The $1 Trillion Supercycle Hidden in Plain Sight | Joseph Shaposhnik
What this covers
On this episode of Excess Returns, Matt Zeigler and Bogumil Baranowski speak with Rainwater Equity ETF portfolio manager Joseph Shaposhnik about how long-term investors should think about markets in an era defined by geopolitical shocks, AI disruption, and unprecedented capital investment cycles. The conversation explores how disciplined investors can stay focused on durable businesses and long-term free cash flow rather than reacting to short-term headlines. Joseph explains how his team evaluates companies during major events, why the AI boom may create both massive disruption and opportunity, and where he believes the most attractive investment opportunities exist today.
Topics covered in this episode
• Why most macro headlines and geopolitical events rarely have lasting impacts on great businesses • How long-term investors should analyze conflicts and market shocks without overreacting • The defense spending supercycle and why aerospace and defense may benefit from rising geopolitical tensions • How Joseph evaluates the AI investment cycle across semiconductors, software, and hyperscalers • Why semiconductor companies may offer a lower-risk way to benefit from AI growth • The risks created by massive AI infrastructure CapEx and concentration around specific AI models • Why some software companies may face significant disruption from AI tools and LLMs • How AI could reshape business models that rely on packaging public or commoditized data • The potential rotation from the Magnificent Seven to the other 493 companies in the S&P 500 • Why capital intensity may change the long-term attractiveness of some technology companies • The role of management quality and capital allocation in navigating technological disruption • Fragile vs anti-fragile business models in an AI-driven economy • Where AI may create unexpected winners across industrial and traditional industries • Why long-term investors should still prioritize durable cash flow compounding businesses
Timestamps
00:00 Introduction and why most headlines have limited long-term impact on businesses 02:00 How experienced investors think about geopolitical shocks and market headlines 04:00 Defense spending tailwinds and the aerospace and defense supercycle 06:45 How investors should react when major market news breaks 11:10 How Joseph evaluates the AI boom and which companies benefit most 14:15 The case for opportunities outside the Magnificent Seven 17:15 How rising AI CapEx is changing the economics of major tech companies 21:25 Why hyperscalers face increasing concentration risk 23:00 Why semiconductor suppliers may be the best positioned AI investments 27:15 Why Joseph reduced exposure to software companies 33:00 The importance of learning organizations and adaptive management teams 37:00 AI, labor markets, and whether high-income jobs face disruption 41:00 Fragile vs anti-fragile companies in the age of AI 46:00 Where AI could create unexpected business winners 52:00 How great management teams adapt during technological disruption 57:00 How AI may accelerate entrepreneurship and innovation 59:00 Why investors should remain focused on sustainable cash flow 01:02:00 What the next generation of long-term compounders may look like
Source description (no synthesized summary yet).
Long-term portfolio success depends on identifying durable businesses with strong free cash flow generation and capable management that can navigate disruption—particularly AI—rather than chasing short-term headlines or betting on specific technology winners.
- Most headlines have no meaningful long-term impact on companies; business durability and compounding power drive returns
- AI creates winners, losers, and disruption across sectors, but concentrated bets on single LLM winners carry unacceptable risk; supplier diversification is safer
- Management quality—specifically leaders with learning cultures, capital allocation discipline, and proven ability to pivot—matters more than sector or trend positioning
This asset isn't compiled yet
You're seeing its claims, ranked. Compile it to build the argument threads, weight them, and check each claim against your library — the full view.
Railroad companies laid the foundations for huge infrastructure investments but didn't capture the value—instead, connected businesses, cities, and companies captured the value, suggesting a pattern where the question is not who builds the infrastructure but who benefits from and uses it.
“You know, Joe, I'm thinking about the other technological evolutions and the companies that laid the foundations, the big investments were not necessarily the ones that would capture value, right? We had railroads, but it wasn't railroads that at the end of the day captured the value. It was the businesses, the cities, the companies that got connected with the railroads.”
