On a day when Berkshire Hathaway closed up a whopping 3.06 percent, we have to hand it to the unsexy businesses that have been part of this mini rotation. Berkshire had a wonderful Tuesday, and I'm sure days like that won't last forever, but think about all the sectors the company operates in directly or holds through its portfolio, including railroads, utilities, energy, Coca-Cola, and of course Apple. The Dow climbed 659 points behind blowout earnings from Sherwin-Williams, which jumped about 6 percent and raised its full-year outlook, and Coca-Cola, which beat on both lines and lifted guidance. Paint, soda, and freight trains are wonderful places to hide while AI stocks take a breather.

While I have many great “boring” type of investments, like Waste Management for example, it got me thinking how frustrating it has been not having exposure to companies like Casey’s General Stores. This C-store is practically a popular pizza chain (the 5th largest in the country), but if you don’t live near these gas stations you would have no clue. Then you have O'Reilly Automotive, which has had a wonderful run of its own, though in my opinion the growing market share of electric vehicles is a headwind I would rather stay away from at today's price.

So the question is “What other companies share this type of business model, the kind that compounds quietly for decades while nobody wants to talk about it at a cocktail party?” I'm surely missing a few, so I decided to put the ‘ole AI chatbot to work to see what it would come up with.

For idea generation I used the default Gemini 3.5 Flash model with a simple original prompt: "I'm looking at unsexy stocks similar to both Casey's General Store and O'Reilly Automotive that have made investors billions. Please provide other ideas that could be the next similar stock." Gemini would only produce three names at a time, so I kept prompting and fine tuning the search until the list grew to 23 names. From there I handed the list to Claude’s Fable 5 model, which ran the online research, allegedly checked the numbers against live market data, and produced the PDF table you’ll find below.

Some of the return figures in that PDF look off to me, and I can’t imagine the document is one hundred percent bulletproof. But that’s part of the exercise here, to show you that even the latest frontier AI models are still not bulletproof. So please treat the PDF as a demonstration and a starting point. Always verify everything yourself when using AI.

As I'm realizing now, the 23 names appear in no particular order, which is kind of nuts, but I think you still get the idea. Some of these were quite interesting and had never been on my radar, which is the entire point of this post. That said, no single stock piqued my interest enough to earn an immediate spot on my current watchlist, though I'll probably revisit a few of these names in the coming days.

For the purpose of the exercise, I’ve included a brief background on what each business does, along with a quick bull and bear case for all 23 names. FYI, this section was also powered by Fable 5, so beware, fellow investors.

Casey’s General Stores (CASY)

Casey’s runs over 2,600 convenience stores across small-town America and happens to be one of the largest pizza chains in the country. The bull case is a decades-long formula of prepared food margins, fuel traffic, and acquisitions in towns too small for competitors to bother with. The bear case is that everyone finally noticed, and the stock now trades around 45 times trailing earnings, the richest valuation in its history.

O’Reilly Automotive (ORLY)

O’Reilly sells auto parts to both do-it-yourself customers and professional mechanics through a distribution network that gets parts to a repair bay within hours. The bull case rests on an aging vehicle fleet, dual-market scale, and one of the most relentless buyback programs in the market. The bear case is a roughly 30 times multiple on mid single digit growth, plus the long-term question of what electric vehicles do to parts demand.

Watsco (WSO)

Watsco is the largest distributor of air conditioning and heating equipment in North America, selling to over 120,000 contractors. The bull case is replacement-driven demand, a fortress balance sheet, and a long runway to consolidate a fragmented industry. The bear case is a soft housing cycle, returns on capital that have drifted lower, and a stock down over the past year while still carrying a premium multiple.

Copart (CPRT)

Copart auctions totaled vehicles for insurance companies through an online platform and a land network that would be nearly impossible to replicate. The bull case is a duopoly with 33 percent net margins, roughly five billion dollars of net cash, and a valuation near 17 times earnings, the cheapest it has traded in a decade. The bear case is that the stock is down about 46 percent in a year for real reasons, with growth slowing and a CEO transition that brought back Jay Adair while investors question what management will do with all that cash.

Rollins (ROL)

Rollins owns Orkin and a family of pest control brands serving millions of residential and commercial customers. The bull case is recurring revenue, pricing power, route density, and an endless supply of small competitors to acquire, because the bugs always come back. The bear case is that even after a 26 percent drawdown the stock trades near 36 times earnings while earnings growth has decelerated.

Pool Corporation (POOL)

Pool Corp is the world’s largest wholesale distributor of swimming pool supplies, chemicals, and equipment. The bull case is a massive installed base of pools that need maintenance every year regardless of the economy, with the stock down roughly 55 percent from its 2021 peak and trading below 20 times earnings. The bear case is that new pool construction remains tied to housing and the post-pandemic hangover has lasted longer than anyone expected.

Carrier Global (CARR)

Carrier manufactures the HVAC equipment that distributors like Watsco sell, along with transport refrigeration. The bull case is surging data center cooling orders, the Viessmann heat pump acquisition in Europe, and a forward multiple near 25 times. The bear case is a trailing multiple near 46 times on depressed earnings, a leveraged balance sheet, and a company still proving it can execute as a standalone business.

