“It’s different this time” may well be the four scariest words in the investment world.

Mark Twain was quoted as saying “history doesn’t repeat itself, but it often rhymes” — yet investors often ignore older news.

If you’ve been reading my columns over the past two years, you know I am careful in not trying to make short-term market predictions but rather suggesting the merits of taking a long-term perspective when investing.

So it seems reasonable to take Warren Buffet’s advice to take a critical look at markets when things look either scary or euphoric.

Especially when I start hearing those four scary words from the “experts” — and with all of the hype about artificial intelligence (AI) and comparisons to the internet tech era, I’m hearing them a lot.

Note that I tend to believe there are good odds that AI will truly be transformative (in fact I use Perplexity and ChatGPT often to research my columns), but I suspect equally good odds of the road ahead having some sizable ups and downs.

Here’s why:

As of this writing, there were 357 stocks with year-to-date returns of more than 100%.

Some of these are very small startups, so let’s look at companies within the large cap S&P 500. There are stocks of six companies up more than 100% YTD; 31 companies with YTD returns of more than 50%; and 126 up more than 20% YTD.

And we are seeing some cracks within the “magnificent seven” with Meta, Microsoft and Nvidia are all up more than 20%, but Apple and Amazon having double-digit negative returns and Amazon and Alphabet showing just modest gains so far this year. 

The tech-heavy NASDAQ 100 is up almost 18% YTD with the Top 10 all up more than 60% YTD.

Palantir Technologies is up 136% YTD with a forward P/E ratio of 210 and earnings estimates of 84 cents on a stock price of $179.

Legendary chipmaker Intel is up 78% YTD, selling at 54 times projected forward earnings even after seriously struggling for quite some time.

Note: We are all shareholders in Intel as the U.S. government now has a 10% stake in the business to protect this chip foundry for national security reasons.

We’ll see how this controversial deal involving government ownership will turn out. Analysts, including Bank of America, have recently downgraded the stock, highlighting ongoing challenges in its CPU lineup and AI strategy.

As you might suspect, many of these incredible returns are for AI-focused companies. A very recent story in Fortune magazine noted “75% of gains, 80% of profits, 90% of capex—AI’s grip on the S&P is total and Morgan Stanley’s top analyst very concerned.”

There are also growing concerns about “vendor financing” as illustrated in the nearby CNBC chart.

Simply put, OpenAI places a massive order for chips from Nvidia — but doesn’t have the cash to pay for them. Nvidia simply invests $100 billion in OpenAI so it can buy the chips.

And not to be outdone, AMD entered into a deal giving warrants (“options”) that would allow OpenAI to pay for chips with money from (hoped) AMD stock price growth.

This is not a new idea: Boeing has financed jets for airlines for years. The difference is that airlines make money and can pay for purchases while these AI developers have yet to make money.

It’s creative but it introduces financial risk and complexity reminiscent of the dot.com era. Kind of reminds me of Wimpy’s “I’ll gladly pay you Tuesday for a hamburger today.”

Please note that I’m not playing Chicken Little and predicting the sky is falling. But there are enough signs of an AI bubble that may at least “rhyme” with the dot.com bubble a quarter-century ago.

Much of this will depend on the ability of AI to become a “general purpose technology” (GPT) such as electricity did a century ago and the personal computer did 50 years ago.

History suggests two conditions generally precede the mass adoption of transformative technologies like this:

• Surging capital investment in the new technology.
• A period of sluggish adoption, productivity growth and profitability — with big winners and losers.

It seems clear that the first condition is being met in spades as companies are spending trillions of dollars on AI.

But the question remains as to whether AI will be widely adopted — and if so, how long that will take and whether it will be profitable.

Remember, it took about two decades from Thomas Edison’s light bulb invention until electricity was widely used.

And if AI becomes widely adopted, which companies will be the winners and losers in this incredible race?

Don’t forget all the dot.com companies that went bust or even the smaller number of big winners like Amazon that still saw huge drawdowns in stock prices before becoming highly profitable.

For those of you nonhistory buffs, Amazon stock was $4.30 a share on April 30, 1999, and fell by 93% to 30 cents on Sept. 28, 2001.

It’s easy with the benefit of 20/20 hindsight to see that riding through this horrifying decline would have paid off handsomely.

Or how about Cisco Systems, which hit a peak price of $77.31 on March 31, 2000, and dropped 86% to $10.48 by Sept. 30, 2002?

Both survived, which is more than you can say about a lot of dot.coms, but ask yourself if you really could have stuck with these investments while watching values decline so much so quickly?

And note that while Amazon’s stock price has increased to more than $200, Cisco is still selling below its 2000 peak.

Bottom line: without that magic crystal ball everyone yearns for, one can only guess about the future.

I am generally optimistic for stocks in general and AI in particular over the long run. But in my mind, it seems prudent to make sure you’re diversified and make sure you’re buckled up for the potential of a wild ride.

It may not really be different this time.