Bank of America Outlook 2025: The most aggressive policy in a century in the United States will fuel a "tech bubble," but "the higher you climb, the harder you fall."

After the chief strategist of JP Morgan, Marko Kolanovic, resigned in disappointment and Mike Wilson, the chief US equity strategist at Morgan Stanley, was demoted and turned bullish, the biggest bear on Wall Street has become Benjamin Bowler, a strategist at Bank of America. He leads a stock derivatives team that has conveyed pessimistic forecasts to clients year after year.

This year is no exception. In the recently released report "2025 Outlook: The Roaring 2020s?", Bowler's team warned that the biggest risk in 2025 is that investors misjudge the "tech bubble," which could be exacerbated by the most aggressive policy shift in the US in a century.

The team wrote that the easing and tax cuts of the 1980s, along with the internet bubble of the 1990s, drove a record stock market boom, but the subsequent crashes of 1987, 2000, and 1929 taught us one thing: "The higher you climb, the harder you fall."

According to Bowler's team's analysis, 2025 could be a year where the "AI bubble and Trump policy experiment collide," a situation not seen in the US since the 1920s. History shows that either of these two factors could drive the US stock market sharply up or down, and now, the market may face both factors simultaneously, leading to even more volatility.

Investors have yet to realize a key risk - the simultaneous impact of the AI bubble and Trump policies

Bank of America believes that the key risk in 2025 is that investors have "not yet realized this is a historic moment," and how significant the tail risks are.

The bubble in the field of artificial intelligence is already large enough, and AI concept stocks may accelerate the bubble or may give back a significant amount of gains, both of which could lead to severe market volatility. Coupled with the most aggressive public policy shift in a century, the market remains very fragile. Meanwhile, macroeconomic uncertainty in the post-pandemic era remains high, and the highest debt levels in the US in a century will only exacerbate market volatility.

What is striking about their argument is that whether it is a stock market bubble or a large-scale policy shift, both will lead to significant fluctuations in asset prices, "and 2025 may witness both situations simultaneously, which may not have been seen since the 1920s."

This means that whether the US stock market fluctuates upwards or downwards, its volatility will be amplified. The historic shock of the VIX this summer indicates a structural fragility in the market, which means the US stock market may be more susceptible to shocks and may have already reached a critical point.

Meanwhile, Bowler's team warns that the mistake investors may make is to "underestimate the risks and remain cautious," hoping the market will provide more clarity and certainty, but this expectation may never be realizedTheir advice is: The key is to act early and recognize that we are in a period of rising risks and rising returns. Although few can accurately predict the market top or the timing of the next major shock, the Bowler team points out that the current market environment (such as reduced volatility after the U.S. elections) and price dislocations across asset classes provide opportunities for investors.

If Trump fulfills his campaign promises...

The Bowler team believes that if Trump fulfills his campaign promises, the U.S. will see the largest policy easing since the Reagan administration, the largest tariff increase since the Smoot-Hawley Act of 1930, and the largest illegal immigrant deportation in U.S. history.

While such policies may bring a brief boom to U.S. stocks, they are often followed by a great depression, such as the Great Crash of 1929, the crash of 1987, and the bursting of the internet bubble in the early 2000s.

The report notes that the largest increases in the U.S. stock market before bear markets since 1921 are often accompanied by the largest declines in the following year.

AI—The tech stocks of the internet era?

Next, we turn to the key support for tech stocks—the "chatbot" revolution. The Bowler team believes that the current AI boom has similarities to the internet bubble of the late 1990s.

At that time, the internet bubble lasted for more than 5 years, during which the Nasdaq 100 index grew 11 times, and as the bubble developed, both risks and returns generally increased.

Now, the AI rebound has already "gone too far." In just two years since the launch of ChatGPT, the Nasdaq 100 has risen by 85%, while the "seven giants" stocks have surged by 200%.

From a technical and valuation perspective, Bank of America believes that today's volatility and valuations indicate that tech stocks are closer to the state of 1996/1997 than to 1998/1999.

However, macroeconomic and historical vulnerabilities may undermine this optimism.

