AI Bubble Fears Trigger Global Semiconductor Selloff as SoftBank Plunges 40%

SoftBank Shares Slip 40% as AI Bubble Fears Worsen | AIM

Bank of America's Survey in October revealed that 54% of respondents believed AI-related assets were in a bubble territory. In the first three quarters of 2025, the global tech companies braced for US President Donald Trump's trade tariff storm. Since October, however, the overriding concern has been the threat of an artificial intelligence (AI) bubble. Investors are turning increasingly cautious about the AI companies and their unsustainable valuations. Japanese technology conglomerate SoftBank Group, with deep exposure to AI companies, saw a sharp market correction ever since. Between October 31 and November 26, the company's shares dropped by 40%, resulting in a loss of nearly $50 billion in market capitalisation. The financial turmoil was not caused by a single event but resulted from a confluence of factors. The increasing scepticism was raised by Palantir Technologies, the AI software company, whose shares fell despite promising Q3 results, suggesting concerns over its valuation. The market reaction signalled a fundamental shift in investor behaviour. Strong performance was no longer enough to sustain the meteoric valuation based on the future AI potential. A Yahoo Finance report noted that Palantir traded at revenue multiples that exceed those of many established AI and cloud leaders and questioned whether current prices reflect long-term fundamentals or short-term enthusiasm. Besides, Bank of America's Global Fund Manager Survey in October revealed that 54% of respondents believed AI-related assets were in a bubble territory, and 60% said global equities were overvalued, Bloomberg cited. "Among other signs of rampant speculation, frantic venture capitalists are throwing money at AI startups at multi-billion-dollar valuations without even being told their plans," Ben Inker, partner at GMO, the Boston-based investment management firm, wrote in their monthly newsletter. He further noted that equity investors are increasing the valuation of large corporations by hundreds of billions of dollars through investment deals with OpenAI, a company whose revenues would have to rise a hundredfold to fulfil its commitment. For SoftBank, its investments in AI companies, such as Arm, OpenAI, Perplexity, Databricks, and ByteDance, among others, make it more vulnerable to AI-related fears. Following the Palantir trigger on November 4, its quarterly results on November 5, and the subsequent announcement on November 11 to offload its entire stake in NVIDIA (32.1 million shares) for $5.8 billion and deepen its investment in OpenAI, market fears were amplified. A segment of investors, however, sought evidence of long-term conviction in the chip sector. "SoftBank's softness has occurred as the market questions whether it sold a golden lottery ticket to buy a whole new stack of scratch-offs," Michael Ashley Schulman, partner and chief investment officer at Running Point Capital Advisors, told AIM in an email interaction. Schulman added that this move "turns a category leader into a messy, high beta roulette wheel", which led some investors to mark the shares down due to timing risk. While major AI players continue to report strong demand, the rapid rise in valuations for companies like OpenAI has heightened concerns about timing risk and long-term sustainability, especially as markets reassess the gap between AI hype and financial results. Around this time last year, the AI giant was valued at $150 billion. Currently, its valuation is 5x at $750 billion. SoftBank's decline had repercussions across global semiconductor markets as well, sparking widespread industry debate. The fall occurred at a delicate moment for investors monitoring Q4 earnings signals, movements in global tech stocks and turbulence within the semiconductor sector. NVIDIA's stock price fell by about 3%, while other major tech firms experienced similar declines. Taiwan's TSMC and China's Alibaba also faced declines. Adding to semiconductor share fluctuations, developments by major AI companies competing with OpenAI also played a role. Google's release of Gemini 3 earlier this month, which delivered standout results across coding, reasoning, multimodal tasks and more, generated intense excitement, with many calling it a leap ahead in the AI race. That buzz flagging Gemini 3 as a formidable rival prompted some investors to question whether OpenAI could maintain its dominance. "The stocks are hit by concerns that the competition environment of OpenAI will become tougher after Google's Gemini 3 received strong reviews," Tsutomu Yamada, from Mitsubishi UFJ eSmart Securities Co, told Bloomberg. Schulman observed that in the short run, SoftBank's sale of NVIDIA stock looked like a wise move, as if the latter foresaw the dip and exited at the right time. However, investors didn't fully buy its strategic bets. "Nonetheless, based on past quirks and eccentricities, every new ambitious Softbank move gets priced with scepticism," Schulman said. The SoftBank stock decline had a rippling effect on the global tech stocks. Between November 11 and 26, while SoftBank shares declined by 28.35%, NVIDIA's shares fell by 6.68% and Microsoft was down by 4.56%. Meanwhile, Apple, Google and Meta, which competed with OpenAI, saw their stock prices rise by 0.84%, 9.78% and 1.04%, respectively. Schulman noted that SoftBank may remain in a phase where it has become a meme of itself, replaying episodes of "overconfidence and reinvention." He further noted that when the company swaps its investments for another ambitious tech bet, investors are unsure whether it marks a good strategy or should remind them of past mistakes, such as WeWork, Oyo, Zemu and others. Investors are watching to see whether the company will adjust its risk approach or shift its strategy as these markets evolve in 2026. "AI is a game-changer and will remain a structural growth theme in 2026," Christian Nolting, global chief investment officer at Deutsche Bank, noted in their market outlook report. At the same time, he urges caution: "Overinvestment and electricity shortages could dampen expectations." Similarly, Vanguard Funds noted that while AI spending on infrastructure, chips and data centres could support faster-than-expected expansion, the investment management firm cautioned that it may not guarantee a bull run in equities.

