February 2009: in the middle of the biggest financial crisis since the depression of 1929, this bull market was born. Today, it is in its 17th year. Despite COVID, Trump’s trade war against all his allies, high oil prices (over 100 USD at times) due to the Middle East war with Iran, it is alive and kicking.

It is all but natural that investors wonder whether things have gone up too high too fast and is it time to sell and wait for the market to correct before getting back in.

We wish we knew but we don’t. So, what do we do?

We think the most rational thing to do is to remember the fundamental reasons for owning equities in the first place: equities reflect economic growth over time and are the best proxies for business in general. Well-managed companies will outgrow their competitors and provide us with a better than average return in the long run. If chosen well, patience is your best ally. Financial reports on companies are your best tools and newspapers, newsletters from marketing sources and especially social media, if misused or misleading, are your worst enemy.

Then there is the concept of compounding effect of money:

• When you own shares of a company in a regular, non-tax-sheltered account, and it appreciates in value, for every dollar of appreciation, you will have to pay the government capital gain taxes of around 25% if you sell it. However, if you keep it for 20+ years, you will only owe it on paper until you sell it. In other words, the government “lends” you the tax money you owe them interest free until you sell your shares. Moreover, if you happen to lose money on an investment, you get to deduct your losses against other gains. Do you think you can get a deal like this one from the banks????
• If you happen to buy well-managed companies that can reinvest their profit to grow their market share, the likes of Alimentation Couche-Tard (ANCTF), CGI (GIB), Microsoft (MSFT) and many more, the compounding effect of your capital would be mind-boggling if you can take a long-term view. For example, Couche-Tard’s return on equity (ROE) averages annually over 20% in the last 15 years, CGI’s average ROE is over 15% in the last 10 years and Microsoft’s has been over 25% in the last 25 years!

In short, you were “borrowing” money interest free from the government and investing it in companies that were compounding it at an above-average rate.

Combining these 2 compounding magic tricks will justify not taking the short-term prognostics from the so-called experts even when the trajectory for the future will certainly not be a straight line.

The AI Boom versus the late 1990s Dot-Com Boom: similarities and differences

As mentioned in our last quarterly letter, the current AI frenzy is reminiscent of the late 1990s dot-com boom. Both eras feature a tectonic, technology shift, heavy capital deployment into foundational infrastructure, and narrow stock market concentration. However, examining the underlying corporate data reveals several significant differences.

The most significant divergence between the two eras lies in the fundamental cash-generation capability of the market leaders.

The Dot-Com Boom (1995–2000): The internet boom was built heavily on “speculative demand.” The median Nasdaq technology company at the peak in 2000 was entirely unprofitable. High-profile IPOs were backed by eyeballs and clicks rather than revenue, creating an ecosystem highly vulnerable to a sudden credit freeze.

The GenAI Infrastructure Cycle: Today’s infrastructure buildout is funded by the most profitable, cash-rich corporate balance sheets in economic history. Market leaders like Nvidia (NVDA), Microsoft, Alphabet (GOOG), and Meta (META) generate hundreds of billions of dollars in positive free cash flow annually. For instance, Nvidia achieved a $5 trillion market valuation backed by trailing 12-month revenue of $215.9 billion and a massive 53% operating margin.

All frenzies will end with pain and this one will not be different. Our job is not to predict the timing of a correction but identify signals that indicate “wretched” excess and problems to come.

Four things are worth watching:

• Free Cash Flow inflection: between Amazon, Google, Meta and Microsoft, they have committed over USD 725 Billion in capital spending in 2026 and promise even more in the years to come. Amazon is projected to turn cash-flow negative this year. If AI capital expenditure (CapEx) begins to exceed operation cash flow in these 4 hyperscalers, funding will have to come from capital markets and will put pressure on valuations and yields.
• As of now, no hyperscaler has even tempted to offer insight into their AI-specific operating margins, separately from their broader cloud revenue. Markets have so far accepted backlog growth as a proxy for AI returns. The price of tokens, the measuring unit for the future profitability of the business of data centres, has declined 90% since 2023 while the total capex spend has roughly doubled since last year. Here lies the structural paradox in the AI economy: data centre builders are spending double the capital to build infrastructure, while the “unit of value” they sell (the token) is rapidly deflating due to hyper-commoditization. For additional context, a token is a small unit of text analyzed or generated by an AI model. The more a company uses AI, the more tokens it consumes. For data centres to remain highly profitable in the long term, token consumption volume must grow exponentially faster than the hardware depreciation costs. Yet, Amazon Chief Technology Officer Werner Vogels recently made several comments on the rapidly climbing cost of AI through uncontrolled consumption of tokens (“tokenmaxxing”):

“We see a shift happening between the cheaper open source models and the bigger expensive models… Cost is a very important part of your architecture, you need to take that into account.”

