We are living through a moment where Artificial Intelligence claims to change everything, yet fears that America's biggest corporations are propping up a massive bubble are dominating the headlines. In his latest memo, legendary investor Howard Marks tackles the question everyone is asking: "Is it a bubble?". He dives deep into the psychology of market manias, comparing the current AI frenzy to historical booms like the railroad and the internet to help us understand where we stand today.

Howard Marks is the Co-Chairman of Oaktree Capital Management and a revered figure in the world of value investing. He is best known for his insightful memos that dissect market cycles, investor psychology, and risk management. In this piece, Marks admits he is "no techie," but uses his decades of experience watching market behavior to analyze the financial craze surrounding AI.

The Anatomy of the Bubble

Marks begins by reminding us that bubbles follow a predictable psychological progression. A revolutionary technology captures the imagination, early investors make huge gains, and then envy takes over. As Marks notes, "Those who merely look on feel incredible envy and regret and... pile in". This creates a dangerous environment where investors stop worrying about price or risk, driven entirely by the fear of missing out (FOMO). He argues that while history doesn't repeat exactly, it certainly rhymes, citing the "word-for-word applicability" of the 1860s railroad boom to today's AI surge.

Not All Bubbles Are Created Equal

One of the most fascinating points Marks raises is that not all bubbles are purely destructive. He distinguishes between "Mean-reversion Bubbles" (like the subprime mortgage crisis, which offer no societal progress) and "Inflection Bubbles". Inflection bubbles, like the dot-com era or the railroad boom, are actually necessary evils. They result in massive losses for investors, but they fund the infrastructure required for technological revolutions. As Marks puts it, "Bubbles based on technological progress are good because they excite investors into pouring in money... a good bit of which is thrown away... to jump-start its exploitation". The takeaway? Society might win from AI, but that doesn't mean your portfolio will if you buy at the top.

The Danger of "Lottery-Ticket Thinking"

Marks observes a shift in valuation logic toward what he calls "lottery-ticket thinking". This happens when the dream of a trillion-dollar payout compels investors to bet on startups with almost zero probability of success. He points to staggering examples, such as a startup raising billions at a massive valuation despite having "refused to tell investors what they're even trying to build". While venture capitalists play this game of expected value, Marks warns that for the average investor, believing that "thinking about a trillion-dollar payout will override reasonableness" is a recipe for disaster.

The "Circular Deal" Warning Sign

A specific red flag Marks identifies is the prevalence of "circular deals," reminiscent of the telecom bubble of the late 1990s. He notes that huge revenue figures in the AI space might be illusory, as companies like OpenAI, chipmakers, and cloud providers seem to be paying each other. For instance, "OpenAI is set to receive billions from tech companies but also sends billions back to the same companies to pay for computing power". This "round-tripping" of money can exaggerate progress and profitability, making the sector look more stable than it actually is.

Debt vs. Equity: The Critical Mistake

Perhaps the most worrying trend Marks highlights is the shift from funding this boom with equity to funding it with debt. In the past, risky tech ventures were funded by equity (ownership), which is safer for the system because if the company fails, the stock just goes to zero. Today, however, "companies are committing amounts that require debt financing". Marks warns that "Winner-takes-all" competitions should be played with equity, not debt. If these massive investments in chips and data centers become obsolete faster than expected, a real risk in tech, the debt remains while the asset value vanishes.

Conclusion & Call to Action

So, is it a bubble? Marks concludes that there is "no doubt that investors are applying exuberance with regard to AI," though whether it is strictly "irrational" remains to be seen. However, he advises against sitting on the sidelines entirely. Because AI is likely a genuine transformative technology, "no one should stay all-out and risk missing out".

His final advice to value investors is to thread the needle: "A moderate position, applied with selectivity and prudence, seems like the best approach". Acknowledge the risk of ruin, avoid the hype of circular economics, but maintain exposure to the future.

For more of my insights on this topic, be sure to follow me.