Comparing to the internet bubble, has AI now reached "1998"?

```

If history is a mirror, the AI wave is reflecting the shadow of 1998. The Wang Xueheng team from Guosen Securities released their annual strategy report on the 27th, stating that the current stage of AI industry development is highly analogous to the early 1998 period of the Internet bubble in terms of timing—namely, the late “1→N” phase.

Compared to the Internet bubble period, the current AI wave in terms of spatial development has only reached about one-third of the limit of that time. Data shows that since October 2022, the S&P 500 has outperformed its fundamentals by about 27 percentage points, far lower than the peak of 85 percentage points during the Internet bubble period. The degree of style divergence between Nasdaq and S&P 500 is also only about one-third of that time. This means that although recent market concerns about high valuations have intensified, from a historical perspective, the valuation pressure of the current AI rally has not reached its extreme.

This assessment directly affects investors by recalibrating expectations for the next two years. The report forecasts that the “1→N” phase of the AI wave will end around mid-2026, at which time the market may experience a moderate correction, with the S&P 500 possibly pulling back to around 6,700 points. However, this is not the end of the rally, but rather the prelude to the “N→N+” stage. In the second half of 2026, as the bubble builds up, the overall market—especially tech stocks—are expected to accelerate, potentially pushing the S&P 500 to a high of 9,600 points by the end of 2027.

Therefore, the first half of 2026 will be a key window for investors to adjust portfolio structures and optimize AI position costs. Unlike the fragile fundamentals during the Internet bubble, this round of the AI wave shows that Nasdaq has stronger profitability support, consistently outperforming the S&P 500 with an expected widening of gains in the future.

The Essence of the Internet Bubble: A Feast Fueled by Capital

The research report notes that looking back, the Internet bubble was not merely an industry carnival, but rather the product of a mismatch between capital and asset supply and demand under special historical circumstances. At the time, the US was enjoying the “peace dividend” after the Cold War; fiscal surpluses dramatically reduced US Treasury supply, resulting in a severe “asset shortage.” Meanwhile, deposit migration and pension reforms (401k) triggered a boom in non-bank finance, compounded by the influx of massive foreign funds due to trade liberalization, bringing a flood of capital into equity markets.

This abundance of capital coincided with the rise of the Internet industry narrative. Due to cognitive bias, the market severely underestimated fiber optic technology supply capabilities, mistakenly believing that network resources would remain scarce long-term, which underpinned aggressive capital expenditure plans. This expectation of “overestimated demand, underestimated supply” allowed financing to transform into industry profits, attracting further investment and resulting in a self-reinforcing positive feedback loop. In the end, the bubble was not burst by high debt leverage, but rather by a concentrated sell-off of primary market equity assets in early 2000 and the reality of supply surplus in fundamentals.

The Anchor of Time: AI in the Late “1→N” Phase

Turning back to today, by comparing key indicators, the report believes that the AI wave’s timeline has reached a point similar to 1998. Looking at industry chain revenue growth, the Internet bubble went through preparatory, “0→1,” “1→N,” and “N→N+” phases. Nvidia, as the upstream representative in this AI wave, has a data center growth curve closely mirroring Cisco of that era. If we extrapolate based on Cisco’s historical trajectory, the current AI wave is in the late “1→N” phase, where revenue growth is marginally tapering but still at a high level.

Another comparison is between AI cloud-focused CoreWeave and AOL of that period. The revenue growth rates show that the AI industry is currently in the application expansion phase. Overall, this period is expected to last until mid-2026. Based on the market performance in 1998, the “1→N” phase typically ends with a moderate correction before entering the most frenzied, bubble-rich “N→N+” phase.

Space Measurement: Valuation Pressure Only a Third of Bubble’s Peak

Though AI concept stocks have risen sharply, their degree of bubble inflation is far less than in the Internet era. The report uses an “intrinsic return model” comparison, finding that the S&P 500 in this cycle has outperformed fundamentals (cumulative reinvestment rate) by about 27%, compared to 69% at the end of the “1→N” phase and as much as 84% at the peak during the Internet bubble. In other words, current market valuation premium is only about a third of the extreme level during the Internet bubble.

The style divergence further confirms this. Since October 2022, the annualized excess return of the Nasdaq over the S&P 500 is about 6 percentage points. In comparison, this was 18 percentage points during the Internet bubble. The difference is mainly due to improved fundamental support: in the current AI wave, the Nasdaq’s component stocks have much stronger profitability than the overall S&P 500, with fundamentals providing solid support rather than pure sentiment-driven speculation.

Looking to 2026: Accelerating After Adjustment

Based on these comparisons, the market’s future path is becoming clearer. The report believes that the first half of 2026 will see the end of the “1→N” phase, with the S&P 500 facing a potential adjustment similar to mid-1998. Estimations show that the index may pull back to around 6,700 points, with a sideways consolidation. This will be a good opportunity for investors to re-evaluate positions and prepare for the next stage.

After passing through this adjustment stage, the market is expected to enter the “N→N+” phase, leading to an accelerated rise similar to late 1999 through early 2000. Under base-case assumptions, by the end of 2026, the S&P 500 may reach a neutral level of 7,500 points. If sentiment plays out further, by the end of 2027, the index could even hit 9,600 points. During this period, the Nasdaq index is expected to maintain its excess gain over the S&P 500, possibly reaching around 23,500 points by mid-2026 and further widening its lead. Of course, this path could be affected by several uncertainties such as the Fed’s monetary policy, macroeconomic outlook, and the pace of AI technology advancement.

Risk Warning and DisclaimerMarkets have risks. Investment requires caution. This article does not constitute personal investment advice, nor does it take into account the specific investment objectives, financial situation, or needs of any individual user. Users should consider whether any opinions, views, or conclusions in this article are suitable for their particular situation. Investing based on this information is done at your own risk. ```