From the Kosovo War to the Dot-Com Bubble Burst: Will AI Repeat the Same Mistakes?

War crises lead to increased trading activity during recessions, but signals of a trend “exit” still require observation.

First, the US recession cycle typically occurs at the end of a rate hike cycle; during rate cut cycles, recessions are less likely to form. Second, even when entering a rate hike cycle, the probability of an economic recession is 6/9, while the market declines during this period are 2/9. Additionally, the peak of the index after rate hikes usually exhibits a “lag,” averaging 17.5 months.

It is important to note that in the past, wars involving oil-producing countries that triggered rate hike cycles saw index peaks roughly synchronized with the initial rate hike. Compared to the oil crises of the late 1990s, rate hike cycles, and industry cycles, there are many similarities to the current situation.

From “Blonde Girl” to “Dot-com Bubble”: Strengthening and Shifting Narratives.

(1) In the mid to late 1990s, the “Blonde Girl” narrative still dominated. After the Fed’s “soft landing” of rate hikes in 1994–1995, the economy rebounded, wage growth accelerated, and inflation showed signs of rising, with the CPI year-over-year exceeding 3% for three consecutive months by late 1996. The Fed raised rates once in March 1997 but was soon suppressed by a strong dollar, low oil prices, globalization, and the IT revolution. Subsequently, the Asian financial crisis triggered global deflation and financial risks, forcing the Fed to pause rate hikes and cut rates urgently in 1998.

(2) Geopolitical wars and oil prices had a clear impact on the “Blonde Girl” narrative. After the Asian financial crisis, the global deflation cycle reversed, and commodity prices began to recover. Coupled with the Kosovo War (March–June 1999), oil prices rose from $10 to over $30 per barrel. At this time, US inflation also surged, prompting the Fed to start a rate hike cycle in June 1999.

(3) However, the market did not immediately price in the end of the “Blonde Girl” narrative. The sequence—(99.3–99.6) Kosovo War -> oil prices rise -> Fed rate hikes (99.06–00.05) -> economic suppression -> dot-com bubble burst (00.03)—was not instantaneous. In fact, after starting rate hikes in June 1999, the Nasdaq still rose 91%. The tech sector led the market throughout. From 1997 to 1999, the Nasdaq’s annual gains were 21.6%, 39.6%, and 85.6%.

How the Dot-com Bubble Ended: Essentially, overvaluation drained the fundamentals, but after the Y2K-driven upgrade wave failed to sustain, the tech boom could not be maintained.

The burst of the dot-com bubble was a concentrated eruption after ignoring several negative signals beforehand. Before the crash in March 2000, aside from rising inflation, tightening liquidity, and policy reversals, there were also negative signals in the tech industry and leading companies’ fundamentals since 1999.

In fact, the fundamentals of tech giants like Cisco, Intel, and Microsoft showed signs of slowdown as early as 1998–1999. The high ROE (over 40%) in 1997–98 was the peak; thereafter, ROE began to decline. In Q3 1998, these companies’ ROE temporarily stabilized but then resumed downward from late 1999, with a central tendency shifting lower.

Profit growth, which had been 50–80% before 1997, sharply declined in 1998–99. The Y2K upgrade demand briefly boosted profits in 1999 but further declined in 2000.

2. The 2000 Y2K upgrade wave did not materialize as expected; preemptive hardware purchases exhausted future demand, leading to a slowdown in hardware capital expenditure and high inventory levels in the industry.

Between 1998 and 1999, companies worldwide accelerated IT spending to prevent Y2K issues, but the anticipated upgrade wave did not occur. This led to reduced demand, order shrinkage for giants like Dell, HP, IBM, and record-high semiconductor inventories by late 2000.

3. In February–March 2000, reports of antitrust investigations into Microsoft intensified, negatively impacting industry sentiment.

In early 2000, US media extensively covered the Department of Justice’s antitrust case against Microsoft, which further dampened market confidence.

4. From March 2000 onward, earnings reports showed poor performance for tech companies during the holiday season.

Microsoft’s April earnings missed expectations, citing slower PC market growth; IBM’s Q1 revenue was also below forecasts. Several once-hot new tech companies went bankrupt, such as Boo.com, APB Online, and Value America.

(4) After the Iran–US conflict: changes in logic and subsequent projections.

In the short term, “geopolitics is unpredictable,” and developments are highly uncertain and random.

In the medium term, the situation’s visibility and predictability increase. Currently, US stock valuations are only “one step away” from the dot-com bubble peak, with long-term US Treasury yields remaining resilient. Both stocks and bonds are highly sensitive to global inflation expectations. Historically, before the 1990s dot-com crash, tech stocks led the market throughout, even during rate hike cycles, with no systematic style shift until the bull market ended, when investors moved to safe assets.

Thus, current conflicts and oil price surges may cause temporary shocks to the tech industry, but based on current valuations and earnings expectations, the industry trend is unlikely to reverse.

Furthermore, considering the 2026 US midterm elections, where Trump has indicated that “price” will be a key issue, and the Republican focus on “cost of living,” ongoing war developments that intensify could be politically unmanageable before the elections.

Previously, Trump’s State of the Union emphasized lowering energy prices as a major achievement, and he explicitly stated that the 2026 midterms would revolve around “America’s economic success” and “cost of living.” If gasoline prices rise from below $3 per gallon, it could undermine voters’ confidence in the government’s inflation control, affecting the midterm results.

From a high-probability perspective, the logic of a global non-US asset bull market in 2026 is unlikely to be overturned by geopolitical tensions. We remain optimistic about Chinese stocks, expecting a “Davis double play” with non-US markets. Therefore, once short-term geopolitical risks subside, the market may present the best bottom-fishing opportunity of the year.

Content has been abridged.

Source: Chenming’s Strategic Deep Thinking

Risk Disclaimer and Terms of Use

Market risks exist; investments should be cautious. This article does not constitute personal investment advice and does not consider individual user’s specific investment goals, financial situation, or needs. Users should consider whether any opinions, views, or conclusions herein are suitable for their particular circumstances. Invest at your own risk.