The second quarter of 2026 was a rewarding stretch for disciplined investors — and a humbling one for forecasters and leveraged speculators alike. Equity markets delivered strong returns despite navigating geopolitical conflict in the Middle East, energy-price volatility, shifting expectations for interest rates, and ongoing economic uncertainty. Once again, markets demonstrated an ability to look beyond today’s headlines and focus on tomorrow’s opportunities.
For our portfolios, the most important development of the quarter was this: the market appears to be recognizing that the great software companies we own may be significant beneficiaries of artificial intelligence. For about a year, a fashionable narrative held that AI would render enterprise software a slow growth sector. This earnings season did not end that debate, but it moved the evidence meaningfully in our direction. We detail the results below.
ECONOMIC & MARKET DYNAMICS
Profit Margins Are Elevated
The net profit margin for companies in the S&P 500 rose significantly in the first half of the year and is on pace to reach approximately 15% for the second quarter. Importantly, this is not just a technology story — multiple sectors, including financial services and industrials, reported net margins above their five-year averages. Corporate America has become more resilient to shocks that once would have triggered an inflationary jump and an economic slowdown.
Earnings Are Doing the Heavy Lifting
This year’s equity returns have been driven by companies earning more money — not by investors simply paying higher prices for the same earnings. That is the healthiest foundation a bull market can have.
The Consumer Remains Strong
U.S. household assets stand at roughly $204.5 trillion against just $21.6 trillion of liabilities, and household debt service consumes only about 11% of disposable income — near historically low levels.
Investment Spending Is Surging
The defining economic force of this cycle is capital investment in artificial intelligence infrastructure. The scale is without precedent, and the overwhelming majority of it is being funded from operating cash flow — not debt.
The five largest cloud-computing companies — the “hyperscalers”: Alphabet, Amazon, Meta, Microsoft, and Oracle. Source: company disclosures and consensus estimates.
THE MARKET IS STARTING TO COME AROUND ON SOFTWARE
For the better part of a year, the fashionable view on Wall Street was that artificial intelligence slow growth of enterprise software — that AI agents would let companies build their own workflows and hollow out demand for the platforms they run on. Quality software franchises were affected indiscriminately, including the ones we own.
We took the other side of that argument. Our view has been that the companies that own the workflow, the system of record, and the customer relationship are precisely the companies AI makes more valuable, because AI is only as useful as the enterprise data and business context it can act on. This quarter brought encouraging signs that the market is beginning to come around to that view — not because the narrative changed, but because the earnings made the old narrative harder to sustain. Consider what the software companies we own just reported:
One quarter does not settle a debate of this size, and the shift in sentiment could stall or reverse. But the direction of the evidence is consistent with a lesson we have repeated often in these letters: narratives move stocks in the short run, but earnings move them in the end.
OUR LARGEST HOLDINGS ARE COMPOUNDING
The same dynamic is playing out, at even greater scale, in our four largest holdings: Amazon, Alphabet (Google), Microsoft, and Meta. These are among the largest enterprises in the history of capitalism, and in their most recently reported quarter each of them grew revenue at a rate one would normally associate with a small, fast growing company:
Taken together, these four businesses generated more than $470 billion of revenue in a single quarter, growing at a blended rate of roughly 22%. Just as striking is what is already contracted but not yet recognized: AWS ended the quarter with a $496 billion backlog, Google Cloud with $514 billion, and Microsoft with $678 billion of signed customer commitments — nearly $1.7 trillion of future revenue, in aggregate, already committed by customers.
This is what we mean when we say the AI buildout is being driven by demand, not hope.
OPEN MODELS ARE A TAILWIND
A related shift is underway in the AI landscape itself: free, publicly available AI models — which anyone can download and run on their own systems — are being adopted at an accelerating pace by enterprises around the world.
Here is why this matters so much for our portfolio. As these free models multiply, raw AI capability is fast becoming a commodity — the cost of using it keeps falling. When a key ingredient gets cheaper, the profits flow to the businesses around it — and those are exactly the businesses we own.
