16 Mar Now You See It (Q4-24)

Speculative enthusiasm is as old as markets themselves. Each generation of investors convinces itself that it is witnessing something unprecedented – a technological breakthrough or economic transformation so profound that old rules no longer apply. Today, amid untethered excitement about artificial intelligence, the belief that we have entered a permanent new era of accelerated growth is once again taking hold.

Yet history tells us that markets have an uncanny ability to make fools of those who believe the past no longer applies. From the rise of the automobile and telephone in the 1920s to the internet revolution of the 1990s, every period of disruptive innovation has fueled the same euphoria, leading investors to abandon arithmetic in favor of fantasy. Arthur C. Clarke famously observed that “any sufficiently advanced technology is indistinguishable from magic.” The problem is that investors, time and again, mistake this perceived magic for a license to ignore economic fundamentals.

Despite all the transformative technological breakthroughs of the past two decades – smartphones, e-commerce, cloud computing, automation, and now AI – both U.S. GDP growth and S&P 500 revenue growth have actually slowed, not accelerated. Since 2000, these measures have averaged just 4.5% annually, lower than in the preceding half-century.¹ One would think that with all the world-changing innovations since 2000, growth would be faster, not slower. It has not.

The lesson? The impact of innovation on markets follows a familiar pattern: early profits surge, valuations reach extremes, competition intensifies, margins compress, and the long-term benefits accrue mostly to consumers, not to speculative investors chasing unsustainable narratives.

Stock market gains have come disproportionately from a handful of mega-cap technology companies, just as past bubbles have been led by their own anointed winners. Investors convince themselves that this time, the winners are different – immune to the laws of competition and destined to outgrow the economy indefinitely. But history suggests otherwise. Every speculative peak has been defined by a cohort of “must-own” companies that were expected to dominate forever. Yet over time, market leadership is constantly disrupted. Consider:

• In 1929, the dominant companies were industrial titans tied to automobiles, oil, and consumer goods.
• In the 1960s and 1970s, it was electronics, aerospace, and the Nifty Fifty, with IBM, Kodak, and Xerox at the forefront.
• In 1999, it was internet pioneers like Cisco, Intel, and Microsoft, commanding valuations that assumed decades of uninterrupted dominance.

Each of these groups was seen as invincible at the time. Yet as the speculative peaks unwound, reality set in: profit margins normalized, growth expectations were revised downward, and stock prices suffered, sometimes for decades.

Today’s AI obsession is following the same script. Companies like Nvidia, Broadcom, Super Micro Computer, and Palantir have been propelled to astronomical valuations based on the assumption that AI will drive perpetual earnings growth. But just as with past cycles, competition will intensify, margins will contract, and much of AI’s economic benefit will accrue as consumer surplus rather than as permanent excess profits for investors.

A common argument supporting today’s elevated stock valuations is that corporate profit margins are structurally higher due to technology and globalization.² While these factors have played a role, they are not the primary drivers of margin expansion over the past three decades. The most significant driver has been declining interest rates. Since 1990, corporate EBIT margins (earnings before interest and taxes) have barely increased. The key reason net profit margins have expanded is that companies have enjoyed steadily lower interest expenses thanks to falling borrowing costs. In 2020-2021, many companies locked in historically low rates, further extending the illusion of permanently high margins.

Now, with interest rates back at more normalized levels, this tailwind is disappearing. Companies that relied on cheap debt to drive earnings growth will face rising borrowing costs, compressing margins and exposing just how unsustainable much of the past decade’s profit expansion really was.

Investors make the repeated mistake of basing valuations on excitement rather than arithmetic. When a company or sector is growing rapidly, it is tempting to project that growth forward indefinitely. But even in industries experiencing massive technological shifts, growth follows a trajectory that is not a straight line:

• In the early 1900s, the rise of the automobile revolutionized transportation, yet the auto industry soon became fiercely competitive, with margins eroding over time.
• The 1960s and 1970s saw the explosion of electronics and computing, yet even dominant players like IBM saw prolonged periods of underperformance as competition caught up.
• The 1990s internet boom saw enormous enthusiasm around online commerce and software, but investors who bought at peak valuations had to wait nearly 15 years just to break even.

Investors today, enthralled by the promises of AI, risk making the same mistake. Yes, AI will be transformative. But the assumption that AI will deliver indefinite profit expansion for a select few companies is an illusion. Capitalism ensures that competition eventually narrows the gap between innovation and economic reality.

The psychology of speculative bubbles follows a predictable sequence:

1. A real, transformative innovation emerges.
2. Investors extrapolate early successes into an unsustainable growth trajectory.
3. Capital floods into the sector, driving valuations far beyond reasonable levels.
4. Margin pressures, competition, or macroeconomic forces begin to erode profitability.
5. The unwinding begins, leaving latecomers with steep losses.

This pattern has played out in every major speculative cycle – from the railroads of the 1800s to the dot-com crash to the SPAC frenzies of recent years. It is unfolding again in today’s AI-driven stock boom.

Speculation thrives on narrative, not numbers. When investors begin justifying extreme valuations with phrases like “new era,” “paradigm shift,” or “this time is different,” history warns us to be skeptical. Yes, technology is evolving rapidly. Yes, AI will reshape industries. Yes, some of today’s dominant companies will continue to succeed. But that does not mean profit margins, revenue growth, and market leadership will defy gravity indefinitely.

Ultimately, long-term wealth is built not by chasing momentum, but by respecting arithmetic, recognizing cycles, and maintaining discipline. The history of markets is clear: those who buy into speculative frenzies based on the belief that they are witnessing magic will, in the end, find themselves in the disappearing act.

[1] John Hussman, November 2024
[2] “AI stocks aren’t in a bubble”, Goldman Sachs, September 18, 2024