The Dividend Cafe - Investment Conviction in a Time of “Revolution”?

Episode Date: October 10, 2025

Today's Post - https://bahnsen.co/3KCGbir Unpacking the AI Revolution: Investment Philosophies and Realities In this episode of Dividend Cafe, David Bahnsen, Chief Investment Officer at The Bahnsen Group, explores the current excitement surrounding AI investments. Bahnsen contrasts contrasting perspectives from investment experts and delves into the cyclical nature of AI-driven market trends. He emphasizes the importance of adhering to a disciplined investment philosophy, particularly dividend growth investing, even amid the AI craze. Through an analysis of market dynamics and speculative risks, David outlines three potential scenarios for AI investments and underscores the need for consistency and prudence in managing client capital. 00:00 Introduction to Dividend Cafe 00:05 Contrasting Views on AI Revolution 02:56 Current AI Market Dynamics 04:41 The Circularity Problem in AI Investments 06:00 Soros' Reflexivity Theory and AI 14:56 Three Potential Outcomes for AI Investments 19:58 The Bahnsen Group's Investment Philosophy 27:35 Conclusion and Final Thoughts Links mentioned in this episode: DividendCafe.com TheBahnsenGroup.com

Welcome to the Dividing Cafe, weekly market commentary focused on dividends in your portfolio and dividends in your understanding of economic life. Hello and welcome to a very special dividend cafe. My name is David Bonson. I am the chief investment officer here at the Bonson Group. And today we're going to be talking about our investment philosophy in this moment of the AI revolution. I want to start with two quotes from money managers, both of whom are personally multi-billionaires, both of whom are people we have known for quite some time, both of whom are people we look up to a great deal and have a track record to warrant our respect. And I want to read these quotes to you so you can contrast as I set up our topic today in the Dividendon Cafe. 1. AI is a revolution and when something is revolution, you don't ask questions. You back up the truck and just buy. I can see something's going up 10x from here with 8 to 10 trillion dollar valuations a norm. Okay. Number two, this entire AI setup is extremely concerning. We are certain we have seen this movie before and we are going to do whatever we can to avoid the carnage.

that is coming. I am being asked more than at any point I can remember in my 25 years of professionally managing money if we're going to adjust what we do or put our normal investment philosophy on pause for this particular moment in which we find ourselves. An AI craze where certain companies are going through the roof, where there is this tremendous optimism, some of which we believe is entirely understandable, some of which may prove not to be. We're going to unpack all of that here today in the Dividing Cafe. But I'm being asked over and over if we want to adjust and then maybe come back in to what we normally do with dividend growth investing a little later after this AI period plays out.

for a bit. Get on the sideline, come back in, violate what you believe in because, hey, this thing is a revolution and you don't ask questions in a revolution. I hope that out of today's Dividend Cafe, you will not only better understand what we believe about managing money and why, but you will better understand what this actual AI moment is, what is happening, pushing certain stock prices higher and what some of the outcomes may very well prove to be. It's a dividend cafe that I enjoyed writing a great deal. I want to now deliver its message to you listening to the podcast watching the video and I hope it will be instructive as we intended. We're going to start with what I think is just the basic reality of what's happening in the AI moment. That is those

those who sell or design semiconductor chips that are related to artificial intelligence applications are doing very, very well in their stock price. Number two, those who purchase those chips, the basic computing systems necessary to drive an infrastructure related to AI are doing very, very well. These types of companies are often referred to as hyperscalers. and number three is those that are just somewhat adjacent to AI but then do a good job marketing or branding themselves as being an AI play and that has always been a dynamic of public market investing when you get a moment like this whether it was dot com or or crypto or cloud or whatever

it does whether your business is fundamentally in that space or not sometimes the third category or just people that decide to go put some branding on it around that phase. We're primarily concerned both with number one and number two today because I think those are where the questions exist as to how this is going to continue and whether or not we need to go jump on in. Whereas with number three, I think there are more reasonable people that are more reasonably prepared for the reasoned response that such scam connectivity to this moment is not necessarily.