Asset-light, capital-light businesses with high recurring compounding cash flow are becoming asset-heavy, capital-intensive businesses as hyperscalers invest massively in AI infrastructure, which is a transformation that makes them less attractive and more difficult investments.
“these businesses, which had been capital light, are unlikely to become capital light uh in the near future. And so, in some ways that makes them less attractive. They're just much more capital intensive businesses than they have been for most of their histories.”
Reduced time to service customer requests, debug products, develop new features, and market new ideas will dramatically improve service quality and reduce barriers to entry for software businesses.
“the ability to develop new products, the develop the ability to debug existing products, the ability to service the customer, uh to receive the input and then know how to respond to the customer, I think, uh, will dram- will be the time to do that will dramatically drop, and I think that will make for a really wonderful world, uh, where service will be much better. Um, as you said, the rate of of iterating on new ideas will dramatically increase, and I think that that makes for, uh, a better world, a a richer pie, uh, for companies to go after. I think it it certainly reduces the barriers to entry in a lot of businesses”
Most headlines don't have meaningful long-term impacts on companies because by their very definition they are relatively short-term in nature and go away; what matters is the strength of businesses, durability of franchises, and their ability to keep compounding.
“Most headlines don't have meaningful long-term impacts on companies because most headlines, by their very definition, are relatively short-term in nature um and they go away. It all comes back to um the strength of the businesses, the durability of the franchises, and their ability to keep compounding.”
Great businesses can power through macro shocks and geopolitical dynamics, but it is important to think about what is going on in the world as you manage a portfolio because events could impact specific businesses and opportunities.
“generally speaking great businesses can power through those those dynamics. But I think that it is important to think about what is going on in the the world as you manage a portfolio and we certainly do. So, we're not an investor that just says, well, we just focus on the long run and we don't think about how events could impact the portfolio and our businesses.”
There are massive beneficiaries of AI, companies in the middle experiencing muddled impact, and companies that will be clearly negatively impacted by the AI cycle.
“in general, you know, there are massive beneficiaries of AI. There are companies in the middle, and there are companies that will be clearly negatively impacted uh by what is going on in the in in uh in the in this in this cycle.”
Technology sector history indicates that it's closer to a winner-take-all market than a four or five horse race, which is why those investing massive money into the future are betting on specific customer duration and specific customers who will win.
“The techno the history of the technology sector indicates that it's closer to a winner-take-all market than it is to a four or five-horse race, which is why those that are investing to your point so much money into the future are making making a bet on the duration of of of the cash flow that's going to be generated by their customers and on specific customers who will win.”
Productivity-enhancing technologies have historically expanded markets over time, which is a pattern AI could follow, suggesting a bull case that AI will make businesses much more efficient across markets and generate significantly better returns.
“I can certainly see it. I think that the the bull case is that we'll become much more efficient across all major markets, across all all businesses and that will generate significantly better returns across most businesses.”
Companies must invest ahead of productivity improvements to capture them, creating an investment period where returns are uncertain, and the payoffs may be several years out before they materialize.
“It's early though. What we're seeing out of this out of some of the software companies is that you have to invest ahead of that. And so to the extent companies are playing catch-up to get where they need to get to there is going to be upfront investment uh, get there. That means AI tools, that means capacity, that probably means more engineers. So, it's company dependent and the payoffs uh, maybe some years out before you see the returns.”
Investors must remember that the ultimate goal is finding sustainable cash flows with decent returns in businesses, which should be weighed equally against companies trying to figure out what products should look like day-to-day due to consumer willingness to defect to competition.
“I think we sometimes get lost with the excitement of the innovation, both as consumers and investors, and I'm thinking that we kind of have to like shake and wake up for a second, because at the end of the day, we're trying to find cash flows, right? And if you can get sustainable cash flows with decent returns in a business that's almost absolutely not exposed to AI, which is hard to do these days, but if there is one, this is this is as attractive not attractive as a company that's trying to figure out how the product should be looking, you know, minute to minute, day to day, because the consumer is willing to go to the competition, right?”