United Rentals (URI)

United Rentals is the largest equipment rental company in North America, renting everything from forklifts to generators for mega projects. The bull case is scale advantages, secular growth from infrastructure and manufacturing buildouts, and a strong second quarter that came with raised guidance. The bear case is that this is a deeply cyclical, capital intensive business trading near all-time highs after a 55 percent run in a year.

Cintas (CTAS)

Cintas rents and launders uniforms and supplies businesses with mats, first aid kits, and fire protection services. The bull case is one of the great recurring revenue route businesses in America, with net margins near 18 percent and returns on equity above 40 percent. The bear case is a 43 times multiple that prices the stock for perfection on high single digit growth.

Waste Connections (WCN)

Waste Connections collects trash and owns landfills, focusing on secondary markets where it operates as the local monopoly. The bull case is pricing power, expanding margins, and the most durable demand imaginable, since the garbage never stops. The bear case is a 40 times trailing multiple against revenue growth in the mid single digits, which means margin gains are doing most of the work.

SiteOne Landscape Supply (SITE)

SiteOne is the only national wholesale distributor of landscaping supplies, from irrigation to fertilizer to hardscapes. The bull case is a classic roll-up story consolidating a fragmented industry, with a long acquisition runway. The bear case is thin net margins under 4 percent, housing-linked demand, and a stock that has gone nowhere for five years.

Driven Brands (DRVN)

Driven Brands franchises and operates automotive service concepts, led by the fast growing Take 5 Oil Change, after divesting its car wash business. The bull case is a cleaned up story at about 13 times earnings, the cheapest multiple on this list, with Take 5 posting steady comps and debt coming down. The bear case is heavy remaining leverage, thin free cash flow through the restructuring, and a history of brand sprawl that investors have not forgiven.

Core & Main (CNM)

Core & Main distributes water, wastewater, and fire protection products to municipalities and contractors. The bull case is sixteen consecutive years of sales growth, a municipal customer base that has to maintain aging water infrastructure, and a reasonable 19 times multiple. The bear case is exposure to residential construction and margin improvement that still has to prove itself durable.

Comfort Systems USA (FIX)

Comfort Systems installs the mechanical, electrical, and plumbing systems inside complex buildings, which now means AI data centers. The bull case is a 14 billion dollar backlog, earnings that roughly doubled in a year, and a front row seat to the largest construction boom in decades. The bear case is that a stock up roughly 300 percent in a year at 42 times earnings has stopped being unsexy, and any pause in data center capital spending would hit hard.

APi Group (APG)

APi provides fire protection, life safety, and specialty services on an inspection-first model that generates recurring revenue. The bull case is a management target of ten billion dollars in revenue by 2028, growing data center exposure, and a forward multiple near 22 times. The bear case is messy reported earnings that make the trailing multiple useless, plus integration risk from a steady diet of acquisitions.

Stella-Jones (SJ)

Stella-Jones manufactures pressure treated utility poles and railway ties from its base in Montreal. The bull case is grid hardening and electrification demand for poles, dominant share in sleepy niches, and a modest 14 times multiple. The bear case is lumber cost pass-through timing, a soft residential lumber segment, and the limited liquidity that comes with a mid cap listed in Canada.

Modine Manufacturing (MOD)

Modine spent decades as a forgettable maker of radiators and has reinvented itself as a data center cooling company. The bull case is that chillers and cooling distribution units for AI facilities have transformed the growth profile, with the stock up about 179 percent in a year. The bear case is a GAAP trailing multiple near triple digits, margin expansion that has come slower than promised, and the memory of twenty years as a value trap.

Dycom Industries (DY)

Dycom builds and maintains fiber optic networks for telecom carriers and has pushed into data center and power work through acquisitions. The bull case is the fiber buildout plus AI infrastructure demand, with a recent quarter showing 56 percent revenue growth and record backlog. The bear case is a 42 times multiple on a low margin construction business with meaningful customer concentration.

EMCOR Group (EME)

EMCOR is the largest electrical and mechanical construction firm in the country, wiring and cooling everything from hospitals to data centers. The bull case is raised 2026 guidance, a conservative balance sheet, and returns on capital near 38 percent, remarkable for a contractor. The bear case is a stock up 74 percent in a year trading around 29 times earnings versus a historical multiple in the teens, with results dependent on backlog conversion.

Tractor Supply (TSCO)

Tractor Supply is the largest rural lifestyle retailer in America, selling feed, fencing, tools, and pet supplies to hobby farmers. The bull case is a loyal needs-based customer, a valuation near decade lows around 15 times earnings, and a 3 percent dividend yield after the stock was cut nearly in half. The bear case is negative comparable sales, a lowered 2026 outlook, weakness in the pet category, and a pressured rural consumer.

Vulcan Materials (VMC)

Vulcan is the nation’s largest producer of construction aggregates, the crushed stone and gravel underneath every road and foundation. The bull case is irreplaceable quarry assets near growing metros and decades of uninterrupted pricing power. The bear case is a 34 times multiple on a business whose volumes remain cyclical and weather dependent.

Murphy USA (MUSA)

Murphy USA runs low priced fuel and convenience stores, often in Walmart parking lots, and has retired roughly half its shares since its 2013 spinoff. The bull case is a disciplined buyback machine trading at 0.6 times sales with a proven low cost model. The bear case is dependence on volatile fuel margins, net margins near 3 percent, and the long-term electric vehicle question hanging over gasoline volumes.