sustainably investable. But others do wonder if the other categories are, and that's what we're going to kind of examine. I have talked a lot, and others have been having this conversation more recently as well, about the so-called circularity problem in this AI investable thesis. And I want to read directly from my own Dividend Cafe here so that I capture what we're really referring to. As succinctly as possible, the lion's share of investor gains so far in this story have come about as some companies are selling the powering ability of AI to customers called hyperscalers, and the stock of hyperscalers has gone up because of their purchasing of AI computing power. Now, in one easy summary, some companies are going up because they're selling and other

companies are going up because they're buying, and it's not necessarily a problem, but it leads to us asking more questions. And before I get into the details of how this circularity cycle is playing out, I think it's going to fascinate you. I think it's helpful to unpack the details so people understand what really is going on. But I do want to bring it back somewhat philosophically to an investor that I do not very often quote or cite in my own analysis and in my own sort of cultural commentary. and that is George Soros, who's a pretty toxic figure politically and not necessarily one that I share a lot of common ground with from a worldview standpoint. But most certainly around his reflexivity thesis provided the investing world with tremendous wisdom. There's plenty of others that could

have formulated the same thing. But I think Soros did a wonderful job in his reflexivity theory explaining how markets are often driven, not just by objective reality, but by a perception of reality. Now, that's simple enough. That's not the heart of the theory, by the way. That sometimes people believe things that aren't true and act on it is true enough. And then there is, of course, always the underlying truth or reality that can drive investment outcomes. But what Soros taught us was that investor perception is material to fundamentals in that it creates a self-reinforcing cycle where these perceptions of reality influence prices, prices impact fundamentals, and then those affected fundamentals now shape perceptions, rinse, and repeat.

That there is a sort of cycle at play that creates a feedback loop, and it can enable this, process to stray further and further from market equilibrium from objective reality, the greater, the distortion caused in this cycle, the more susceptible we are to booms and busts. And I don't think you have to be a fan of Soros to understand the wisdom of what he's referring to here. And, you know, look, people's perceptions can be wrong and that creates a bad outcome. That's simple enough. But what he's saying here is, is that in a distorted impact to prices, you get fundamentals that appear different and you respond to those different fundamentals now, and so it feeds on itself. And there is this

kind of reflexivity process that can become very, very distorted over time or create severe distortions in markets. And the course correction by which we go about rediscovering what is objective of reality, the actual market truth and dynamic. It can be very skewed. It can be very prolonged. It can be very complicated and, of course, do some significant damaged investors along the way. The circularity in the AI thesis is that the hope and promise of a new technology, a hope and promise of a new technology that I share, by the way, has understandably led to higher valuations. No problem. But that then those higher values, valuations are attracting more and more capital. And this chasing of hope and promise,

what aforementioned person referred to as a revolution, is then used as the proof of the thing the capital is chasing. The proof of what we're after with this AI story is the high valuations that are coming about as a result of all the capital chasing it. That's the circularity issue here. What is it we're supposed to be after? Genuine productivity gains. Genuine profit gains. There is no need for this new technology to be monetized for investors when the monetization is coming to them before the actual monetization of the thing. The monetization is basically the elevated valuations, higher prices that are coming about from those who have the perception of this thing, this revolution. Now, this is structurally very dangerous.

It has a number of ways in which it can play out. But I want to explain to you in practice what I mean by this and how there's a cyclicality going on that I don't. don't think very many people understand. I'm going to quote right now from the head of investment strategy at J.P. Morgan, Michael Symbolist, who had a quote that really kind of went viral on the internet and social media a week or so ago. And there's a link at Diven Cafe about all this, and it's been heavily quoted in the news. But Oracle stock jumped by 25% after being promised $60 billion a year from Open AI. An amount of money Open AI doesn't earn. to provide cloud computing facilities that Oracle hasn't yet built and which will require

4.5 gigawatts of power about four nuclear plants and two and a half Hoover dams that we don't have, as well as increased borrowing by Oracle whose debt to equity ratio is already 500%. The tech capital cycle is about to change. None of this, by the way, is a negative comment per se on Oracle or Open AI or Nvidia or any of the companies that are involved in this. It's to explain what the cycle of gains exactly is. Okay, a company like OpenAI, who's the maker of ChatGPT, was given a $500 billion valuation in their latest funding round.

