Goldman Sachs warns: The AI frenzy may repeat the 1999 internet bubble; five major signals to watch out for

Author: Jin10 Data

The market is currently worried that the US technology sector is showing signs reminiscent of 1999. Although debates about whether AI is a bubble are intense, history offers some signals that reveal what investors should specifically focus on.

Goldman Sachs strategists say they believe the market’s AI frenzy is at risk of repeating the dot-com bubble burst of the early 2000s.

In a report to clients on Sunday, Dominic Wilson, senior advisor for global market research at Goldman Sachs, and Vickie Chang, macro research strategist, wrote that US stocks do not yet resemble 1999. However, they noted that the AI boom is increasingly resembling the frenzy of the early 2000s.

“We see increasing risks that, as the AI investment craze continues, the imbalances accumulated in the 1990s will become more apparent. Recently, some echoes of the late 1990s boom have appeared in the market,” the bank wrote, adding that current AI trading looks similar to tech stocks in 1997, a few years before the bubble burst.

Wilson and Chang pointed out several warning signals that appeared before the dot-com bubble burst in the early 2000s, which investors should be wary of.

1. Investment Spending Peaks

In the 1990s, investment in technology equipment and software surged to “extraordinarily high levels,” peaking in 2000, when non-residential investment in telecommunications and technology accounted for about 15% of US GDP.

According to Goldman Sachs analysis, investment spending began to decline in the months leading up to the dot-com bubble burst.

“Therefore, high valuation asset prices had a significant impact on real spending decisions,” the strategists said.

Since the beginning of this year, investors have become increasingly cautious about large tech companies’ massive spending on AI. Amazon, Meta, Microsoft, Alphabet, and Apple are expected to spend about $349 billion on capital expenditures in 2025.

Goldman Sachs states that technology investment peaked in the early 2000s, just as the internet stock bubble was beginning to burst.

2. Corporate Profits Begin to Decline

Corporate profits peaked around 1997 and then started to decline.

“Profitability peaked long before the boom ended,” Wilson and Chang wrote. “Although reported profit margins were stronger, in the late stages of the boom, a decline in profitability in macro data coincided with accelerating stock prices.”

Currently, corporate profits are performing strongly. According to FactSet data, the S&P 500’s blended net profit margin in Q3 was about 13.1%, higher than the five-year average of 12.1%.

Corporate profits peaked at the end of 1997, a few years before the bubble burst.

3. Rapid Increase in Corporate Debt

Corporate debt as a percentage of profits peaked in 2001.

Before the dot-com bubble burst, company debt levels increased. Goldman Sachs analysis shows that the percentage of corporate debt relative to profits peaked in 2001, right at the time of the bubble burst.

“The combination of rising investment and declining profitability pushed the corporate sector’s financial balance—i.e., the savings-investment gap—into deficit,” the strategists said.

Some large tech companies are financing their AI expenditures with debt. For example, Meta issued $30 billion in bonds at the end of October to fund its AI spending plans.

However, Goldman Sachs adds that most companies now seem to be financing capital expenditures with free cash flow. The ratio of corporate debt to profits is also well below the peak levels seen during the internet bubble.

Compared to 2000, the debt-to-profit ratio appears quite low.

4. Federal Reserve Rate Cuts

In the late 1990s, the Federal Reserve was in a rate-cutting cycle, which was one of the factors fueling the stock market. Goldman Sachs wrote, “Lower interest rates and capital inflows fueled the stock market rally.”

The Fed cut rates by 25 basis points at its October policy meeting. According to the CME FedWatch tool, investors expect the Fed to cut rates again by 25 basis points in December.

Other market experts, such as Ray Dalio, have also warned that the Fed’s easing cycle could contribute to market bubbles.

5. Widening Credit Spreads

Credit spreads widened in the early 2000s.

The bank pointed out that credit spreads expanded before the dot-com bubble burst.

Credit spreads—the difference in yields between bonds or credit instruments and benchmark rates like US Treasuries—widen when investors perceive higher risk and demand higher compensation.

While credit spreads remain near historic lows, they have started to widen in recent weeks. The ICE BofA US High Yield Index Option-Adjusted Spread rose to about 3.15% last week, up 39 basis points from the low of 2.76% at the end of October.

Wilson and Chang noted that these warning signals appeared at least two years before the internet bubble truly burst in the late 1990s. They added that they believe there is still room for AI-related trading to rise.