Fastenal (FAST)

Fastenal distributes fasteners and industrial supplies through vending machines and onsite locations embedded inside its customers’ facilities. The bull case is 15 percent plus net margins in a distribution industry where most peers earn half that, driven by technology and service intensity. The bear case is a valuation around 41 times earnings, the richest in the company’s history, for a business growing high single digits.

Thank you all for reading. A bit different of a post today, but it was fun, and that’s what Rigatoni Capital is all about, right?

If you enjoyed it, please subscribe for more content on long-term investing.

The Rigatoni Capital Podcast is now on YouTube, Spotify, Apple Podcasts, Amazon Music, and iHeart Radio.

In Case You Missed:

Yesterday’s podcast with special guest (below)

Top Holdings as of July 28th:

References:

Yahoo Finance company statistics pages for the tickers discussed (finance.yahoo.com)

StockAnalysis.com company statistics pages for the tickers discussed (stockanalysis.com)

Simply Wall St earnings coverage of Rollins, Cintas, Waste Connections, Comfort Systems, Fastenal, and EMCOR (simplywall.st)

Company press releases and SEC filings: Comfort Systems USA Q2 2026 release, Dycom fiscal 2026 fourth quarter release, Vulcan Materials Q1 2026 release, Tractor Supply Q2 2026 release

Morningstar, TipRanks, GuruFocus, and MarketBeat quote pages for supplemental market capitalization and valuation data

Disclaimer: This blog is for informational purposes only and does not constitute financial advice. All opinions are my own, and I am not a financial advisor. The information provided reflects my personal views and is intended to encourage discussion and thought among readers. Investments involve risk, including the loss of principal, and past performance is not indicative of future results. Always conduct your own research or consult with a qualified professional before making any financial decisions.

As Marc pointed out at the beginning of the podcast, this sell-off in semis and AI infrastructure theme names has been breaking down for a few weeks now. It's Marc's opinion that this has been a leverage unwind rather than a valuation or earnings problem.

But before we recap the podcast episode, Marc is almost at 800 subscribers as I write this post. If you enjoy his content, I think you will be very pleased with his Substack newsletter - link below:

Learning.Investing.Thriving.

Learning.Investing.Thriving. is built on a simple idea: If you want to build wealth, you need more than stock tips. You need to understand how markets actually work.

So back to the leverage unwind. That has certainly impacted the South Korean market, and we must agree that while memory chips like SK Hynix have lower PEs due to their cyclicality, they still have tripled or even quadrupled earnings within the last year. The results have been extremely impressive, with Samsung posting a roughly 19x jump in profit and Micron putting up record revenue north of $40 billion.

Marc wrote a piece on the Korean market earlier this month showing how heavily Korean investors were levered into options and leveraged products on Samsung and SK Hynix, with daily turnover in those vehicles running double to triple what a comparable US name would see. The same structures that exaggerated the move up, where the funds buy more to hedge as prices rise, now have to sell into the decline, and in Korea that shows up as ten percent down days at the index level. In the US it is the leveraged ETFs doing the same thing but on a smaller scale. Stripped down, the move looks a lot more like margin calls than a rerating.

Mark to market - Alphabet Earnings

Alphabet’s latest quarter carried roughly ninety nine billion dollars in gains on its holdings in Anthropic and SpaceX, which pushed GAAP net income up almost three hundred percent and made the headline look spectacular. Strip that paper gain out, which any professional will do, and the core business came in a touch light against expectations.

Regardless of the large capex spend, and free cash flow going negative for the first time, it was nice to hear Marc state that Alphabet is still likely a good long-term investment. Too often we get wrapped up in the spooky AI headlines and need to realize that the business as a whole is still firing on all cylinders. This is why I personally believe the stock has a floor here, and there will be dip buyers around its 200 day moving average (at least in the short term). That is assuming we don’t have a broader tech sell-off or any unexpected macro or Google-specific headlines in the coming weeks.

Full disclosure: I do not own GOOGL directly.

Rotation into financials?

I made the case that financials and industrials deserve more attention than they get, and Marc took the financials side further. He called the setup a Goldilocks moment. Bank earnings this quarter were strong across the board rather than propped up by one line, and the tailwinds look set to continue, with more large IPOs coming, Anthropic likely among them, a steepening yield curve, and volatility that feeds the trading desks. He noted that JP Morgan is knocking on the door of a trillion dollars in market value while still trading around twelve times earnings. His other angle was European index funds, where the big banks carry far more weight in their home indices than any US bank does in the S&P, so buying the index there gives you real financials exposure and some diversification in one shot.

Jobless claims

We closed on initial jobless claims, which came in at 187,000, the lowest reading since 1969. Marc's caution is that this looks less like a booming economy and more like a frozen one. Adjusted for how much larger the labor force is today it is arguably the lowest ever, but continuing claims have held steady around 1.8 million, which is not what you would see if the low number meant people were getting rehired quickly.

His read is a hiring and firing stalemate, where companies are not letting people go but are not adding either, which is why the unemployment rate keeps sitting in the low fours no matter what gets thrown at it.

If you enjoyed this podcast, please consider checking out Marc’s work at https://rudajev.substack.com/.

The Rigatoni Capital Podcast is now on YouTube, Spotify, Apple Podcasts, Amazon Music, and iHeart Radio.