They have committed to buying $300 billion of computing power from Oracle, $60 billion a year for five years. Oracle is committed to buying tens of billions of dollars of chips from NVIDIA. NVIDIA, well, first of all, Oracle's stock went up on the news Open AI was doing it. Open AIs valuation went up on the news of their purchasing of this from Oracle. And in the meantime, the Nvidia stock went up as Oracle promised them a big order for chips. Now, how does Open AI pay Oracle for the money that Oracle is in a pay? pay NVIDIA for the chips?

NVIDIA made $100 billion investment in Open AI. Open AI has done similar deals with AMD, Broadcom, but what has never been provided is any estimate of the return on investment of all of this capital expenditure. And we don't need to pick on Oracle and Open AI and Vivida, they're part of this process. There are all sorts of hyperscalers spending hundreds of billions of dollars. that they then in tune need to go turn and monetize. And the valuations and asset prices and all these three-way transactions

continue to go higher. And that gives the appearance of a validation that there is a real value creation going on. And it's certainly rewarding investors, especially those who decide to sell out of these big gains. Checks are getting written. There's real-life money being deposited in accounts.

but the fundamental underlying deep down question, what is the value creation from the underlying product? How do the big AI CAPEX customers monetize the amount that they are spending? That is the question in front of us. Now, first of all, let me make clear. I am not suggesting that there isn't an answer or that the eventual answer is going to prove to be a disappointing one. I'm 100% open to the idea, that there is a path here for tremendous profitability to this massive amount of money. In aggregate, it's trillions of dollars being committed or spent. But what I'm not open to is the idea that anyone knows the answer or knows how or knows when. There are a lot of ifs, hows, when's out there. And that's different than saying AI CAPEX will prove to be a disaster.

I'm not saying it proved to be a disaster. I'm saying that there is a wide. widely speculative backdrop to trillions of dollars of investment right now. Where will it go from here? I'm not in the prediction business. I am in the business of risk-reward calculations. But what I would suggest to you, there are three basic categories that are useful as a kind of compartmentalization. There's thousands of subcategories. and different options and optionalities as to where a lot of these things could go, subsets of subsets and all of that stuff. But I would say that we're not out of bounds to suggest that there is, number one,

a sort of revolution, euphoria outcome, a kind of best case whereby all of those most desirable outcomes materialize, outlandish valuations, become rationalized, and we spend the next 10 years just blown away at how much all this AI CapEx story was monetized. In this scenario, the unknowns and risks become known payoffs. And when I say monetized, they do not mean just simply that stock prices go up of the customers, but that the end user, what the customer is buying it for results in a tremendous path to revenue growth. And in this scenario, risk takers are rewarded and done so without any consequence for the tremendous risks that they have taken. That is scenario number one.

Scenario number two is what I'm going to call the doomsday panic outcome, where the AI story ends like the metaverse for those old enough to remember that thing from three years ago, only with a lot more zeros and commas involved than the metaverse had, where there is just a kind of eventual conclusion that there is not going to be an adequate ROI relative to the amount of money that has been spent. The monetization path fails to materialize and the circularity problem reverses. And so the basis for stock appreciation and hyperscalers and computing providers becomes a basis for depreciation of both. That's a kind of indiscriminate crash outcome. That's what we'll call scenario two, a doomsday panic. But then scenario

Number three, we're going to call a nuanced reality outcome where there are winners. There are losers. There's a lot of unexpected developments along the way. And those who do go into this with eyes closed, no diligence, they do get punished. Some who do more prudent analysis along the way. They get rewarded. Some who avoid it altogether are regretful. And so you can call it a more middle ground issue, But what happens is there is some end-user monetization and it rationalizes some CAP-X. It makes a mockery of other CAP-X. But it just takes years to parse out winners and losers, but there are winners and losers. You can call this the Internet outcome because it's a very similar story to what did materialize with the World Wide Web. And I think you can call it a history repeats outcome because I think this is basically