Anxiety over AI spending returns to global markets - The Economic Times

Asian stock markets, including South Korea, Taiwan, and Japan, experienced significant declines on Friday, reversing previous day's gains. Investor anxiety over a potential artificial intelligence investment bubble, fueled by concerns that AI hardware spending might outpace demand, led to a broad sell-off in tech and chip companies. Stocks in South Korea, Taiwan and Japan tumbled Friday, reversing gains from a rally the previous day, as investor anxiety about a potential investment bubble in artificial intelligence bled over to Asia from the United States. Benchmark indexes in South Korea and Taiwan fell nearly 4% in Asia on Friday. The losses were driven by declines in chip companies, including South Korea's SK Hynix, which plummeted more than 8%. Taiwan Semiconductor Manufacturing Co. shed more than 4%. In Japan, the Nikkei 225 index declined 2%. Advantest, a Japanese manufacturer of semiconductor equipment, was down more than 12%, while shares of SoftBank, a big investor in AI companies including OpenAI, sank almost 11%. In recent weeks, investors have been increasingly concerned that tech companies' lavish spending on AI hardware like semiconductors and data centers -- which has helped to fuel a rally in AI-related stocks over the past three years -- might be outpacing actual demand. That has prompted a slump in the shares of many large global technology firms. On Thursday, global tech stocks briefly surged after chipmaker Nvidia unveiled strong earnings, signaling to investors that demand for chips essential to AI projects had not abated. The Nikkei 225 index in Japan saw its strongest performance in a week, thanks to surges in the prices of many of the country's semiconductor-industry companies. But the relief was short-lived, and investors began to lose their nerve again. In the United States on Thursday, Nvidia opened more than 5% higher but lost those gains by the afternoon, ending the day down more than 3%. The S&P 500 tumbled after climbing as much as 1.9% in morning trading, closing the day 1.6% lower. The market rally and subsequent reversal underscore how investor jitters about potential overvaluation in the AI sector are prompting swings in technology stocks, which sway the overall performance of global indexes. During Asia hours, futures trading on the S&P 500, for which Nvidia is the largest component, edged higher, suggesting that the index might open slightly up Friday in the United States.