In Case You Missed:

Disclaimer: This blog is for informational purposes only and does not constitute financial advice. All opinions are my own, and I am not a financial advisor. The information provided reflects my personal views and is intended to encourage discussion and thought among readers. Investments involve risk, including the loss of principal, and past performance is not indicative of future results. Always conduct your own research or consult with a qualified professional before making any financial decisions.

The open weight AI models coming out of China are going to keep getting better, because the technology scales and compounds and there is no obvious ceiling that stops a well funded lab from shipping more capable open model every few months.

These open models are not going to replace American labs like OpenAI and Anthropic overnight, and I think the closed labs can hold onto the business model that monetizes the model itself through API pricing and subscriptions, at least for the high end work where quality and reliability justify the premium.

All-In co-host David Friedberg on a recent podcast explains it as the enterprise gets to treat AI models like an à la carte menu. A company can keep paying Anthropic for the specific workloads where its model earns the cost, and at the same time pull in a cheap Chinese open weight model for the high volume daily use cases where paying frontier prices makes no sense. The enterprise plugs models in and out based on what each job needs, and the savings on the routine work are dramatic.

This lines up with the argument I made last week in my post titled "Why Open Weights Are Good For the Hyperscalers and AI Infrastructure."

As AI scales and compounds into something even more useful, the enterprise still needs the infrastructure and cloud layer to run all of it, while gaining switchability at the model layer. The enterprise keeps paying for the cloud that runs everything and loses the lock-in that used to tie it to a single model provider.

The Rigatoni Capital Podcast is now on YouTube, Spotify, Apple Podcasts, Amazon Music, and iHeart Radio.

The prupose of this post is to try and get inside the head of the Chinese Communist Party and President XI Jinping in the context of a very heated AI Arms Race with the United States. President Xi called on countries to encourage open source, openness, collaboration and sharing, and he positioned China as the leading advocate for open source AI against Washington's more closed and export-restricted approach.

It seems to me that China wants to commoditize the model layer that frontier labs like Anthropic and OpenAI depend on for margin.

The first idea worth understanding is what is China ultimately better at than the U.S.? I would argue manufacturing and building things. Think of everything that sits in the layers beneath the AI model, in manufacturing, robotics, hardware, energy, cloud capacity, and the physical economy. Xi has framed AI as moving out of the digital world and into the physical world. This is why cheap and commoditized AI models help China, the manufacturer of the world.

Underneath the business model there is a founder motivation that was highlighted in the news last week. Moonshot AI founder Yang Zhilin allegedly has one real goal, which is to reach artificial general intelligence. A founder can sincerely want AGI above everything, and the state can be delighted to let him chase it, precisely because his open, research first, margin indifferent approach delivers the national outcome for free. In other words, Beijing has made the decision to get out of the way.

It’s my opinion that if China is smart, open weight is a win-win for their country. It points back to China’s larger goal, which is to stay the world’s leading manufacturer and exporter of physical products. AI robotics is the clearest example of that ambition.

A commoditized, cheap, globally adopted model layer feeds into the cloud and compute infrastructure China is building. It should also accelerate the robotics and physical economy that President Xi envisions. If a Chinese AI company reaches AGI in the meantime, that is also something to be proud of and a win for the world.

Thank you for reading, and don’t forget to subscribe for more content on long-term investing!

Top Holdings As Of 7-27-26

In Case You Missed:

Disclaimer: This blog is for informational purposes only and does not constitute financial advice. All opinions are my own, and I am not a financial advisor. The information provided reflects my personal views and is intended to encourage discussion and thought among readers. Investments involve risk, including the loss of principal, and past performance is not indicative of future results. Always conduct your own research or consult with a qualified professional before making any financial decisions.

Mytwocent$

Casual observations on the passing scene

By Bagholder

Before we get started on the podcast recap, as a friendly reminder the Rigatoni Capital Podcast is on all major platforms that include YouTube, Spotify, Apple Podcasts, Amazon Music, and iHeart Radio.

The podcast is mostly a morning show where I go over finance news and dissect what it means for long-term buy and hold investors. I'm also fortunate to have wonderful guests come on from time to time like .

Silver and the Silver Miners

Silver has come down hard. It peaked somewhere around $110 to $120 an ounce in late January and now sits near $58, which puts it roughly half off its highs. Kevin has traded the metal since the early 1990s and leans on the Edwards and McGee school of technical analysis, and he does not find the drawdown alarming. He reads the fifty year chart as a textbook cup and handle formation, and his argument is that a pattern which takes forty five years to build does not resolve itself in a three month rally. What usually follows a breakout above a long watched level, in this case the $50 area, is a run and then a pullback all the way back to that breakout point before the next leg higher. He thinks silver is doing exactly that, and that the $50 to $55 zone is where a lot of the people who sold on the way up turn into buyers.

Kevin has been working through the SILJ ETF, the silver miners index, whose two largest holdings are First Majestic (AG) and Hecla (HL) at a combined fifth or so of the fund. His observation is that the metal and the miners have diverged badly over the last decade. By his math, silver is up something like three and a half times over ten years while the SILJ has crept from around $20 to $24 and effectively flatlined, even though miners are supposed to lever the metal's moves. Looking at the top two names over the last five years, he figures First Majestic has grown operating cash flow around sixfold and Hecla around twentyfold, with the group averaging roughly a thirteenfold jump in free cash flow against a doubling in the metal. Those multiples are his own quick math and I would hold them loosely, but the width of the gap is what carries his point. As he put it, "if that doesn't scream bargain, I don't know what does." He reached for a line from short seller Jim Chanos to frame it, paraphrasing him to the effect that bull markets pay a premium for what could be while bear markets discount reality.