When evaluating companies, the fund asks: How are you incorporating AI into what you do? What's your plan for becoming more efficient and serving customers more effectively with available technology? This is a conversation topic but not the top-down focus.
“one of those questions is, How are you incorporating AI into what you do? What's your plan for becoming more efficient in serving your customer more effectively with the technology that is available today and will be available in the not-too-distant future. So, it's certainly a a topic of conversation, but we're not we're not building a portfolio with a top-down view of trying to find direct beneficiaries of the technology.”
Less activity during aggressive headlines and many headline events generally yields better portfolio performance than reactive trading; the key is staying focused on company investment theses and understanding whether anything has meaningfully changed.
“In my experience, uh less activity during headline during uh aggressive headlines or many many headlines coming out, less activity generally uh has yielded better performance or good performance uh from my portfolio. So I'm always cognizant of that, uh not overreacting to the headlines, just going back to the companies, going back to our investment thesis on these businesses, and trying to understand whether anything has meaningfully changed.”
Management quality and leadership—particularly those with history of successfully navigating transitions like on-premise to SaaS, making successful capital allocation pivots, and demonstrating long-term thinking—is one of the best predictors of success in an AI-disrupted environment.
“what you're trying to do is you're trying to invest with people that have a history of doing that. They have a history of um having gone through the transition from on-premise to SaaS. They've done that successfully. They've pivoted capital allocation in the past successfully when they ran into a roadblock or an end market where they were saturated.”
Hock Tan at Broadcom had been acquiring semiconductor businesses for 15 years as CEO, attempted a large Qualcomm acquisition that was blocked, pivoted to software acquisitions in an underexplored end market, and then pivoted again to XPUs, which are now critical backbone infrastructure for hyperscalers—demonstrating excellent capital allocation and adaptability.
“what comes to mind is is Hock Tan at Broadcom where he'd been acquiring businesses in the semiconductor industry for 15 years as CEO of of what was called Avago and then became uh Broadcom and he tried to make his largest acquisition, which is Qualcomm and it was blocked by the government. And he and his team made a decision at that point that pivoting the software and making software acquisitions was a better use of their capital and was an end market where they'd done no acquisitions and so there was opportunity and they made that pivot. And then some years after that they pivoted to making XPUs.”
Previously thought-to-be very proprietary data and contributory software services that were triple-A highly durable assets could be disrupted by AI and LLM advances.
“But what we began to believe were businesses that we thought had contributory data that were the only providers of contributory data as an example at some point those contributory services could become disrupted. And for the longest time those were thought to be believed uh triple-A highly durable assets.”
There is significant uncertainty about whether software will continue to be one of the highest-quality business models in the world, as it has been thought to be for the longest time.
“For the longest time, software was thought to be one of the highest-quality business models out there in the world. And I think that there's significant uncertainty about whether that will continue to be the case.”
The better route for capturing AI upside with less risk is investing in semiconductor suppliers of picks and shovels who serve multiple LLMs and hyperscalers rather than betting on specific LLM winners, because suppliers have more diversified customer bases and healthier access to capital.
“I think that where the opportunity to be, uh, to take on less risk and benefit from, uh, what is a very, very strong cycle, and certainly a secular growth end market, is on the semi side where you can you can invest in businesses that have a customer base that is far more diversified uh a customer base that is generally healthier uh and has a greater access to capital uh and has less concentrated sets of capital investors in it.”
The Buffett 1987 letter quote: 'Experience indicates that the best business returns are usually achieved by companies that are doing something quite similar today to what they were doing 5 or 10 years ago...a business that constantly encounters major change also encounters many chances for major error...Such a franchise is usually the key to sustained high returns.' This principle drove the software exit.
