Deutsche Bank sees echoes of dot-com era as software selloff prompts market rotation

The stock market experienced a notable shift this week, as investor sentiment began to re-evaluate the intrinsic value of corporate software and IT services. This re-evaluation comes amid growing speculation that artificial intelligence could increasingly handle these functions internally, potentially reducing the need for external solutions. While the market has not seen a significant overall decline, analysts point to a clear rotation of investments into other sectors, a pattern that may feel familiar to long-term observers of financial markets.

This recent movement, particularly away from IT and Software as a Service (SaaS) assets, followed an update from Anthropic. The company launched plug-ins for its Claude Cowork agent last week, tools designed to streamline operations in areas such as data analysis, legal, marketing, and sales. The implications of such advancements appear to have prompted investors to reconsider their positions in traditional software providers.

Despite the sector-specific rebalancing, the broader market impact has remained relatively contained. Neither the S&P 500 nor the Nasdaq experienced declines exceeding 2% recently, and European and Asian markets largely maintained their stability. However, even a relatively minor cycling out of certain affected sectors and into others draws comparisons to past technological shifts that reshaped market dynamics.

Henry Allen of Deutsche Bank highlighted to clients that the software component of the S&P 500 has fallen nearly 30% from its October peak. Given this drop, one might expect a broader market correction. Yet, the overall S&P 500 closed only 2.6% below its record high from the previous month. This resilience, Allen noted, is due to a significant rotation, where other sectors, including energy, materials, and consumer staples, have absorbed the capital flowing out of technology.

This pattern, Allen explained, bears a resemblance to the market behavior observed in 2000 as the dot-com bubble began to deflate. In that period, equities started falling from March 2000 as technology stocks saw substantial declines. However, sectors like consumer staples, utilities, and healthcare rallied considerably in the subsequent months. By September of that year, the S&P 500 had recovered to within a percentage point of its record high from six months earlier. This historical parallel illustrates how a market can endure a prolonged sector rotation without immediately showing significant stress at the index level. However, Allen cautioned that a deeper and more sustained sell-off in a dominant sector eventually makes it harder for the broader index to withstand the drag, as evidenced by the S&P 500 ending 2000 over 10% lower due to continued tech losses.

While the current sell-off in software is not easing, with Amazon’s recent earnings report reflecting a substantial $200 billion investment in AI, the comparison to the dot-com era does not automatically imply an identical outcome. Eric Sheridan, a senior equity research analyst at Goldman Sachs, pointed out in an October report that discussions around an “AI bubble” are far more prevalent today than conversations about the dot-com or housing bubbles were when they were actively inflating. In the late 1990s, many companies with exuberant valuations had little to no revenue. Today, many of the “Magnificent 7” companies, trading at an aggregate price-to-earnings ratio of 31x compared to the market’s 23x, generate significant free cash flow, engage in stock buybacks, and pay dividends—practices uncommon among the high-flying dot-coms of 1999.

As global GDP becomes increasingly linked to AI capital expenditure and the optimistic outlook for its future returns, investors are naturally scrutinizing when and for whom these bets will ultimately pay off. Jamie Dimon, CEO of JPMorgan Chase, articulated this perspective at the Fortune Most Powerful Women Summit in October, stating that while some individual AI investments might constitute a bubble, AI as a whole is unlikely to be one, and will “probably pay off” in the long run. The current market movements suggest a repositioning in anticipation of this future, rather than a broad retreat.

Ruth Forbes

See Full Bio