Gold and the Gold Miners

Gold is trading near $4,055/oz and sits only about 20% off its highs, and Kevin owns the GDX miners ETF along with several of its components. His top five there are Newmont, Agnico Eagle, Barrick, AngloGold Ashanti, and Franco-Nevada, and he says their cash flows have tripled on average over three years, with a few names up six or seven times and Newmont lagging at a doubling. Yet the GDX itself has climbed a little less than gold over the same stretch, so the operating leverage that shows up clearly in the companies' cash flow is barely showing up in the fund's price. It is the same disconnect he sees in silver, priced into the miners rather than the metal.

On why gold itself should head higher, Kevin leans on demand rather than industry. Only a small slice of gold goes into jewelry and a tiny fraction into electronics, so the bid comes from central banks and from countries moving away from dollar denominated assets, a trend he thinks is still running at record levels with China among the larger buyers. Pair that structural demand with the same technical setup, a pullback to the breakout level after the momentum crowd has been chased out, and he sees enough buyers left to push it back up.

Bitcoin, Strategy, and the Stretch Preferred

Bitcoin is off roughly half from its highs and trades near $64,000, which brings us to Strategy (MSTR). I opened by admitting I am now a bagholder of the common, and I framed what I see as the core tension. Strategy executive chairman Michael Saylor can look after the preferred holders, the Bitcoin holders, or the common shareholders, and it is hard to keep all three groups happy at once. Kevin disagreed with the premise. His view is that there is almost always an accretive move available. When Saylor retired a $1.5 billion tranche of debt at 92 cents on the dollar about a month ago, the roughly $200 million discount flowed straight to shareholders, and the same logic applies to buying back preferreds below par.

China, Quality, and Open Weight AI

The most memorable stretch came out of Kevin's family trip to Iceland. Hertz was out of Teslas, so they handed him a Hongqi, China's luxury marque, and after ten days he called it the finest car he has ever driven, better in feel and quality than anything coming out of Germany or the United States, with around 500 kilometers of range and a fast charge. A Huawei hotspot they threw in ran better than the internet at his house. His larger point was generational. The old assumption that the American version of a product was the reliable one and the Chinese version was junk may simply have flipped, and once he started noticing it in the car and the hardware, he started wondering about it everywhere.

China is graduating engineers by the millions each year, far more than the United States, and pushing its models out as open weight while the American labs stay closed. As it happened, Nvidia CEO Jensen Huang chose that very day to make his first post on X a coalition letter titled "Open Weights and American AI Leadership," signed by around twenty five firms including Microsoft, Meta, and Palantir, with OpenAI, Anthropic, and Google conspicuously absent. I made the case that the fear of Chinese open weights as a vector for IP theft misses how the technology works, since you can download the weights and run them on your own hardware, and that open weight arguably helps Nvidia, Huawei, and the cloud providers rather than hurting them.

Moonshot's Kimi K3, released a week or so earlier, is the current example. ‘The Nvidia Way’ author Tae Kim made a similar point on Fox Business, that open weight is good for the ecosystem and the infrastructure names while the closed labs sit on models several steps ahead, waiting on approval and perhaps an IPO. My own conclusion is that all of this commoditizes the model layer, which is why I would be a seller rather than a buyer of an OpenAI or Anthropic listing when it comes.

I would like to thank you all for reading, watching, or listening.

The Rigatoni Capital Podcast is growing, and I cannot thank the finance Substack community enough for the great support I have here.

In Case You Missed:

Disclaimer: This blog is for informational purposes only and does not constitute financial advice. All opinions are my own, and I am not a financial advisor. The information provided reflects my personal views and is intended to encourage discussion and thought among readers. Investments involve risk, including the loss of principal, and past performance is not indicative of future results. Always conduct your own research or consult with a qualified professional before making any financial decisions.

Full Disclosure (I own the following positions):
Lockheed Martin, 1.66% of my holdings, total gain of +87%.
RTX Corp, 1.55% of my holdings, total gain of +152%.
Union Pacific Corp, 0.60% of my holdings, total gain of +105%.

As industrial earnings near the end, we have the market rewarding industrials against the backdrop of a theme I have been discussing on here for months, which is a mini wartime economy and potentially a small defense industrial revolution. This theory of mine has not come to fruition completely as I would have envisioned it, but that may be because we are early in earnings. This is more of a geopolitical and macro theme, and those are extremely difficult to predict the timing of.

There are plenty of times when we see strong earnings in industrials led by the defense manufacturers that only lead to investors taking profits the next day. The common sense investing thesis is that we need more weapons, and we need them as fast as possible. The more complicated long-term theory is that the West, or the U.S., is figuring out a way to become less dependent on materials and supply chains that have China brokering, and is therefore trying to build here as fast as possible so we are not left so dependent on our competitors. That means building data centers, robotics, defense missiles, jets, drones, energy, pharmaceuticals, and just about anything defined in the Defense Production Act that would incentivize reshoring critical materials.