“Experience indicates that the best business returns are usually achieved by companies that are doing something quite similar today to what they were doing 5 or 10 years ago. That is no argument for managerial complacency. Businesses have opportunities to improve service product lines, manufacturing techniques, and the like, and obviously these opportunities should be seized. But a business that constantly encounters major change also encounters many chances for major error. Therefore, economic terrain that is forever shifting violently is ground on which it is difficult to build a fortress-like business franchise. Such a franchise is usually the key to sustained high returns.”
A simple innovation like interacting with CRM using natural language instead of menu clicking has been a game-changer, and one can only imagine 5 years out how much better lives will be and the kinds of businesses that will be created and existing businesses improved.
“I'll give you a very simple innovation which has been game-changer for us is uh just the ability to interact with your CRM using natural language is the most basic of of innovations, you know, before you you had to go through and and uh click through all kinds of menus to to put somebody into your CRM or to to to to edit your CRM or or add somebody uh to your database. Now now you can just go to Claude and instruct Claude to to do X, Y, and Z and it does it. And that's incredibly valuable.”
Most software businesses are top-heavy with management making decisions at top layer then sending down to developers/engineers, making them poorly positioned for AI disruption era.
“most software businesses most businesses in general are very top-heavy with management making decisions kind of at the the top layer and then sending that down to the developers and the engineers. And so in an era of of uh AI and and AI disruption, I think there's no substitute for organizations that are decentralized, have appropriate incentives uh put in place for those decentralized businesses, and are a learning culture that have a a track record in the history of being learning cultures.”
On the fragile end of the spectrum, businesses that are digital services collecting financial data and packaging it on a nice user interface with subscription models are highly vulnerable because LLMs can now do much of that work for a fraction of the tens of thousands of dollars professionals spend annually on subscriptions.
“if you're um a financial uh if you're a digital service that that collects financial data and then puts it on a nice uh user interface, and has a subscription business and some folks build models on top of that that data, I'm pretty certain that your business is going to rapidly change and that your your ability to price that product where where it was 2 years ago is going to change significantly over the next couple of years because today the LLMs can do much of that work for you and the cost of uh that LLM is a small fraction of the tens of thousands of dollars today that financial professionals spend on an annual basis”
If a company lays out hundreds of billions in CapEx and bets on one or two significant horses that don't prove to be winners, they will have made a very large investment that doesn't pay off regardless of future CapEx normalization.
“So, in other words, if if a company has laid out, uh, you know, 200 or 300 or 500 billion dollars of capex, uh, and has bet on one or two significant horses, uh, to to generate the level of returns that are necessary, and those two horses don't prove out to be the winners, uh, of the race, then despite the fact that that capex will normalize at some point in the future, they will have made a very, very large investment that doesn't doesn't doesn't pay off.”
Forecasting the economy is a very difficult exercise, so the fund tries not to do it and instead invests in businesses that are agnostic to different economic outcomes and sub-segments of the economy.
“generally speaking forecasting the economy is a very, very difficult exercise. We try not to do it, and we try to be We try to invest in businesses that are agnostic to the outcomes of the economy or different sub segments of the economy.”
The fund runs a bottoms-up portfolio looking for recurring revenue businesses run by great leaders that can generate free cash flow for long periods of time, not a top-down portfolio obsessively focused on investing in AI businesses.
“I I don't want this conversation to lead people to believe that we're obsessively focused with investing in AI businesses. We're not. Um we're we're looking for businesses that that that meet that criteria, and they happen to come from a many different sectors, for sure.”
Defense spending by NATO countries could increase by a trillion dollars over the next 10 years as those countries get up to NATO targets, creating a super cycle tailwind for aerospace and defense businesses.
“We think that over the next 10 years spending on um defense could increase by trillion dollars just from those company uh just from just from those countries as they get up to to the NATO targets that have been set.”
The Magnificent Seven experienced extraordinary earnings and cash flow growth as a percentage of S&P 500 that went disproportionately to those seven stocks, while valuation multiples of the rest of the market (Mag 493) diverged massively relative to the Mag 7.