Our hyperscalers win on computing power. An open model is free to download, but it is not free to run. Every enterprise that deploys one still needs vast amounts of secure, reliable computing power — and rents it from AWS, Microsoft Azure, and Google Cloud. Cheaper intelligence expands the number of economically viable AI applications, which expands total demand for computing power. The cloud growth rates in the table above — 37%, 43%, 82% — are that dynamic showing up in reported numbers.
Our software companies win on margins and moats. Falling model costs turn AI features from an expense into a profit engine for ServiceNow, Salesforce, and SAP: they can embed intelligence into their platforms ever more cheaply while charging for the business outcomes it delivers. And because open models are available to everyone, they confer an advantage on the companies that own what models cannot replicate — proprietary enterprise data, entrenched workflows, distribution, and customer trust.
Meta wrote the playbook. Meta open-sourced its own models precisely because commoditizing the model layer steers the industry toward its strengths — unmatched distribution and the world’s largest advertising dataset. Its 28% revenue growth this quarter suggests the strategy is working.
In short: free AI models make intelligence cheaper and increase demand for everything around it — and our portfolio owns the toll roads on both sides.
THE REST OF YOUR PORTFOLIO IS WORKING TOO
Because this letter has spent so much time on technology, we want to be clear about something: your portfolio is not a technology portfolio, and the strength this quarter extended well beyond it.
More than a third of your portfolio sits in businesses that have little to do with artificial intelligence and everything to do with steady compounding. Some are global consumer brands — in nicotine alternatives, beauty, and luxury goods — built on decades of pricing power and customer loyalty; our largest holding among them passed $11 billion of quarterly revenue for the first time in its history this quarter, with adjusted earnings per share up 15% from a year ago. Others form the financial plumbing of the world’s markets: the payment networks that settle card transactions, the exchanges, the credit-rating and index providers, the risk and insurance advisers, and the payroll processors. These companies collect a small fee on an enormous and still-growing volume of everyday economic activity. They raised revenue and widened margins again this quarter — one of them now operates at a 56% profit margin, a level very few enterprises of any kind achieve.
Our health care holdings supply the instruments and diagnostic testing that human and veterinary medicine depend on, rather than the drugs themselves. Both grew revenue roughly 10% and both raised their guidance for the full year — and they did it in a quarter when health care was the only sector of the S&P 500 whose earnings declined industry-wide.
None of these are exciting businesses in the way an AI story is exciting. They are toll collectors: essential, entrenched, highly profitable, and largely indifferent to the mood of any given month. That is precisely why we own them alongside the growth companies, and it is why a violent few weeks in AI-related shares does not determine your year.
THE JULY UNWIND: VOLATILITY, NOT RISK
Beginning in late June, markets saw a rush for the exits by speculators who had piled into the same AI-related chip trades — nowhere more so than in the high-flying memory companies, none of which we own. The declines in these names, from their late-June highs to their subsequent lows, tell the story of just how much speculation had built up:
What was most telling about this sell-off was that it had almost nothing to do with fundamentals. Shares of many companies dropped sharply immediately after reporting earnings in which they significantly beat expectations and raised guidance. The selling was mechanical, not informational: a confluence of global deleveraging events, including forced selling by investors in funds that use borrowed money to double or triple their market bets and, in the U.S., the collapse of a prominent hedge fund that had reportedly borrowed four dollars for every dollar of its own capital to bet on AI stocks. The unraveling of these stocks — and others positioned similarly — produced a big unwind. We own none of the speculative names at the center of this.
MR. MARKET, REVISITED
Nearly eighty years ago, Benjamin Graham — the father of value investing and Warren Buffett’s teacher — created an allegory to explain how investors should think about the stock market’s daily mood swings. In his classic The Intelligent Investor, Graham asked readers to picture the entire stock market as a single business partner, and Buffett has retold the story ever since. July gave us a textbook illustration of it. The parable goes like this:
Imagine that you own a successful business with a partner named Mr. Market. Every day, Mr. Market knocks on your door and offers either to buy your share of the business or to sell you his. The unusual thing about Mr. Market is that his mood changes constantly. On some days he is wildly optimistic and offers an extremely high price. On other days — like several days this July — he becomes deeply pessimistic and offers to sell at a bargain-basement price because he fears disaster.