Plenty of times have I discussed the technological upgrades needed to run manufacturing better here in the West, and that is when I usually bring up Palantir CTO Shyam Sankar. If you have listened to many of his previous interviews, you will know that he lays out a vision that aligns with the same investing thesis we have here at Rigatoni Capital.

We cannot simply ignore old economy industrials or old economy financials as long-term buy and hold investors. These sectors have too much overlap now with technology as a sector, and there is a convergence that has been happening for many years and will likely compound on itself. In other words, old economy HALO-type businesses are integrating AI technology and technically are running on custom proprietary software applications. Financials are also becoming fully integrated with AI technology, and this trend will continue over time.

This allows what Bill Ackman would describe as a productivity boost in the overall economy, which is a far more useful frame than the fearmongering about job losses. Profit margins in both financials and industrials continue to impress many investors even while we see inflationary pressures in materials, energy costs, regulatory hurdles, supply chain bottlenecks, and even wage growth.

Lastly, if you follow me, I do not go on and tell you to hurry up and buy financials and industrials today, although I do feel we could be on the cusp of Goldilocks for both sectors. That would just be unfair to you, the reader, while I am doing the opposite. I am already in these names, and some of them, like Lockheed Martin for example, I entered into in the year 2015. I made the decision to allocate to these sectors more than ten years ago.

The idea is to be focused in three sectors, tech, industrials, and financials. I am not going to do an updated breakdown today, but I would imagine the portfolio mirrors something like tech at 70%, while financials and industrials sit at 15% each. Again, that is my portfolio as of today, July 23rd.

As usual, my only suggestion if you are a long-term investor is to look for bargain deals, decide what each sector's or individual stock's intrinsic value is, and then really buy below that intrinsic value.

Even after today's wonderful day in industrials, led by the defense manufacturers, this is not a victory lap post. Too many times have we seen a quick sell-off where capital gets distracted back into another tech rotation, which is highly likely here. The next rotation is more likely to happen inside the tech sector, whether that be to the hyperscalers or to software. No one knows where the money goes next, and anyone claiming they do is in theory a trader. While momentum looks strong in the XLI, that does not mean it will continue over the next few weeks. I could even argue that financials are in more of a Goldilocks setup today than industrials are.

Top Holdings as of 7-23-26

To help visualize my holdings that would be considered old economy stocks, I have highlighted them in yellow.

Lockheed Martin (LMT) Q2 Earnings

Second quarter sales of $20.06 billion against $18.16 billion a year ago reads as 11% growth, but the prior year quarter carried $1.6 billion of reach-forward program losses and $169 million of other charges. CFO Evan Scott put underlying growth at 7% once those adjustments come out.

The stock ended the day up 10.54%, but I'm hoping it was not just due to seeing headlines of 11% revenue growth and EPS going from $1.46 to $7.94. The real story that should have been driving it up is the record backlog of $230.4 billion, on $65 billion of orders in a single quarter. This is with a backdrop narrative of a "wartime economy" that we have been discussing here on Substack. This mostly lives inside Missiles and Fire Control, where backlog nearly doubled from $46.7 billion to $87.9 billion on the back of a seven-year, $35 billion THAAD interceptor contract.

Every line of guidance moved up, with sales going to $79.75 billion to $81.75 billion, EPS to $29.95 to $30.65, and free cash flow to $7.0 billion to $7.2 billion.

The capital return picture is exactly what I expected to find after looking at Northrop Grumman's earnings on Tuesday. Lockheed bought back zero stock in the first six months of this year versus $1.25 billion in the same stretch of 2025. Dividends remain the same, with no increase. Cash is indeed going into factories, missiles, and acquisitions.

CEO Jim Taiclet gave the clearest articulation of the thesis when he told analysts the company is "investing in the manufacturing and design capabilities, before the orders come in." He also laid out the scale of what is being asked, describing government requests to triple PAC-3 MSE output, quadruple THAAD, and ramp PrSM faster than the company ever planned for.

RTX Corp Q2 Earnings

Backlog hit a record $289 billion, up 22% year over year, split $170 billion commercial and $119 billion defense. Raytheon booked $19.9 billion in the quarter for a book-to-bill of 2.42. Again, this is my opinion on why defense manufacturers led the market today. On top of that, forward guidance went up across the board, with sales to $95.0 billion to $96.0 billion, organic growth to 8% to 9% from 5% to 6%, and adjusted earnings per share to $7.10 to $7.25

To no surprise, we have zero buybacks in the first half against $50 million last year, and no dividend increase. CEO Chris Calio listed priorities as investing in the business first, the dividend second, and debt reduction third. Calio pointed to a 2027 base budget request of $1.1 trillion, roughly a 25% increase year over year with bipartisan support, and said the five framework agreements RTX signed this year carry significant demand over the next ten years and are not in the backlog yet.

These defense stocks are trading on forward guidance, backlog, and internal investments in what many may call a "wartime economy." This is the narrative today for this quarter of earnings, and it doesn't mean it will always be this way.

The Rigatoni Capital Podcast is now on YouTube, Spotify, Apple Podcasts, Amazon Music, and iHeart Radio.

Union Pacific Corp Q2 Earnings

Now if we are really on the cusp of a wartime economy, then we would need to rely on the transports as a leading indicator, like the ole Dow Theory. Industrials cannot be hot without transports, and usually we see it in the transports run first. When trucking capacity gets this tight, freight goes to the rails, and Union Pacific is the largest market cap in the group collecting the benefit.