“the rate of earnings growth and cash flow and rate of earnings growth and so not cash flow growth as a percentage of the S&P 500 that has gone to the Magnificent Seven had been extraordinary. Uh and valuations ex-Mag Seven on the Magnificent 493, as we call them, had become uh relative to the Mag Seven uh very very had experienced a very very large divergence.”
Microsoft is heavily tied to the success of a single LLM, so if that LLM doesn't prove to be the winner of the AI race, they will have a hard time justifying massive CapEx investment supporting a single customer.
“Microsoft, as an example, is heavily uh is heavily tied to the success of a single LLM. So, if that LLM doesn't prove out to be the winner of of the AI race, they will have a very hard time justifying the massive investment and CapEx that they've laid out to support a a a single customer.”
Job opening data for programmers and engineers have been growing very rapidly over the last 6 months, suggesting that AI may actually increase the need for software engineers and developers rather than displace them.
“I think one of the most interesting pieces that has come out in response to the Citrini piece is the is the job opening or job search data coming out of I think there was a presentation by Citadel piece of research that came out by Citadel which said that job openings for programmers and engineers have been growing very very rapidly over the last 6 months.”
The disruption from AI to many businesses will not show up in P&Ls today but over the next couple of years many businesses will be negatively impacted, making it important to get out of the way of significant disruption.
“Um the it is not quite showing up in the P&Ls today, but we definitely believe that over the next couple of years many many businesses will be negatively impacted, and it's important to get out of the way of what what we think will be significant levels of disruption.”
On the magnitude vs. timing of AI investment payoff: It's producing significant growth for OpenAI and other LLMs right now; they're monetizing IP and growing revenue very rapidly. The question is whether they will generate significant free cash flow at some point, and whether there will be multiple LLMs at scale generating high returns, which technology history suggests is unlikely given winner-take-all markets.
“Well, it's producing it's producing a heck of a lot of growth for for OpenAI and for some of the other LLMs that are they're monetizing their IP. They're growing revenue very, very rapidly. They are gaining investments from their suppliers like Microsoft and Nvidia and they are continuing to be able to generate very, very high rates of revenue growth. The question is will they be able to generate significant amounts of free cash flow at some point in time?”
For companies negatively impacted by geopolitical conflict, investors must build in a large buffer for how long the conflict could extend, because conflicts could last a week or drag on for months or longer, affecting position sizing.
“on the flip side of that certainly um companies that could see negative outcomes uh from this conflict you have to think about how how long will this conflict go on for uh and can those businesses uh sustain themselves as the conflict carries on, and certainly build in a large buffer uh for how long it could go, you know. Um the conflict could be wrapped up in in a week, or it it could drag on for a couple more months or or a longer period of time. So, as I think about companies that will be negatively impacted, certainly you have to build that in and as you think about position sizing, certainly hopefully you've sized investments in businesses that could be negatively impacted by conflicts in a way that gives you the ability to endure whatever short-term impacts the business could receive”
Semiconductors will have long-term benefits and significant upside from the AI cycle due to essential role in LLM and infrastructure CapEx buildout.
“we've certainly invested in semis, which we believe will have long-term legs uh and beneficial benefit significantly from uh the AI cycle.”
Semiconductor suppliers don't lay out massive CapEx to drive revenue growth; they rely on TSMC for that, which allows them to generate incredibly high free cash flow growth and high returns while investing in technology development.
“They're relying on TSMC to to do that for them which has allowed them to generate incredibly high free cash flow growth incredibly high returns while reinvesting in technology development and being able to generate uh really really strong growth and high returns at the same time.”
There is a spectrum of AI risk in software: proprietary data and sticky software at low-risk end, and companies packaging publicly available data with nice UI at high-risk end.
“So, that's on one end of the spectrum. The other end of the spectrum are companies that are providing uh or packaging up publicly available data uh and reselling them by putting a a nice UI on top of what is publicly really publicly available data. And so, we had always believed that businesses on that end of the spectrum were likely to be disrupted.”