The key insight is that you are under no obligation to accept Mr. Market’s offer. As Warren Buffett has often observed, the market is there to serve you, not to guide you. Successful investing comes from owning outstanding businesses over the long term — not from following Mr. Market’s ever-changing emotions.
OUR PRINCIPLES FOR BUILDING LASTING WEALTH
Successful investing is about discipline, not prediction. Long-term wealth is built through a well-designed investment process rather than trying to identify the next winning stock or market trend. A financial plan becomes the GPS system guiding all investing decisions.
Behavior matters more than market timing. Investors who remain calm during periods of volatility are far more likely to achieve their long-term financial goals.
Costs matter. Trading charges, income taxes, and the cost of being out of the market all quietly erode returns.
Inflation quietly erodes purchasing power. Excessive cash holdings may feel safe but often lose real value over time as the cost of living rises.
Distinguish risk from volatility. Risk, as we define it, is the possibility of permanent loss of capital. Volatility is simply the manifestation of the market’s pricing mechanism.
Every investment should have a purpose. Different financial goals require different strategies, time horizons, and levels of risk. Asset allocation drives long-term results.
Focus on value, not simply low fees. Quality advice, tax planning, estate planning, appropriate insurance coverage, and disciplined portfolio management add significant long-term value and are critical elements of risk mitigation.
LOOKING AHEAD
We believe the second half of 2026 will continue to reward owners of financially strong, well-managed businesses. The companies in your portfolio are growing revenue at rates rarely seen at their scale, funding the largest infrastructure buildout in history from their own cash flows, and converting artificial intelligence from a promise into earnings — quarter after quarter. Just as important, these are durable, conservatively financed businesses — enormous cash reserves, little debt, and investments funded from their own profits — chosen so your portfolio can weather difficult markets without putting the wealth you have built at risk. Short-term volatility of the kind we saw in July will surely visit us again. When it does, we intend to treat it as we always have: as the price of admission for long-term returns, and occasionally as a gift.
Thank you for the trust you have placed in us. It is a responsibility we take seriously every day.
“The intelligent investor is a realist who sells to optimists and buys from pessimists.”
— Benjamin Graham
Respectfully,
Jonathan Dash CEO & Chief Investment Officer
Forward-Looking Statement Disclosure
The discussion of our investments represents the views of the Company’s portfolio manager at the time of this report and is subject to change without notice. References to individual securities are for informational purposes only and should not be construed as recommendations to purchase or sell individual securities. As portfolio managers, one of our responsibilities is to communicate with clients in an open and direct manner. Insofar as some of our opinions and comments in our letters to our partners are based on current management expectations, they are considered “forward-looking statements,” which may or may not be accurate over the long term. While we believe we have a reasonable basis for our comments and we have confidence in our opinions, actual results may differ materially from those we anticipate. You can identify forward-looking statements by words such as “believe,” “expect,” “may,” “anticipate,” and other similar expressions when discussing prospects for particular portfolio holdings and/or the markets, generally. We cannot, however, assure future results and disclaim any obligation to update or alter any forward-looking statements, whether as a result of new information, future events, or otherwise. Further, information provided in this report should not be construed as a recommendation to purchase or sell any particular security.
MARKET COMMENTARY
by Jonathan Dash
Good Evening, May 17, 2023
To Our Partners,
The second quarter of 2026 was a rewarding stretch for disciplined investors — and a humbling one for forecasters and leveraged speculators alike. Equity markets delivered strong returns despite navigating geopolitical conflict in the Middle East, energy-price volatility, shifting expectations for interest rates, and ongoing economic uncertainty. Once again, markets demonstrated an ability to look beyond today’s headlines and focus on tomorrow’s opportunities.
For our portfolios, the most important development of the quarter was this: the market appears to be recognizing that the great software companies we own may be significant beneficiaries of artificial intelligence. For about a year, a fashionable narrative held that AI would render enterprise software a slow growth sector. This earnings season did not end that debate, but it moved the evidence meaningfully in our direction. We detail the results below.