“Do you really need to have the biggest, highest-end model to solve this? The answer is no, you don’t.”

Vogels’ comments are part of a broader corporate reckoning regarding token efficiency. Within Amazon itself, the pushback against uncontrolled token spend hit a boiling point when Senior VP Dave Treadwell sent a memo to staff demanding they stop “using AI just for the sake of using AI.” Amazon actually had to kill an internal developer leaderboard that tracked token consumption because employees began “tokenmaxxing” – pointing AI agents at pointless, repetitive loops just to climb the rankings, running up massive, empty cloud infrastructure bills for the company. Similar stories have leaked from Uber (which reportedly burned through its entire annual AI tooling budget in just four months) and Meta, proving that buyers across the board have suddenly become deeply sensitive to the raw cost of token transactions.

• There is a lot of circular financing going on in the computer chip industry, not dissimilar to the same scheme during the dot-com era in the telecom industry: as an example, Nvidia invests in OpenAI (OPENAI); OpenAI commits to purchasing Nvidia GPUs; Microsoft funds OpenAI; OpenAI runs on Azure. The OpenAI-Nvidia commitment alone is estimated to account for as much as 13% of Nvidia’s projected $272 billion in 2026 revenue. This is precisely the structure by which Lucent and Nortel financed telecommunications carriers in 1999, equipment makers were lending customers the money to buy their equipment, and it ended badly, in waves of bankruptcies from the carriers and revenue collapse at the suppliers. It could happen in AI…
• While memory chips, GPUs, and skilled engineering labour are the visible bottlenecks of the AI cycle, electricity is the quieter one. Power supply constraints are already delaying data-centre projects in Virginia, Ireland and parts of Texas. Not only could the demand prove uncertain, we have to ask whether the supply also could prove impossible.

While the dot-com boom was a bubble of speculative valuation — unprofitable companies trading on astronomical multiples of non-existent earnings, the generative AI cycle is a bubble of capital expenditure. The risk today is not that the market leaders will go bankrupt; the risk is that they are building a $725 billion infrastructure footprint that may take a decade for enterprise adoption and monetization to fully justify, leaving them vulnerable to an aggressive capex correction if returns fail to materialize fast enough.

Assessing the current landscape and areas of uncertainty…

First, (almost) everyone believes artificial intelligence has the potential to be one of the biggest technological developments of all time, reshaping both daily life and the global economy.

We also know that in recent years, economies and markets have become increasingly dependent on AI:

• AI is responsible for a very large portion of companies’ total capital expenditures.
• Capital expenditures on AI capacity account for a large share of the growth in U.S. GDP.
• AI stocks have been the source of the vast majority of the gains of the S&P 500.

Further, it’s important to note that whereas the gains in AI-related stocks account for a disproportionate percentage of the total gains in all stocks, the excitement AI injects into the market must have added a lot to the appreciation of non-AI stocks as well.

Yet, many questions linger:

• Who will be the winners, and what will they be worth? As Warren Buffett pointed out in 1999: “The automobile was the most important invention, probably, of the first half of the 20th century… If you had seen at the time of the first cars how this country would develop in connection with autos, you would have said, ‘This is the place I must be.’ But of the 2,000 companies, as of a few years ago, only three car companies survived. So, autos had an enormous impact on America but the opposite direction on investors.”
• What’s a share in an upstart worth? IPOs indicate obscene valuations that the market is willing to pay for companies that have no revenues to show for, let alone profits. The mentality of “lottery-ticket thinking” seems to be pervasive on anything with AI in its name.
• Will AI produce profits, and for whom? For vendors? Or users?

Derek Thompson, an American journalist, podcaster and author, wrote in one of his newsletters with some terrific historical perspective:

“The railroads were a bubble and they transformed America. Electricity was a bubble, and it transformed America. The broadband build-out of the late-1990s was a bubble that transformed America. I am not rooting for a bubble, and quite the contrary, I hope that the US economy doesn’t experience another recession for many years. But given the amount of debt now flowing into AI data centre construction, I think it’s unlikely that AI will be the first transformative technology that isn’t overbuilt and doesn’t incur a brief painful correction. AI Could Be the Railroad of the 21st Century. Brace Yourself”.

Conclusion?

Sometimes, we find writings that can be so insightful that we would rather reprint them as is instead of trying to paraphrase. We should give credit where credit is due. Howard Marks in Oaktree Capital Management has one of the best conclusions and bottom line regarding AI:

“…But do I have a bottom line? Yes, I do. Alan Greenspan’s phrase, mentioned earlier, serves as an excellent way to sum up a stock market bubble: “irrational exuberance.” There is no doubt that investors are applying exuberance with regard to AI. The question is whether it’s irrational. Given the vast potential of AI but also the large number of enormous unknowns, I think virtually no one can say for sure. We can theorize about whether the current enthusiasm is excessive, but we won’t know until years from now whether it was. Bubbles are best identified in retrospect.

