The AI Gold Rush

Investor enthusiasm for the transformative potential of artificial intelligence (AI) has continued to drive the stock market. The S&P 500 Index has achieved numerous record highs this year, largely propelled by mega-cap technology stocks. Indeed, the so-called Magnificent 7 companies have accounted for over half of the index’s price return in 2024. As of mid-year, this cohort was up 33% while the other 493 stocks had risen by 5%.

The past eighteen months have witnessed outsized performance by a select few names, leading the stock market to become increasingly concentrated. Apple, Microsoft, and Nvidia each surpassed the $3 trillion market cap threshold and now comprise one-fifth of the S&P 500’s value. Together, the 10 largest companies represent nearly 40% of the index. In contrast, at the peak of the dot-com boom in 2000, the top 10 stocks accounted for 27% of the market’s total value.

Rhyming not repeating

Drawing parallels to the 1999-2000 period, some observers suggest that history may be repeating itself. In hindsight, the March 2000 stock split by Cisco, a highflyer of that era, coincided with the peak of the IT-fueled bull market. Meanwhile, the current cycle’s darling — chipmaker Nvidia — just split its shares 10-for-1 in early June.

Sir John Templeton famously cautioned that the four most dangerous words in investing are “this time is different”. However, valuations today seem significantly less stretched than they were 25 years ago. Back then, many surging technology and telecommunications companies were burning through their cash, and some were only marginally profitable. Even “old economy” stocks such as GE and Walmart traded at 40-50 times their earnings.

Today’s market leaders have sizable, growing revenues and earnings, fortress-like balance sheets, and worldwide franchises. Investors will ultimately determine whether stock prices remain overvalued. Excluding the top 10, the stock market’s valuation is on the higher side of historical averages but within its 20-year range. So, while this time might not be different, neither is it entirely analogous to the dot-com mania of the late 1990s.

Generating a buzz

As a formal discipline, artificial intelligence has been in existence since the mid-20th century. However, its impact had been largely imperceptible. For many, Deep Blue’s (IBM) 1997 defeat of reigning world chess champion, Gary Kasparov, marked the first memorable AI milestone. More recently, the AI-powered voice assistants, Alexa and Siri, have become integrated into many of our lives.

AI hit an inflection point in public consciousness and adoption with OpenAI’s November 2022 release of its ChatGPT tool. This generative AI system, easily accessible and profoundly useful, acquired 100 million users within two months of its launch. To put this in context, Facebook required 4.5 years to hit the 100-million mark, Instagram took 2.5 years while TikTok achieved it in 9 months. ChatGPT’s explosive global popularity has allowed many to sense AI’s disruptive potential. And it is not the only game in town. Both Microsoft (Copilot) and Google (Gemini) have integrated comparable systems into their widely used platforms.

Content is king

Generative AI refers to a type of artificial intelligence that can create original content in response to user prompts or requests. This content can range from text, images, video, audio, to more complex data analysis. Rather than being explicitly programmed to perform a task, Gen AI systems are trained on massive amounts of data. They use patterns and insight “learned” from that data to make predictions and decisions or create content.

Businesses’ adoption of Gen AI remains in its early stages due to the cost and complexity of developing models to incorporate the technology into work processes. But real-world applications are emerging. In healthcare, for instance, AI tools are being used to facilitate drug discovery. Financial services firms utilize AI models to assess credit risks and detect fraud. The entertainment industry increasingly leans on AI to develop storylines, scenes, and visual effects.

A next wave

The introduction of personal computers revolutionized business operations, enabling more efficient data processing and communication. The internet further accelerated this transformation by providing a global platform for information sharing and e-commerce. AI builds upon these advancements, with the potential to automate and enhance a wide range of cognitive tasks, from data analysis to decision making. Its ability to learn and adapt sets AI apart from previous technological advances.

In 2023, enterprises invested $19 billion to develop Gen AI solutions for their operations according to International Data Corporation. IDC expects this pace to double in 2024, reaching $150 billion of annual spending by 2027. Meanwhile, high-tech giants like Amazon, Microsoft, and Google continue to make massive investments in IT infrastructure, specifically data centers that provide computing power to train and run AI models. And unlike many dot-com era firms that invested heavily in capex in the late 1990s, these juggernauts have deep pockets.

Disruptive possibilities

The IT revolution of the 1990s, driven by the extensive computer deployment and software advances, led to a decade-long surge in productivity. U.S. labor productivity growth accelerated from an average annual rate of 1.6% in the early 1990s to a 3.2% pace by the early 2000s. Rising productivity represents a key driver of economic growth and a higher standard of living. The hope is that the AI revolution will also have a salutary effect, eventually jumpstarting the meager productivity growth rate of the past ten years.

Since the release of ChatGPT, two-thirds of the rise in the overall equity market has stemmed from expanding stock multiples (as opposed to higher earnings forecasts). This suggests that investors are optimistic about AI’s ultimate impact on economic growth. Gen AI has the potential to transform labor productivity by automating routine tasks, enhancing decision-making, and accelerating innovation. A number of bullish private sector forecasts subscribe to this view.

For instance, Goldman Sachs estimates that AI could boost U.S productivity growth by 1.5 percentage points annually over a 10-year period, assuming “widespread” adoption. The Brookings Institute sees an annual boost of 1.8 points over ten to twenty years. Studies by McKinsey suggest that AI-related technologies could enhance global productivity by 0.5 to 3.4 percentage points through 2040. These bullish forecasts assume that AI will automate a significant share of existing work tasks and lead to the creation of new tasks, as previous technology waves have done (prior examples include roles such as webpage designers, software engineers, digital marketing professionals).

However, there are also analysts skeptical of the hype, disputing the notion that artificial intelligence will supercharge economic growth. The counterargument points to the fact that enterprises will find AI exceptionally expensive and challenging to implement. For businesses to earn an adequate return on investment, AI applications must solve extremely complex and important problems. Fewer tasks may prove cost-effective to automate than some forecasts suggest. Additionally, as job functions are automated, the resulting productivity may not entirely translate into economic growth if workers are displaced.

Revolutionary potential

Bill Gates wrote last year in his blog that he had seen two demonstrations of technology over his lifetime that had struck him as “revolutionary”. He said the first instance occurred in 1980 when he was introduced to the graphical user interface, the forerunner of every modern computer operating system. The second event occurred in 2022 at a meeting with the OpenAI team who showed him the capabilities of a Gen AI model.

AI’s transformative potential will unfold over time. While large tech companies continue to invest heavily in AI infrastructure, investors will eventually demand evidence of “killer apps” to justify expectations embedded in some stock prices. During the gold rush, the providers of the picks and shovels proved the true winners, not the miners themselves. If the current wave of innovation ultimately proves to be revolutionary, a much broader spectrum of companies and industries will reap the commercial benefits in the years ahead. In the meantime, history suggests investors take care not to become overly exuberant.

— Christopher J. Singleton, CFA, Managing Director
July 17, 2024