Much value will be generated over 10-15 years from utilizing innovations the Magnificent Seven have developed, suggesting downstream beneficiary thesis is correct.
“I think that your your view or or your conclusion that there'll be a lot of ca- there'll be a lot of uh value generated from utilizing the innovations that the the seven have developed over the next 10 to 15 years is right on. Uh there's no there's no doubt. They they have laid the the railroads out there for all of the entrepreneurs out there to to to see and to use.”
Shaposhnik couldn't be more excited about investment opportunities in AI era, but distinguishing which businesses generate sustainable cash flows for long periods will be investor's responsibility.
“I I just couldn't be more excited about it, and I think the investment opportunities will also be really, really significant. And it'll be on us to kind of sift through um which one of these businesses can generate real cash flow for a sustainable period of time. Um and and many won't, and so that'll be on us to figure out.”
Broad world of opportunities exists across many end markets, and there are great opportunities in many different sectors, not just AI-related ones.
“So, it's a it's a very broad broad world out there, and so there are great opportunities in a lot of different end markets, and certainly we're looking for that, and so thinking about how AI will impact and benefit businesses over the long term, for sure.”
In an era of AI and AI disruption, there is no substitute for organizations that are decentralized, have appropriate incentives put in place for decentralized businesses, and have a track record of being learning cultures.
“In an era of of uh AI and and AI disruption, I think there's no substitute for organizations that are decentralized, have appropriate incentives uh put in place for those decentralized businesses, and are a learning culture that have a a track record in the history of being learning cultures.”
AI may be enabling a wave of innovation with thousands of little specialized software providers based on infrastructure someone else paid for, many of which may never scale to billion-dollar size but still capture significant value.
“are we going to see this wave of innovation that will be very different in the sense that it will be thousands of little software providers based on the infrastructure that somebody else paid for. But they might never scale to be multi-billion dollar companies. They don't have to be.”
Building a plane before it crashes (building carefully with incremental progress) is different from jumping off a mountain and building a plane on the way down (moving fast and risky), requiring different leadership skills.
“I think a lot about this idea about this very popular venture capital term of like you jump off the building and then you you or the mountain and you build the plane on the way down. With the idea of being you could go really far. But you also might not build the plane and you crash. And there's a totally different skill set at the manager at the operator level at the leadership level between people who are trying to like build a bridge before it crashes or build a plane before it crashes and build a bridge where it's like you have some semblance of an idea of where you're going to and there's like a paste ordering of getting out there in front of it.”
Industrial distribution businesses like Granger and Fastenal can benefit from AI and robotics through increased warehouse efficiency, robotic deployment, and autonomous delivery, though they also face potential disruption if robots eventually manage their own factories and delivery.
“I think that one of the end markets that can benefit certainly is industrial distribution. Uh as I think about it, AI and robotics will significantly enhance their ability to be more efficient to deliver product uh from from where they are to the customer. And so that's Granger, Fastenal, those businesses. I think they will become more efficient, particularly as they deploy robotics in the warehouse, particularly as they deploy autonomous delivery and autonomous driving from the warehouse to the customer.”
The primary insight people might be missing is that suppliers of picks and shovels are likely to continue performing very well because they have diversified customer bases and don't require massive CapEx to drive growth.
“So if there's anything that I think people could be missing it's that the suppliers of the picks and shovels are likely to continue to do very very well uh because their customer bases are more diversified and they're not laying out massive amounts of capex to drive revenue growth.”
When Microsoft was asked whether it could return to being a capital-light cash flow compounding machine, management answered that CapEx growth for their business will grow more slowly than the cloud business growth rate, which means Microsoft is unlikely to get back to being capital-light for the foreseeable future.
“I asked them, do you think that Microsoft can can get back to being a capital-light cash flow compounding machine? And their answer was interesting. They said they thought that at some point CapEx growth for their business would grow more slowly than the growth rate of the cloud business of of of um you know, a business that they're investing a significant amount a massive amount of CapEx into today. And that to me says that it's unlikely that that Microsoft will get back to being a capital-light business for about as far as the eye can see.”