This is why I wanted a separate section for Union Pacific. As boring as the company is, it is more important to pay attention to than ever.

What's so impressive is how well the stock has traded while we have the overhang of an unresolved $85 billion acquisition of Norfolk Southern. A stock hitting all-time highs amid this much regulatory uncertainty tells us the operating business is being valued on its own merits.

Union Pacific reported operating revenue of $6.86 billion and freight revenue of $6.52 billion, both up 12%, but this mostly had to do with diesel doing the work. CFO Jennifer Hamann attributed 750 basis points of the freight revenue increase to fuel surcharge, which grew roughly $460 million as the average price per gallon went from $2.42 to $3.86.

Year to date operating cash flow rose 21% to $5.52 billion and free cash flow reached $1.81 billion against $1.11 billion, though cash income taxes paid dropped to $236 million from $785 million, so a meaningful piece of that improvement is timing.

The quarterly dividend went to $1.38 from $1.34 for a 3% raise, and the buyback is paused entirely for the merger.

Management raised the 2026 outlook to high-single-digit reported earnings per share growth, mostly due to roughly 200 industrial development projects in the pipeline and international intermodal expected to turn positive once it laps last year's tariff volatility in August.

On the merger, CEO Jim Vena was blunt, saying "the merger's going to close" and calling it too compelling for the country to block, which is no surprise here.

Final Thoughts

I am happy to announce to you all that I think industrials earnings week is officially over here at Rigatoni Capital, except for Caterpillar, which is scheduled to report on August 4th.

While we had an up day in industrials, again I want to emphasize that it would be too soon to get excited. We would need to see a continuation for a few more weeks. Let's all agree on one thing, that between ISM, initial jobless claims at an unexpected low of 187,000, and the recycling of Middle East war postures, there is a reshoring of manufacturing here in the West that could be the cusp of a mini wartime economy under Trump's second term. Something we kind of saw coming, minus the actual US-Iran war.

The stocks discussed above are all holds in the portfolio, and no adjustments are being made.

Thank you for reading, and please subscribe for more content on long-term investing!

In Case You Missed:

References:

Lockheed Martin Corporation, Q2 2026 Financial Results news release, Exhibit 99.1, July 23, 2026.

Lockheed Martin Corporation, Q2 2026 earnings conference call transcript, July 23, 2026.

RTX Corporation, Q2 2026 Results news release, Exhibit 99, July 23, 2026.

RTX Corporation, Q2 2026 earnings conference call transcript, July 23, 2026.

Union Pacific Corporation, Q2 2026 earnings call transcript, July 23, 2026.

Union Pacific Corporation, Q2 2026 earnings press release and 2026 outlook, July 23, 2026.

Disclaimer: This blog is for informational purposes only and does not constitute financial advice. All opinions are my own, and I am not a financial advisor. The information provided reflects my personal views and is intended to encourage discussion and thought among readers. Investments involve risk, including the loss of principal, and past performance is not indicative of future results. Always conduct your own research or consult with a qualified professional before making any financial decisions.

Full Disclosure: I am long Northrop Grumman (NOC), which is 0.65% of my portfolio, up roughly +65%. I am also long Chubb Limited (CB), which is 1.68% of my holdings and up roughly +105%.

Today I had the wonderful pleasure of speaking with Kanaan at on Substack, which you can find the full conversation here:

With the market still deciphering if we are in an AI bubble, or close to one, there are many uncertainties and risks in tech as a whole that include the hyperscalers, software, and semis. The conversation ended on something I have been thinking is a theme in my own personal investing journey, and that is not to forget about two other sectors that could theoretically always catch some momentum, which would be financials or industrials.

This is exactly why I want to spend time going over what you may think of as "boring" companies and their quarterly earnings. You know the names, Travelers Insurance last week, old economy banks like JPMorgan, and this week we have the industrials like Northrop Grumman, and a few financials reporting like Chubb Insurance.

These companies are so boring versus what is going on in the AI tech space, it will literally put you to sleep. But that is the entire point, these are the hedges that I have in mind for someone that is a long-term buy and hold investor. This is the type of risk management or diversification that any long-term investor should be exposed to all the time, not just when there is so much uncertainty around Artificial Intelligence frothiness.

I thought the topic was so important during what us home gamers call "industrials week" that it was worth jumping on a quick Rigatoni Capital podcast today (Link Here) to follow through on my great conversation with .

Before we get on with Northrop Grumman and Chubb earnings, let me give you a brief background of why I subscribe to on Substack. While Kanaan lives half a world away from me, this allows her to have geopolitical insight that I could never have. And yet, we still have converging and overlapping feelings when it comes to market trends, tech investing, and a long-term vision on what we are all chasing, which is better than average returns.

On top of that, our writing styles converge as well, as she has a broader macro understanding that helps her make conclusions on micro positions. By the way, Kanaan is up 1,000% on Micron, if I understood her correctly. Even though I get the sense that Kanaan is more active in trading than I am in my own portfolio, she is not afraid of very large total gains if the fundamentals remain intact.