Asia's Semiconductor Selloff Signals Market Repricing of AI Expectations | Investing.com UK

Asian equity markets faced a sharp correction in semiconductor names on Friday, triggered by a renewed reassessment of global AI valuations and earnings sustainability. Technology-heavy indices across South Korea, Taiwan, Japan, and Hong Kong fell as investors moved to reduce exposure to high-multiple chip stocks. The key risk identified is the misalignment between aggressive AI capital expenditure plans and the near-term cash flow visibility of AI supply chain companies. Asian chip makers are directly exposed to this repricing because the region hosts the backbone of AI hardware manufacturing, including memory, logic, foundry, packaging, and testing. With semiconductors representing a dominant weight in benchmarks such as the Kospi and Taiex, index-level declines mirrored the concentrated weakness in key names. The immediate trigger was the reversal in U.S. semiconductor shares, led by Nvidia (NASDAQ:NVDA), after an initial rally fueled by strong quarterly results. The reversal forced investors to question whether AI infrastructure spending can translate into earnings fast enough to sustain current valuations. Samsung Electronics and SK Hynix, the two largest suppliers of high bandwidth memory for AI training systems, fell as much as six percent and ten percent in morning trade and remained down over five percent. The Kospi dropped around three percent, reflecting heavy foreign outflows. Investor positioning had been extended. Bank of America's recent institutional survey showed that forty-five percent of respondents identified an AI bubble as the top tail risk. Fifty-three percent believe AI stocks are already in a bubble. The same survey highlighted concerns over the financing and magnitude of AI capital expenditure, indicating that sentiment had already been stretched before this selloff. The reassessment is not only valuation-driven. Reports of memory chip buyers diversifying toward Chinese DRAM suppliers raised concerns about pricing power and demand concentration. This adds a structural dimension to what initially looked like a sentiment shift. In Taiwan, the Taiex lost more than three percent, dragged down by Taiwan Semiconductor Manufacturing (NYSE:TSM) and Foxconn Technology Group, both falling more than four percent. Investors reassessed medium-term AI demand and questioned the sustainability of data center spending, highlighting rising inventories across the supply chain. The concern is not about long-term AI adoption but about near-term revenue realization against aggressive capacity expansion. Japan faced an additional layer of pressure. Renewed yen volatility magnified market declines, particularly in exporters sensitive to currency movements. SoftBank Group and Advantest each fell more than ten percent, pulling the Nikkei down over two percent. Investors worried that earnings projections based on a weaker yen could be vulnerable if currency swings persist. Hong Kong's technology segment was not immune. Semiconductor Manufacturing International and Hua Hong Semiconductor both fell more than five percent, indicating concerns that even the broader Asia chip ecosystem could see profit compression from global price competition, inventory normalization, and uncertain U.S. export controls. The Philadelphia Semiconductor Index and related Asian semiconductor ETFs are likely to face further volatility as investors measure short-term earnings expectations against long-term adoption narratives. Spillover into broader equity indices will depend on whether the correction remains valuation-oriented or evolves into a margin compression story from oversupply and inventory buildup. Institutional flows suggest that investors are de rating Asia's semiconductor sector based on near-term earnings visibility rather than denying the long-term AI adoption trajectory. The distinction matters. A valuation correction can stabilize once expectations realign. An earnings compression cycle suggests further downside. In the near term, upcoming earnings from major foundry, memory, and logic chip producers will be critical in recalibrating expectations. December guidance will help investors assess whether AI spending is translating into sustainable order books or merely booked for future capacity. If revenues show sequential growth and inventory trends normalize, markets could stabilize. In the alternative scenario, if spending growth remains front-loaded while revenue recognition lags, earnings estimates may face broader cuts in early 2025, prolonging the sector correction. Medium-term valuation support will depend on data center spending trends from major AI hyperscalers and cloud firms. Any signs of spending moderation could accelerate derating pressure. Investors can maintain selective exposure to AI hardware leaders with proven pricing power and order visibility, but should reduce leveraged or momentum-driven positions in high multiple semiconductor names until earnings guidance confirms the durability of AI infrastructure demand. The key risk to monitor is further inventory build-up combined with signs of demand diversification toward lower cost suppliers, which could signal that pricing power has already peaked.