The AI-driven surge in technology stocks is sparking intense market debate over whether a "bubble" is forming.

Recently, Goldman Sachs analysts Peter Oppenheimer, Sharon Bell, and others provided a definitive answer in their latest research report: Although the current market exhibits some characteristics of historical bubbles, we are not yet in one. Goldman Sachs believes the key difference from past bubbles is that this round of tech stock gains is primarily driven by strong fundamentals and genuine earnings growth, rather than pure speculative frenzy.

The report acknowledges that investment mania surrounding transformative technologies is a common catalyst for bubble formation. While the market's rising absolute valuations, highly concentrated market structure, and surging capital expenditures by tech giants do share similarities with historical bubble periods, Goldman Sachs emphasizes several crucial distinctions.

First, the technology sector's price appreciation has been supported by solid earnings growth thus far. Second, the leading companies driving the market possess exceptionally strong balance sheets, contrasting sharply with the high-leverage-driven bubbles of the past. Finally, current AI sector competition is primarily dominated by a handful of existing giants, whereas most bubbles typically form during periods when numerous new entrants flood in, creating chaotic competitive landscapes.

While Goldman Sachs views systemic risk as relatively low, the firm simultaneously warns that the current extreme market concentration is "unsustainable," and the heavy dependence on earnings growth means markets could face significant corrections if performance falls short of expectations.

**Elevated Valuations But Not at Bubble Levels**

Regarding valuation concerns most pressing to investors, Goldman Sachs concluded through multi-dimensional comparisons that while tech stock valuations are at high levels, they have not reached the extreme levels of historical bubbles.

The report compares the current "Magnificent Seven" U.S. tech stocks with historical market leaders. Data shows these seven companies have a median forward price-to-earnings ratio of approximately 27x, well below the roughly 52x level of leading companies at the peak of the 2000 tech bubble.

Furthermore, whether compared to Japanese bank stocks during the 1989 financial bubble or U.S. blue-chip stocks during the 1973 "Nifty 50" period, current tech giants' valuations appear more rational.

From a PEG ratio (price-to-earnings relative to earnings growth rate) perspective, U.S. tech stocks' current levels remain below the bubble highs of the late 1990s. Historical earnings-based PEG ratios show this metric reached 3.7x at the tech bubble peak, compared to just 1.7x currently.

Goldman Sachs' dividend discount model analysis also reveals that current U.S. TMT (telecommunications, media, and technology) industry market pricing implies a need for 25% annual dividend growth over the next 10 years. While this level is high, it remains below the 35% growth expectations implied during the tech bubble period.

**Earnings-Driven Rather Than Speculation-Driven**

The biggest difference from past bubble periods lies in the core driving forces behind this rally. Goldman Sachs emphasizes that tech stocks' recent exceptional performance more directly reflects their strong earnings capabilities rather than unrealistic speculation about the future.

The report shows that since 2009, global technology sector earnings per share (EPS) growth has far exceeded that of non-tech sectors, with this gap continuing to widen after the financial crisis.

To further substantiate this point, the report compared current return compositions with those in the year before the 2000 tech bubble burst. During the 1999-2000 period, a significant portion of tech stock returns came from valuation expansion, with earnings growth contributing relatively little.

Currently, earnings growth serves as a key pillar supporting stock price performance, providing a more solid foundation for market stability. The report projects that the combined return on equity for the tech "Magnificent Seven" will reach 46% in 2025, with net profit margins of 29%, far exceeding the 16% net profit margins of dominant companies during the 2000 tech bubble.

**Risk Signals Remain**

Despite an overall optimistic tone, Goldman Sachs has not ignored market risks, particularly surging capital expenditures and record market concentration.

The report notes that since ChatGPT's emergence, "hyperscale computing companies" have rapidly increased capital expenditures, with 2024 spending expected to reach $239 billion, more than double the 2018 level. Historically, technology-driven overinvestment has often led to overcapacity and declining returns, such as in the telecommunications industry during the late 1990s.

The key difference lies in financing methods. Currently, tech giants' capital expenditures primarily come from their abundant free cash flow rather than massive borrowing. Their balance sheets are exceptionally robust, with ample cash reserves and generally negative net debt ratios. This contrasts sharply with 1990s telecom companies that funded investments through large-scale stock and bond issuances, thereby reducing systemic risk to the entire financial system.

The report shows that dominant companies currently hold cash averaging 2.7% of market capitalization, with net debt-to-equity ratios of negative 22%, representing far more stable financial conditions than leading companies during the 2000 tech bubble period.

However, Goldman Sachs also reminds in the report that recent debt financing activities by tech companies have increased, representing a noteworthy shift.

Additionally, market concentration has reached historical highs. Report data shows the U.S. market accounts for over 60% of global equity markets, while the collective market capitalization of America's top five tech companies now exceeds the combined total of Europe's Stoxx 50 index, UK, India, Japan, and Canada markets, representing approximately 16% of global public equity markets.

Goldman Sachs considers this highly concentrated situation "unsustainable" but points out this does not inherently constitute a bubble. Historically, market-dominant industries maintaining leadership positions for extended periods has been the norm. From the early 19th century to present, financial, transportation, energy, and technology sectors have successively dominated markets, with each period lasting several decades.

**Focus on Diversified Allocation**

In conclusion, Goldman Sachs states that markets have not entered a full AI bubble, but investors cannot be complacent.

The report indicates that if investor confidence or patience in AI themes weakens, market correction risks remain. However, given relatively healthy corporate and bank balance sheets currently, the possibility of economy-wide impacts similar to those following past bubble bursts is relatively small.

Facing risks from high valuations and high concentration, Goldman Sachs' core recommendation to investors is "diversification," suggesting approaches across several dimensions:

Regional Diversification: Although markets outside the U.S. have lower tech stock weightings, major indices in Europe, Japan, and China have delivered returns in local currency terms comparable to the S&P 500 this year.

Style Diversification: As interest rate environments change, traditional boundaries between "value" and "growth" styles are beginning to blur, providing more opportunities for cross-style investing.

Sector Diversification: AI's robust development relies on physical infrastructure support, with demand for power, energy, capital goods, and resources creating growth opportunities for other industries.

Intra-Technology Sector Diversification: While focusing on existing giants, investors should also seek the next wave of technology "superstars" that can capitalize on current capital expenditure momentum to create new products and services.