Episode Summary
Executive Summary: Howard Marks analyzes whether AI is in a bubble, distinguishing between mean-reversion and inflection bubbles. He argues AI's transformative potential is real, but investor enthusiasm may be excessive, leading to speculative behavior, circular deals, and risky debt use. He advises a moderate, selective approach, acknowledging both the opportunity and the risk of losses.
Main Topics: Bubble Dynamics and History (Priority: 5/5): Marks discusses the regular progression of bubbles, from new technology capturing imagination to FOMO-driven investment, and the role of newness in justifying limitless valuations. He contrasts mean-reversion bubbles (e.g., subprime) with inflection bubbles (e.g., railroads, internet) that accelerate technological progress. AI as an Inflection Bubble (Priority: 5/5): Marks explores whether AI is an inflection bubble, where speculative investment funds infrastructure that enables future deployment. He cites Perez and Hobart/Huber to argue that such bubbles can be net beneficial for society, though destructive for many investors. Current AI Landscape and Speculative Behavior (Priority: 5/5): Marks details the massive capital expenditures, astronomical stock performance (e.g., NVIDIA), and speculative behaviors like $1 billion seed rounds and circular deals. He highlights the uncertainty around AI's commercial application, winners, and profitability. Debt and Financial Engineering in AI (Priority: 4/5): Marks examines the use of debt to finance AI infrastructure, including 30-year bonds and off-balance-sheet SPVs. He warns that debt magnifies losses and increases systemic risk, drawing parallels to the telecom bust and Enron. Societal Impact of AI: Job Loss and Purpose (Priority: 4/5): In a postscript, Marks expresses concern about AI's potential to eliminate jobs, reduce income tax receipts, and create social division. He questions whether universal basic income can replace the non-financial benefits of work. Historical Parallels and Lessons (Priority: 3/5): Marks compares AI to past bubbles like railroads, electricity, radio, and the internet. He notes that while parallels are inescapable, believers argue 'this time it's different,' a phrase that often precedes bubbles.
Key Arguments: Bubbles are not caused by technology itself but by excessive optimism applied to it. Inflection bubbles (like AI) can be beneficial by accelerating infrastructure build-out, but they destroy wealth for many investors. The use of debt in AI financing is risky because the outcome is uncertain; debt is appropriate for predictable cash flows, not speculative ventures. AI's impact on employment is terrifying; it may eliminate jobs without creating enough new ones, leading to social and political division. A moderate, selective investment approach is best: neither go all in nor stay all out. Historical parallels (e.g., radio, aviation) suggest that transformative technologies often lead to overinvestment and painful corrections.
Data Points: NVIDIA market cap appreciation: ~8,000x (40% per year for 26+ years) - From $626 million at IPO to briefly $5 trillion. AI-related stocks' contribution to S&P 500 gains: 75% of gains, 80% of profits, 90% of capex - Fortune headline, October 7, 2024. Estimated AI infrastructure build-out cost: $5 trillion - JPMorgan estimate, not including tip. Thinking Machines Lab valuation increase: From $12 billion to $50 billion in 5 months - Startup with no product, raised $2 billion seed round. Anthropic revenue growth: 10x each of last two years (100x total) - Claude Code revenue running at $1 billion annual rate. AI's potential time savings for jobs: 43% of time on work tasks - Vanguard's Joe Davis estimate for 4 out of 5 jobs.
Pivotal Quotes: "We'll build this sort of generally intelligent system and then ask it to figure out a way to generate an investment return from it." — Sam Altman (paraphrased by Howard Marks): Captures the uncertainty of AI's business model and the speculative nature of investment. "For those who believe, no proof is necessary. For those who don't believe, no proof is possible." — Stuart Chase (attributed by Howard Marks): Applied to AI, gold, and cryptocurrencies, highlighting the faith-based nature of the enthusiasm. "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." — Derek Thompson (quoted by Howard Marks): Argues that AI could be the railroad of the 21st century, with a painful correction but lasting transformation.
Implications: Investors should adopt a moderate, selective approach to AI, balancing the potential for transformative gains with the risk of significant losses. The use of debt in AI financing increases systemic risk, and society must prepare for job displacement and social division. Historical patterns suggest a correction is likely, but the technology's long-term impact will be profound.
From the Transcript
Die or bust, Dr. Koronick said. The New York Times, November 20th. The yet-to-be-determined nature of the industry under construction is best captured in remarks from Sam Altman, the CEO of OpenAI, that have been paraphrased as follows. We'll build this sort of generally intelligent system and then ask it to figure out a way to generate an investment return from it. This should be a source of pause for people who heretofore fully comprehended the nature of the business they invested in. Clearly, the value of a technology that equals Or surpasses the human brain should be pretty big, but isn't it well beyond calculation? A word about the use of debt. To date, much of the investment in AI and the supporting infrastructure has consisted of equity capital derived from operating cash flow. But now, companies are committing amounts that require debt financing. And for some of those companies, the investments and leverage have to be described as aggressive. The AI data center boom was never going to be financed with cash alone. The project is too big to be paid for out of pocket. JPMorgan analysts
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 behavior 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.
1929, it left us with the Great Depression. AI is the bubble to burst them all. Brian Merchant, Wired, october twenty seventh. Please note the Depression had many causes beyond the bursting of the radio aviation bubble. Derek Thompson, who supplied the quote with which I opened this memo, ended his newsletter 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 U.S. economy doesn't experience another recession for many years. And given the amount of debt now flowing into AI data center 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 twenty first century. Brace yourself. November fourth.
About The Memo by Howard Marks
On October 12, 1990, Oaktree Co-Chairman Howard Marks published his first memo to clients. In the decades since, he has periodically released memos reflecting his viewpoint on the investment landscape, as well as more general business insights. On this podcast we'll hear the latest memos by Howard, released in tandem with or shortly after their publication.