the suggestion that history will repeat in a similar way that has before with many other technological revolutions. And I think it is a kind of natural order of things. It is a more likely scenario out of what we know about not only economics and business cycles, but human tendencies through all of this as well. I would tend to pour water on theory number one, outcome number one. And that is simply because the risk reward tradeoff and it strikes me as highly unlikely. The entire business sector right now is generating about $50 billion of revenue. I have no doubt in my mind that number can and will go higher. But the estimates that we're reading that are very rationally intelligently assembled suggest that $2 trillion is a low number and I've seen

sum as high as $3 trillion for what is going to be needed in eventual revenue streams to justify the current capital expenditures to deliver any kind of real-life AI. And I mean in four or five years. Two trillion a revenue is more than all of these hyperscalers, the biggest five, six, seven companies in the world, name brand companies you've heard of every day, all of their revenue put together. Not from AI, from everything that they do. Okay. We're just simply talking about a tremendous expectation for revenue growth without any clarity as to what exactly is going to create outside of that circularity story and technology. The reason I'd pour water on outcome number one is that I think it is hardly new that an overinvestment period would take place that

results in overvaluation, overcapacity, and eventually the need for a boom to be solved by a bust so that there can be appropriate level of liquidation, right sizing of capital, a correction, adjustment. And this is a very familiar tune. and I believe to be not altogether unlikely one. Now, our job at the Bonson Group is not to invest client capital on rank speculation. The AI story contains speculation. It also contains a lot of promise, a lot of opportunity. I do not consider the AI story to be entirely speculative.

What I believe is that the way in which many people are investing in it is far more speculative of them people understand, that the investable part of the AI story results in companies that make goods and services that meet human needs and wants, driving a better efficiency and productivity, and that if that doesn't materialize in a way that they, by investing and paying for something, can generate that higher productivity, higher efficiency, then the whole thing becomes a really, really big problem. I think that there will be those efficiencies, and some of them are to be determined, but that ultimately the success of AI comes down to its ability to be used by real-life companies, not only in the technology sector buying from one another, but in

consumer staples, in health care, in financials, in industrials, and all of these other stories that are just real-life companies producing goods and services that meet human needs and wants. and along the way, what people want to do to get exposure to the space is only by the hyperscalers at 58 times earnings or 39 times earnings or 84 times earnings, then it will either end really badly for them or it will not. But that is not an outcome that I consider an attractive risk-reward trade-off. What I do think is extremely attractive is allowing the benefits of technological revolution to accrue to real-life companies. Now, that becomes a more boring story. That is not immediately visible with a price signal in the form of evaluation boom. That is less

exposed to Soros's reflexivity thesis. And I believe is a much more prudent way for us to approach this story. People will say, well, along the way, you see some of these other things going up. Are we going to get left behind? Is it time to at least pause the fundamentals of dividend growth to kind of join this euphoric party? And the answer to that question is no, it is not. Because the answer is always no, as to whether or not a smart thing to do is abandon principles you believe in. I have spent over 20 years essentially managing money in a dividend growth orientation. And we have studied the realities of dividend growth investing going back a hundred years of its real life impact in markets and so forth. I believe that what dividend growth is and the things that people right now are talking

about as a bug are in fact a feature. And that is the sustainability, the predictability, the consistency. This is not free of risk. This is not free of some companies that become overvalued. This is not free of some companies that face fundamental headwinds. We have some dealing with that now. There are challenges around operational execution. But those things do not represent existential threats and boom-bust cycles that we're talking about many investors exposing themselves to right now. We are talking about a far lower volatility profile, but with significant upside. to the future, including the upside that comes from its adjacent to the artificial intelligence possibilities. I do not believe dividend growth is divorced from the AI story. I think it is a very

responsible and perhaps the most prudent way to play the AI story. If AI is growing profit margins for real-life companies, then companies with propensity to pay a growing dividend investors are going to have more free cash flow through higher margins that allows them to share more dividends of investors. That's a cyclical self-reinforcing cycle that we like a lot. But the point being that an investment philosophy that you put on pause to let a fad mania or euphoria play out is not an investment philosophy. It's a gimmick. And it is not something one believes in at all. If you went to a church and a pastor told you believed in something and preached a sermon on it, but then he did the exact opposite in his life. He would say, well,