Amazon has experienced significant disruption and bugs from employment of AI in their development processes and has had to throw a lot of people at fixing those bugs created by AI, suggesting AI complexity requires more skilled human resources.
“I was reading reading um some news out of Amazon yesterday uh which said that they've experienced a lot of disruption and and bugs from the uh employment employment of AI in in their development processes and they've they've had to throw a lot of people at fixing those bugs that were created by by uh by AI”
Microsoft's engineers today are using the same off-the-shelf AI tools available to everybody else developing software, not proprietary tools, which means the competitive advantage will come from which teams, companies, and software teams can most competitively deploy these tools to generate value for customers.
“I asked them a lot of follow-up to get an answer to this question. But what was revealing to me is that Microsoft's engineers today are using the same off-the-shelf tools that are available to everybody else that is that is developing software. So today they're they have an agent that is working with them side by side using the LLM side by side with the development by the human engineer to develop software. So today there isn't there aren't proprietary tools that even the the best and most well-financed developers are using.”
Next generation of long-term compounders are companies who make best use of available tools, regardless of whether business is boring or topical.
“Inside of this theme of long-term compounders, it seems like that next generation of long-term compounders are the ones forward the companies who make the best use of the best available tools. It sounds like that's something you're looking for. Cuz it doesn't matter how boring or non-sexy your business is, or how top of the world, fascinating, topical, headline generating it is, it's just are you making great use of great tools to actually drive capital allocation and drive profitability?”
Significant opportunities exist outside Mag 7 because of percentage of market cap, multiple differential, and presence of healthy businesses in the 493 that continue to perform well but aren't seeing appreciation like Mag 7.
“And so we believe that there'd be significant opportunities outside of the Mag 7 because of the percentage of market cap, because of the multiple differential, and because we saw a lot of healthy businesses in the 493 that were just not not seen they continue to perform well, but they were not seeing the level of appreciation that was occurring in the Mag 7.”
The fund's focus on recurring revenue businesses that are needed by everybody or by everybody in a particular market provides some protection against K-shaped economy outcomes because if the high end experiences pressure but the low/middle end picks up, it should be a wash.
“generally speaking recurring revenue businesses are um they're wide in their exposure across the economy. We we try to find businesses that are needed by everybody or everybody in a particular market. And if if we're successful in doing that to the extent that the high end uh uh experience pressure, but the low end or the middle part of the economy uh picks up hopefully that'll be a wash”
Great managers have been investing in AI for the last several years, not just starting their investments today; companies that only recently began AI investment will likely fall behind.
“I think that that question and that topic brings me back to management and the importance of uh, investing with great great managers and you know, great managers have been investing behind AI, uh, for the last several years, not just starting, uh, and starting their investment today.”
The fund reduced software exposure dramatically in mid-2025 due to two factors: (1) realization that AI would disrupt many software businesses at an accelerating pace as LLMs improve rapidly, and (2) the illness and departure of Mark Leonard as CEO of Constellation Software, which triggered a disciplined reevaluation.
“What we saw in the middle of the year or uh the later part of last year uh was multiple fold. Number one, um we had been long-term investors in Constellation Software, and our CEO had fallen ill and was no longer uh capable of being CEO of the company. And so, because of our discipline, when a CEO leaves, particularly in in in a very abrupt fashion, that causes us to reevaluate the investment uh and step back and and just take our time before uh we come back to it. And so, uh number one, we saw the disruption that AI was likely to bring to many software businesses.”
Shaposhnik mentioned in mid-2025 that the year 2026 could be the year of the 493 stocks remaining in the S&P beyond the Magnificent Seven, with AI benefits broadening across the market and earnings re-rating occurring.
“Joe, at some point last year you mentioned how 2026 could be the year of the 493 493 stocks remaining stocks of the S&P, not the seven that everybody's talking about. You talked about AI benefits across the board, maybe some re-rating.”