ECONOMIC & MARKET DYNAMICS
Profit Margins Are Elevated
The net profit margin for companies in the S&P 500 rose significantly in the first half of the year and is on pace to reach approximately 15% for the second quarter. Importantly, this is not just a technology story — multiple sectors, including financial services and industrials, reported net margins above their five-year averages. Corporate America has become more resilient to shocks that once would have triggered an inflationary jump and an economic slowdown.
Earnings Are Doing the Heavy Lifting
This year’s equity returns have been driven by companies earning more money — not by investors simply paying higher prices for the same earnings. That is the healthiest foundation a bull market can have.
The Consumer Remains Strong
U.S. household assets stand at roughly $204.5 trillion against just $21.6 trillion of liabilities, and household debt service consumes only about 11% of disposable income — near historically low levels.
Investment Spending Is Surging
The defining economic force of this cycle is capital investment in artificial intelligence infrastructure. The scale is without precedent, and the overwhelming majority of it is being funded from operating cash flow — not debt.
The five largest cloud-computing companies — the “hyperscalers”: Alphabet, Amazon, Meta, Microsoft, and Oracle. Source: company disclosures and consensus estimates.
THE MARKET IS STARTING TO COME AROUND ON SOFTWARE
For the better part of a year, the fashionable view on Wall Street was that artificial intelligence slow growth of enterprise software — that AI agents would let companies build their own workflows and hollow out demand for the platforms they run on. Quality software franchises were affected indiscriminately, including the ones we own.
We took the other side of that argument. Our view has been that the companies that own the workflow, the system of record, and the customer relationship are precisely the companies AI makes more valuable, because AI is only as useful as the enterprise data and business context it can act on. This quarter brought encouraging signs that the market is beginning to come around to that view — not because the narrative changed, but because the earnings made the old narrative harder to sustain. Consider what the software companies we own just reported:
One quarter does not settle a debate of this size, and the shift in sentiment could stall or reverse. But the direction of the evidence is consistent with a lesson we have repeated often in these letters: narratives move stocks in the short run, but earnings move them in the end.
OUR LARGEST HOLDINGS ARE COMPOUNDING
The same dynamic is playing out, at even greater scale, in our four largest holdings: Amazon, Alphabet (Google), Microsoft, and Meta. These are among the largest enterprises in the history of capitalism, and in their most recently reported quarter each of them grew revenue at a rate one would normally associate with a small, fast growing company:
Taken together, these four businesses generated more than $470 billion of revenue in a single quarter, growing at a blended rate of roughly 22%. Just as striking is what is already contracted but not yet recognized: AWS ended the quarter with a $496 billion backlog, Google Cloud with $514 billion, and Microsoft with $678 billion of signed customer commitments — nearly $1.7 trillion of future revenue, in aggregate, already committed by customers.
This is what we mean when we say the AI buildout is being driven by demand, not hope.
OPEN MODELS ARE A TAILWIND
A related shift is underway in the AI landscape itself: free, publicly available AI models — which anyone can download and run on their own systems — are being adopted at an accelerating pace by enterprises around the world.
Here is why this matters so much for our portfolio. As these free models multiply, raw AI capability is fast becoming a commodity — the cost of using it keeps falling. When a key ingredient gets cheaper, the profits flow to the businesses around it — and those are exactly the businesses we own.
In short: free AI models make intelligence cheaper and increase demand for everything around it — and our
portfolio owns the toll roads on both sides.
THE REST OF YOUR PORTFOLIO IS WORKING TOO
Because this letter has spent so much time on technology, we want to be clear about something: your portfolio is not a technology portfolio, and the strength this quarter extended well beyond it.