While the parallels to past bubbles are inescapable, believers in the technology will argue that “this time it’s different.” Those four words are heard in virtually every bubble, explaining why the present situation isn’t a bubble, unlike the analogous prior ones. On the other hand, Sir John Templeton, who in 1987 drew my attention to those four words, was quick to point out that 20% of the time things really are different. But on the third hand, it must be borne in mind that behaviour based on the belief that it’s different is what causes it to not be different!

Today’s situation calls to mind a comment attributed to American economist Stuart Chase about faith. I believe it’s also applicable to AI (as well as to gold and cryptocurrencies):

For those who believe, no proof is necessary. For those who don’t believe, no proof is possible.

Here’s my actual bottom line:

There’s a consistent history of transformational technologies generating excessive enthusiasm and investment, resulting in more infrastructure than is needed and asset prices that prove to have been too high. The excesses accelerate the adoption of the technology in a way that wouldn’t occur in their absence. The common word for these excesses is “bubbles.”

AI has the potential to be one of the greatest transformational technologies of all time.

As I wrote just above, AI is currently the subject of great enthusiasm. If that enthusiasm doesn’t produce a bubble conforming to the historical pattern, that will be a first.

Bubbles created in this process usually end in losses for those who fuel them.

The losses stem largely from the fact that the technology’s newness renders the extent and timing of its impact unpredictable. This in turn makes it easy to judge companies too positively amid all the enthusiasm and difficult to know which will emerge as winners when the dust settles.

There can be no way to participate fully in the potential benefits from the new technology without being exposed to the losses that will arise if the enthusiasm and thus investors’ behaviour prove to have been excessive.

The use of debt in this process – which the high level of uncertainty usually precluded in past technological revolutions – has the potential to magnify all of the above this time.

Since no one can say definitively whether this is a bubble, I’d advise that no one should go all-in without acknowledging that they face the risk of ruin if things go badly. But by the same token, no one should stay all-out and risk missing out on one of the great technological steps forward. A moderate position, applied with selectivity and prudence, seems like the best approach.

Finally, it’s essential to bear in mind that there are no magic words in investing. These days, people promoting real estate funds say, “Office buildings are so yesterday, but we’re investing in the future through data centres,” whereupon everyone nods in agreement. But data centres can be in shortage or in oversupply, and rental rates can surprise to the upside or the downside. As a result, they can be profitable… or not. Intelligent investment in data centres, and thus in AI – like everything else – requires sober, insightful judgment and skillful implementation “.

Of note:

Alphabet (Google’s parent company) replaced Verizon in the Dow Jones Industrial Average (DJIA) on June 29, 2026, representing a significant change to one of the United States’ main indices. While it makes a major splash in financial headlines, the actual mechanical impact on portfolios and the market is more nuanced.

The Dow is a price-weighted index, meaning a company’s influence is determined entirely by its absolute dollar share price, not its total market cap.

Before the change, Verizon was trading at roughly $47 USD per share, meaning that it was only 0.5% weight in the index, therefore its daily movements have little to no impact on the index.

Because Alphabet’s Class A shares (GOOGL) trade at a much higher price of ~$355 USD per share as of writing this, it immediately commands roughly a 4% weight in the index. This places it among the top 10 most influential companies in the Dow, meaning a big day for Google can move the index quite a bit.

The Dow Jones Industrial Average hasn’t been strictly industrial for some time, but this specific swap marks the end of an era:

Bumping Verizon means the Dow has officially eliminated its last dedicated traditional telecommunications constituent. S&P Dow Jones Indices explicitly noted that Alphabet’s vast digital footprint better represents the modern “Communication Services” landscape.

Alphabet becomes the fifth “Magnificent Seven” mega-cap tech stock to be placed into the exclusive 30-member club, joining Microsoft, Apple, Amazon, and Nvidia.

Historically, the Dow was viewed as a boring, stable, value-oriented safe haven during tech selloffs. By swapping stable dividend-payer Verizon for a relatively volatile growth company like Alphabet, the index ties its fate even closer to the tech sector. If market anxieties regarding massive AI capital expenditures flare up, the Dow will now feel those shocks much more than it used to.

Ultimately, the move cements Google’s status as a foundational pillar of the American corporate world, even if it makes the nightly Dow report a little more tech heavy.

Have a good summer!

– Alain Chung, CFA, Chairman and CIO, on behalf of the Claret team.

Original Post

Editor’s Note: The summary bullets for this article were chosen by Seeking Alpha editors.