This is the time of year when the oracles of the investment industry tell us where the economy is headed in 2026. After all, as investment managers, we should have clear sight into where the markets are headed and should be able to position client portfolios correctly so as to avoid calamities.

In reality, an outlook for the coming year is the equivalent of meteorologists attempting to forecast next year’s hurricane season. Sure, there is a chance there are no disruptions and no volatility, and the year is filled with a mild breeze, sunshine and smooth sailing for our investment portfolios. That would help investors sleep well at night, but history shows that is not a likely outcome.

As we begin 2026, one of the areas that has the potential to significantly impact the markets is artificial intelligence (AI) and whether we’ve reached the peak of what has been a frenzied few years.

The Internet Buildup Example

There are many historical examples of investor frenzies around market themes over the years, including railroads, tech stocks, housing and now artificial intelligence. While the themes have always been different, what happens has been the same.

Excessive optimism leads to speculation. Speculation leads to leverage (more of a good thing is always better, right?). Then herd instincts (or fear of missing out, or greed, or envy) pull more investors in, pushing valuations even higher. Ultimately, it results in a disconnect from fundamentals where promises exceed economic reality.

Let’s look at the example of the internet. When the backbone of the internet was constructed, companies like Global Crossing (established in 1997) set out to connect continents to the web by laying undersea transmission cables. By the year 2000, Global Crossing’s market capitalization had grown to $47 billion. Yet just two years later, the company filed for bankruptcy. Even though Global Crossing had great revenue growth, it never turned a profit.

After numerous lawsuits, Global Crossing was eventually acquired for $3 billion in 2011 by Level 3 Communications, which was then acquired by Lumen Technologies in 2017. Lumen has had only two profitable years since – 2017 and 2021.

Cisco Systems was also a big player in the backbone of the internet. The company’s network routers connect companies, governments and individuals to the internet. 

Everyone wanted and needed their product, and sales and revenues boomed. The company’s market capitalization at the end of 2000 was $448 billion (the present value in today’s dollars, using the return of the S&P 500 as the growth rate, would be $4.33 trillion; Nvidia’s market capitalization today is $4.44 trillion). Cisco had gross operating margins of 65% (Nvidia’s gross operating margin today is 75%). Cisco was going to be the first company to have a half trillion-dollar market cap, but it never made it. Now, 25 years later, the company’s market cap is $306 billion.

Sun Microsystems, meanwhile, introduced the JAVA programming language that revolutionized software development. Sun reached a market capitalization of $200 billion. Oracle purchased Sun in 2009 for $7.4 billion.

Or, take the example of America On Line (AOL), which was the biggest internet service provider in 2000 with 20 million customers. The company’s instant messaging feature became a cultural phenomenon, and the phrase “You’ve got mail” was synonymous with AOL. Time Warner purchased the company for $106 billion in 2001. Verizon acquired AOL from Time Warner in 2015 for $4.4 billion. AOL was merged with Verizon-owned Yahoo in 2017, and Verizon ultimately sold the package to Apollo for $5 billion in 2021.

At the other end of the spectrum is a company like Nvidia. Yes, Nvidia was around at the same time as the companies mentioned above. Nvidia’s initial public offering was in January 1999, and the company’s market capitalization by the end of 2000 was $1.16 billion. Nobody paid any attention to it. Today, of course, Nvidia is the world's most valuable company…for now.

For comparison, Google’s initial public offering was in August 2004 with a valuation of $23 billion. Google didn’t exist during the buildout of the internet backbone, but is one of the largest companies in the world today with a market capitalization of $3.8 trillion.

I am taking a long time to arrive at my point, which is that titans at any point in time may not be titans in the future. Some of the most valuable companies today were not even involved in the buildout of the internet, but managed to harness the resulting utility.

What Does This Mean for AI?

Now we know we have seen this movie before, and today’s market theme is AI. Is Google’s AI, called Gemini, any better than other models, like Grok or ChatGPT or Anthropic’s Claude? These models seem to be “trained” by scraping everything on the internet and social media. How much of that data is fact vs. fiction, or opinion vs. journalism? Who or what decides?

If an individual has a poor dietary intake, there are consequential risks, such as cholesterol, blood pressure, diabetes or weight gain. If AI models are trained by internet and social media data, I am afraid it may be the equivalent of a poor diet, complete with consequential risks. Those risks will be less evident to users and may become accepted as reality even though they could be factually and blatantly incorrect.

In the end, are these AI offerings just a homogeneous commodity, with only one bound to survive? Will Meta, Apple, Google, Microsoft, TikTok and the rest allow a single AI assistant to access their system protocols, or will it be blocked in favor of their respective individual offerings? You can probably tell by this point that I feel the development of a usable AI may take longer than many of us believe. For now, it is perhaps more of a useful search assistant than anything else.

AI Developments in 2026

Has AI reached the point where investor frenzy has disconnected from economic reality, similar to what happened with Global Crossing, Cisco and other cautionary examples from the 1990s or early 2000s? Indeed, that is the last step of the frenzy process.

Will future AI revenues support the massive investment up to this point? Will the leaders of the backbone buildout also be the ones to monetize the platform moving forward? Which companies will be the Global Crossing and wither away? Which companies will be Google or Microsoft and harness the future of the platform moving forward?

In our view, one of the most important questions for 2026 is whether AI develops in the manner and speed as currently envisioned (and expected). Over the past couple of years, AI has been the primary driver of stock market returns, and as a result, expectations – and valuations – are high. If the AI trajectory is altered in a negative manner, it could have a significant downward impact on the market. For investors, this underscores the importance of discipline, diversification, and keeping a long-term perspective on your overall plan.

This is part one of Greg’s 2026 outlook. Read part two here.

Greg Sweeney is the Chief Investment and Economic Strategist at Bell Institutional Investment Management. He guides the investment strategy, and this outlook is his perspective on the latest market trends and what they could mean for investors. Any views, strategies or products discussed in this article may not be appropriate or suitable for all individuals and are subject to risks.