More than a third of your portfolio sits in businesses that have little to do with artificial intelligence and everything to do with steady compounding. Some are global consumer brands — in nicotine alternatives, beauty, and luxury goods — built on decades of pricing power and customer loyalty; our largest holding among them passed $11 billion of quarterly revenue for the first time in its history this quarter, with adjusted earnings per share up 15% from a year ago. Others form the financial plumbing of the world’s markets: the payment networks that settle card transactions, the exchanges, the credit-rating and index providers, the risk and insurance advisers, and the payroll processors. These companies collect a small fee on an enormous and still-growing volume of everyday economic activity. They raised revenue and widened margins again this quarter — one of them now operates at a 56% profit margin, a level very few enterprises of any kind achieve.
Our health care holdings supply the instruments and diagnostic testing that human and veterinary medicine depend on, rather than the drugs themselves. Both grew revenue roughly 10% and both raised their guidance for the full year — and they did it in a quarter when health care was the only sector of the S&P 500 whose earnings declined industry-wide.
None of these are exciting businesses in the way an AI story is exciting. They are toll collectors: essential, entrenched, highly profitable, and largely indifferent to the mood of any given month. That is precisely why we own them alongside the growth companies, and it is why a violent few weeks in AI-related shares does not determine your year.
THE JULY UNWIND: VOLATILITY, NOT RISK
Beginning in late June, markets saw a rush for the exits by speculators who had piled into the same AI-related chip trades — nowhere more so than in the high-flying memory companies, none of which we own. The declines in these names, from their late-June highs to their subsequent lows, tell the story of just how much speculation had built up:
What was most telling about this sell-off was that it had almost nothing to do with fundamentals. Shares of many companies dropped sharply immediately after reporting earnings in which they significantly beat expectations and raised guidance. The selling was mechanical, not informational: a confluence of global deleveraging events, including forced selling by investors in funds that use borrowed money to double or triple their market bets and, in the U.S., the collapse of a prominent hedge fund that had reportedly borrowed four dollars for every dollar of its own capital to bet on AI stocks. The unraveling of these stocks — and others positioned similarly — produced a big unwind. We own none of the speculative names at the center of this.
MR. MARKET, REVISITED
Nearly eighty years ago, Benjamin Graham — the father of value investing and Warren Buffett’s teacher — created an allegory to explain how investors should think about the stock market’s daily mood swings. In his classic The Intelligent Investor, Graham asked readers to picture the entire stock market as a single business partner, and Buffett has retold the story ever since. July gave us a textbook illustration of it. The parable goes like this:
OUR PRINCIPLES FOR BUILDING LASTING WEALTH
LOOKING AHEAD
We believe the second half of 2026 will continue to reward owners of financially strong, well-managed businesses. The companies in your portfolio are growing revenue at rates rarely seen at their scale, funding the largest infrastructure buildout in history from their own cash flows, and converting artificial intelligence from a promise into earnings — quarter after quarter. Just as important, these are durable, conservatively financed businesses — enormous cash reserves, little debt, and investments funded from their own profits — chosen so your portfolio can weather difficult markets without putting the wealth you have built at risk. Short-term volatility of the kind we saw in July will surely visit us again. When it does, we intend to treat it as we always have: as the price of admission for long-term returns, and occasionally as a gift.
Thank you for the trust you have placed in us. It is a responsibility we take seriously every day.
— Benjamin Graham
Respectfully,
Jonathan Dash
CEO & Chief Investment Officer
Forward-Looking Statement Disclosure
The discussion of our investments represents the views of the Company’s portfolio manager at the time of this report and is subject to change without notice. References to individual securities are for informational purposes only and should not be construed as recommendations to purchase or sell individual securities. As portfolio managers, one of our responsibilities is to communicate with clients in an open and direct manner. Insofar as some of our opinions and comments in our letters to our partners are based on current management expectations, they are considered “forward-looking statements,” which may or may not be accurate over the long term. While we believe we have a reasonable basis for our comments and we have confidence in our opinions, actual results may differ materially from those we anticipate. You can identify forward-looking statements by words such as “believe,” “expect,” “may,” “anticipate,” and other similar expressions when discussing prospects for particular portfolio holdings and/or the markets, generally. We cannot, however, assure future results and disclaim any obligation to update or alter any forward-looking statements, whether as a result of new information, future events, or otherwise. Further, information provided in this report should not be construed as a recommendation to purchase or sell any particular security.