The report points out that U.S. debt/GDP is currently close to a record 120%, compared to just 30% in the 1920s, even exceeding the peak during World War II. With higher macro uncertainty and policy risks, prolonged interest rates, and potentially higher risks of policy shifts in budget deficits (higher debt service costs amplifying deficits), there is a risk that the bond market will retaliateThe Bowler team wrote, "This is the biggest visible left-tail risk today, which will be exacerbated by the already high valuations due to the AI boom so far." And as Bank of America said, after a great boom, a great depression is inevitable.

When Will the AI Bubble Burst? Let the Bullets Fly a Little Longer

Interestingly, Bank of America believes that AI is already a bubble, while also predicting it will grow into a larger bubble.

The core argument of Bank of America is that historically, when technological developments signal significant productivity growth, they often lead to the formation of asset bubbles. This is because investors are overly optimistic about the transformative and growth potential of new technologies, leading them to pay high prices for these expected future returns.

Three historical periods are cited as examples: the stock boom of British railways in the 1830s-1860s, the radio and automobile boom in the United States in the 1920s, and the internet bubble of the 1990s.

In light of this, Bank of America points out that the current AI boom has similarities to the internet bubble of the late 1990s, but there has not yet been a significant increase in volatility similar to that seen in the late stages of the 1990s internet bubble.

Moreover, technology valuations are still far from the peak levels seen during the internet bubble of the 1990s.

A notable feature of the internet bubble of the 1990s was its resilience to macro risks. At that time, despite experiencing significant macro crises such as the Asian financial crisis (1997) and the failure of Long-Term Capital Management (1998), as well as the Nasdaq entering a bear market and the Federal Reserve raising interest rates by over 100 basis points, the internet bubble continued to amplify.

If the Federal Reserve further cuts interest rates in 2025 as the market expects (assuming the economy does not collapse), this could increase the upside risk in the market and fuel the bubble.

The current AI boom differs significantly from the internet bubble of the 1990s, particularly in terms of the concentration and scale of market leaders. Today's AI "leaders" are massive, holding a larger market share compared to internet companies in the 1990s

For the biggest beneficiaries of the AI boom, such as Nvidia, reaching the valuation levels recorded during the 1990s internet bubble may be more difficult. If Nvidia reaches the valuation that Cisco had at the peak of the tech bubble, its market capitalization would reach about $11 trillion, accounting for more than 45% of the current U.S. GDP, a scenario that is extremely unlikely to occur in reality.

Therefore, beneficiaries of the AI bubble may need to expand to smaller companies to generate similar valuation effects. As the report states, the current momentum of the AI boom may not yet have reason to exhaust, and the bubble may last for some time.

The Bowler team also pointed out that the record concentration of U.S. stocks has led to a historic concentration of active risk. If stock volatility rises further and concentration increases, this will further amplify the challenges of managing stock portfolios relative to market-cap-weighted benchmarks as asset bubbles expand. In the S&P 500 index, the contribution of tech stocks to the overall market return volatility (or risk) has reached record levels.

Thus, as the bubble progresses, this risk may further increase. Additionally, individual stocks are becoming increasingly fragile, making it more necessary to manage active risk around key catalysts such as earnings. The good news is that the options market is increasingly underestimating this risk, providing potential arbitrage opportunities for investors.

Both Market and Individual Stock Fragility Are Soaring

Finally, turning to fragility, as the Bowler team pointed out, the structural fragility of U.S. stocks has significantly increased over the past few decades. This is due to investors tending to flock into limited momentum trading (buying on the rise and selling on the fall), facing liquidity tightness when exiting the market, especially when liquidity is most needed.

Bank of America first proposed the view of a market shift towards a more peaked distribution in its 2016 outlook, meaning that the market would experience longer periods of calm followed by larger fluctuations. The S&P 500 index has experienced two of the four largest fragility shocks in nearly a century, and its fragility has increased fivefold since the global financial crisis.

More importantly, the fragility of individual U.S. stocks is also rising, with both frequency and magnitude approaching extreme levels not seen in thirty years. Fragility typically spikes around earnings announcements.

Bank of America warns that with the rapid development of AI technology, the valuations of tech stocks have become very high, which may lead to excessive market concentration, thereby increasing market vulnerability. Additionally, the Trump administration may relax regulations on AI and strengthen export controls, which will further elevate market uncertainty and risk