that person doesn't believe in what they said. Now, you might just think they're a hypocrite, but you also might believe they don't even believe it to begin with. An investment professional who is willing to put aside a diligently constructed means of investing client capital because of a temporary moment is not really committed or convicted by that. investment philosophy whatsoever. And I think that what we need more than anything right now in a period of mixed macroeconomic signals, in a period where a lot of things happening at these levels of valuations and whatnot rhyme with things we've seen before in periods of what proved to be tremendous market vulnerability. I think what investors need is investment

professionals that actually believe in what they're doing and do it because they're they believe in it and allow it to generate the successful results it has generated over time, even though there will be periods where it may be in favor and out of favor. There is no such thing as an investment philosophy that is always in favor. That's not the way investment works through the cyclicalities and dynamics of all sorts of things. Certain approaches find themselves in favor at times, out of favor at times. But what we want to avoid, first of all, as a professional, legal fiduciary, but also, in my very humble opinion, as someone who is desperately desirous of practicing ethics, integrity, and trustworthiness, what you want to do is be rooted to something you believe in and the dividend growth mantra, whereby free cash flows are growing over time and being shared more and more with investors. Constantly allowing them to de-risk their investment through the actual real-life receipt of a reward,

while that can be used to reinvest in the company or used to pay real-life bills, both with drawers and accumulators in practicality, create different advantages out of the approach. But this, to me, represents a very sensible way to invest money, even when some stocks are going up 10x on a craze. And it also represents exposure to the underlying promise of the craze itself, that if this AI moment is real, and many parts of it are going to prove to be, in my opinion, then those things will be monetized through actual portfolio companies, not merely inside that circularity trap where many people find themselves in the moment. I do believe that there is more at stake right now than just people sticking to their guns. believe that there is the need for consistency. I believe there is the need for that kind of integrity and investment, but I also believe that it is the right way to manage money, not

merely because it reflects someone's conviction and consistency, but because there is need to have a path outside of this moment whereby there could be a significant amount of carnage on the other end. Option three of the way of the three sort of broad components. apartments I suggested, seems to be the highest likelihood. And in option two of total carnage, I don't think that our approach gets hurt that badly. Option one, I think is very unlikely. And option three, I think, is not only the most likely one, but the one that will do the most good for most investors. This is the way we're approaching this current AI moment, constantly looking for ways in which our portfolio companies can benefit from it.

but at the same time not being guilty of closing our eyes, buying both hands, not asking questions. That is not what clients pay us to do. We're asking the questions, and hopefully today you feel we've given you some answers. Thank you for listening. Thank you for watching, and thank you for reading the Dividing Cafe. The Bonson Group is a group of investment professionals registered with Hightower Security's LLC, member FINRA and SIPC, and with High Tower Advisors, LLC, a registered investment advisor with the SEC.

Securities are offered through Hightower Securities LLC. Advisory services are offered through Hightower Advisors, LLC. This is not an offer to buy or sell securities. No investor process is free risk. There is no guarantee that the investment process or investment opportunities referenced torian will be profitable. Past performance is not indicative of current or future performance and is not a guarantee. The investment opportunities, reference TIRAN, may not be suitable for all investors.

All data and information referenced herein are from sources believed to be reliable. Any opinions, news, research, analyses, or other information contained in this research is provided as general market commentary and does not constitute investment advice. The Bonsor Group in Hightower shall not in any way be liable for claims and make no, express, or implied representations or warranties as to the accuracy or completeness of the data and other information, or for statements or errors contained in or omissions from the obtained data and information referenced here in.

The data and information are provided as of the date reference, such data and information are subject to change without notice. This document was created for informational purposes only that opinions expressed are solely those of the Bonson Group and do not represent those of High Tower Advisors LLC or any of its affiliates. High Tower advisors do not provide tax or legal advice. This material was not intended or written to be used or presented to any entity as tax advice or tax information. Tax laws vary based on the client's individual circumstances and can change at any time without notice. Clients are urged to consult their tax or legal advisor for any related questions.