When I first discovered her writing on Substack, I was hooked, and I still look forward to rolling over in bed in the morning to a fresh newsletter in my email reminding readers that the US-Iran war is nowhere near over and to ignore any political headlines that say otherwise. This is the type of macro insight you will get from her, alongside her individual holdings and the thinking behind them.

The best part of being part of the Substack finance community is when you can meet all sorts of financial writers and investors across the globe, connect with them on a Zoom call for a quick conversation, and then record it for the world to hear. Now, could that have been this easy 20 years ago? How fortunate are we, really? Thank you again for meeting me today, it was a pleasure!

Northrop Grumman (NOC) remains a hold:

Northrop’s headline number looked soft, and in this environment you cannot afford that. However, the stock only ended down ~2.23% today. Net earnings fell 7% YoY to $1.1 billion, and diluted earnings per share came in at $7.68 against $8.15 YoY.

The decline was mostly an artifact of last years quarter that was padded by a gain from selling the training services business, worth a bit over a dollar a share. If we strip that out, operating income still fell but only 5%. Taxes also did a lot of the heavy liftning this quarter, with the effective tax rate dropping to 6.3% from 17.7% on a “remerasurement of some uncertain tax positions after developments with the IRS, which alone saved $179 million.” So of course, we don’t want to see a company leaning heavy on a tax line to cushion any earnings report.

What actually pressured the operations were two programs. Space Systems took a charge on its GEM 63XL rocket motor program for higher projected costs and materials, which pulled that segment's margin down to 8.6% from 10.6%. Defense Systems took a charge on its Stand-in Attack Weapon program during qualification testing, which knocked that segment's margin all the way down to 7.5% from 12.7%. This is essentially the entire margin story this quarter. If we pull the missile charge out of Defense Sytems, the rest of that segment ran near 11%.

On a positive note, net awards were $20 billion in the quarter, a book-to-bill of 1.84 times, and backlog hit a record $104.7 billion, up 9% since year-end and 17% from a year ago. Sentinel, the company's program to replace the Air Force's aging intercontinental ballistic missiles, alone added $7.6 billion to backlog. Management raised full-year guidance on both sales and earnings, taking sales to roughly $44 billion at the midpoint and lifting the EPS range by $1.20.

The obvious headline risk you will read in mainstream news is the midterms, and the idea that Democrats will pressure defense budgets. Unless you are a short-term trader, this is not a reason to avoid the sector. The base budget that funds defense manufacturers has more bipartisan support than ever. In other words, the war machine in the US does not stop regardless of which party is in control, sadly.

The missile squeeze, and Trump’s contradiction

In January, Trump ordered defense contractors to stop buybacks and dividends until they sped up weapons production. Then in June, Trump invoked the Defense Production Act to fix the solid rocket motor shortage himself, effectively admitting the slowness is a structural supply problem and not the primes hoarding cash.

Northrop is one of only four companies qualified to make those motors, and it is spending to build capacity, which is why buybacks nearly vanished this year, $68 million in the first half against $891 million a year earlier. However, Northrop did raise its dividend in May, to $2.47 from $2.31.

Chubb Limited (CB): the long-term compounder of wealth

I know how excited you all get for Chubb Insurance earnings, but unfortunately the earnings call is not until tomorrow morning, so we will just go over the report's key highlights.

Net income fell slightly, to $2.85 billion from $2.97 billion a year ago, and per-share net income was basically flat at $7.30. But that comparison is distorted by large investment gains in last year’s quarter that inflated the prior number. Strip those out and look at core operating income, which is the cleaner measure of the actual insurance business, and it rose 14.6% to $2.84 billion, or up 18.2% per share. The operating business got materially stronger even as the headline drifted backward.

Chubb also reported a combined ratio of 83.3%, which means Chubb is keeping roughly sixteen cents for every premium dollar as underwriting profit. They did this while premiums in North America commercial actually fell 2.3%, because CEO Evan Greenberg is deliberately walking away from soft property pricing. In his words, the company "will not underwrite knowingly at a loss," and he expects the growth penalty they are paying in property to fade over time.

Chubb returned $1.37 billion to shareholders in the quarter, $979 million of it through buybacks and $395 million in dividends, and tangible book value per share. The stock remains a hold in my portfolio.

Thank you for reading, and don’t forget to subscribe for more content on long-term investing! Thank you again for having me on your podcast today!

The Rigatoni Capital Podcast is now on YouTube, Spotify, Apple Podcasts, Amazon Music, and iHeart Radio.

Top Holdings as of July 21st:

In Case You Missed:

References

• Northrop Grumman, “Second Quarter 2026 Financial Results,” news release, July 21, 2026 — investors.northropgrumman.com

• Northrop Grumman Q2 2026 earnings call transcript, July 21, 2026

• Chubb Limited, “Chubb Reports Second Quarter 2026 Results,” news release, July 21, 2026 — investors.chubb.com

• Presidential Determination 2026-15, Defense Production Act, June 11, 2026 — whitehouse.gov

• CSIS on solid rocket motor stockpiles, via Breaking Defense, June 2026 — breakingdefense.com

Disclaimer: This blog is for informational purposes only and does not constitute financial advice. All opinions are my own, and I am not a financial advisor. The information provided reflects my personal views and is intended to encourage discussion and thought among readers. Investments involve risk, including the loss of principal, and past performance is not indicative of future results. Always conduct your own research or consult with a qualified professional before making any financial decisions.