Portfolio comes from many sectors including aerospace and defense with unclear exact AI impact but clear structural growth tailwinds.
“Um we're we're looking for businesses that that that meet that criteria, and they happen to come from a many different sectors, for sure. We talked about one of the themes, which is um aerospace and defense, and, you know, it's unclear exactly how that'll be impacted by AI today. It's certainly I think there'll be lots of opportunity for the incorporation of of uh autonomous uh vehicles, autonomous uh navigation, and and many other incorporations of it in defense, but you know, we're looking for businesses that are generating free cash flow today”
The rotation toward the Mag 493 reminds Shaposhnik of the year 2000 rotation after the tech bubble, when the rest of the market was left behind for several years and then benefited greatly from 2000-2002 as there was a significant rotation back to traditional businesses performing well.
“It reminds me to a certain extent of the rotation that we saw in the year 2000 after multiple years of very very strong returns for the technology sector. The rest of the market was essentially left behind. And the rest of the market benefited greatly from the year 2000 to 2002 to 2001 2002 where there was a significant rotation back to traditional businesses that were performing well.”
Software companies have derated massively over the last 9 months and valuations are as low as they've been in a very long period of time, creating opportunity; however, the fund has maintained only a very small position because the disruption risk remains significant and it's very early in the cycle.
“Uh and so we're studying those businesses. We've maintained a very small position uh in in software. But as I as I think about it I think it's important to be invested uh to the extent that somebody chooses to be invested in software in in businesses that have a learning culture. And so as we think about most software businesses most businesses in general are very top-heavy with management making decisions kind of at the the top layer and then sending that down to the developers and the engineers.”
Constellation Software is a learning culture that has been studying and deploying AI for a long period of time, which provides some confidence, and they've recently reoriented capital allocation strategy by investing in publicly listed software businesses at far lower valuations.
“And so, that does give you some level of confidence and comfort. And I think that um the the additional piece of news is uh or the other thought is who's taking advantage of this of this disruption today in the software space? And uh you know, we've gone through now I would call it half an earning cycle, maybe an earning cycle and a half of evaluating results from software businesses. It's so early.”
In context of greater future uncertainty, Shaposhnik prudently reduced software risk many months ago, which has proven to be a reasonably good decision given what has occurred in software.
“And so in the context of far greater future uncertainty I think that we've prudently reduced our our risk or we did that many months ago, which I think is proved to be reasonably reasonably good thing to do given what's occurred in software.”
Software company selling to municipal government market only began investing in AI last year, causing concern about management's forward-thinking capability.
“I was talking with a company who is in the software industry, sells a very very sticky product to the, uh, municipal market. And so, sometimes companies that sell to the government, uh, are are somewhat too relaxed. And they said they've only begun to start investing in AI, uh, last year. And so, that's a company that you've got to be very very concerned about.”
Mark Leonard, despite stepping back as CEO of Constellation, continues to be involved in the business, particularly in leading new capital allocation strategy toward public market software acquisitions.
“And while we're uncertain about where the world is going to go, I certainly like to be aligned with managers and management teams that are evolving as the opportunity set evolves and are not just wedded to the way it's always been done because the way it's always been done may not be the right way. It's especially in a very very dynamic in in changing in a changing world. I'd also say you know, for the longest time software was thought to be one of the highest quality business models out there in the world. And I think that there's significant uncertainty about whether that will continue to be the case. And so in the context of far greater future uncertainty I think that we've prudently reduced our our risk or we did that many months ago, which I think is proved to be reasonably”
Microsoft, which had been a top-5 position for the funds over the last 10 years, is no longer a top-5 position because the concentration and capex risk is more significant than it has been since the initial investment.
“one of the reasons Microsoft, which had been a top-five position uh for the funds that I've managed over the last 10 years or so, is no longer a top-five position because the risk there is more significant uh than has been the case uh since we made the investment a long time ago.”