Masayoshi Son\u2019s sudden U-turn on chipmaking sensation Nvidia is fueling a bull market in questions about what, oh what, the SoftBank founder is thinking.<\/p>\n
The popular theory is that Japan\u2019s richest man is selling his stake<\/a> \u2014 savvily pocketing\u00a0US$5.8 billion \u2014 to bankroll other artificial intelligence (AI) investments.<\/p>\n It follows that Son isn\u2019t making a statement about Nvidia\u2019s future so much as reacting to SoftBank\u2019s constraints. Son, the thinking goes, is merely freeing liquidity to go big on the next boom in the AI capital cycle.<\/p>\n Yet it\u2019s hard not to see Son\u2019s exit as a statement on the super-pricey valuations of all too many\u00a0AI-focused companies. Especially since this follows more and more investors questioning if the AI space is experiencing bubble troubles.<\/p>\n The broader market is questioning the dot-com era d\u00e9j\u00e0 vu being triggered by the AI craze, as evidenced by recent volatility. Here, Michael Burry\u2019s return to the headlines is hardly a welcome development for AI true believers.<\/p>\n The investor made famous by Michael Lewis\u2019 \u201cThe Big Short<\/a>\u201d is betting against AI. He\u2019s doing so based, in part, on his contention that some top tech companies are padding AI-related profits via creative accounting. Specifically, Burry thinks \u201chyperscalers,\u201d or tech giant companies building gargantuan data centers, are underestimating depreciation.<\/p>\n \u201cUnderstating depreciation by extending useful life of assets artificially boosts earnings \u2014 one of the more common frauds of the modern era,\u201d Burry wrote on social media. \u201cMassively ramping capex through purchase of Nvidia chips\/servers on a 2-3 year product cycle should not result in the extension of useful lives of compute equipment. Yet this is exactly what all the hyperscalers have done.\u201d<\/p>\n Goldman Sachs strategist Dominic Wilson sees uncomfortable echoes with the early 2000s\u00a0dot-com bubble<\/a>\u00a0burst in the early 2000s. Though the AI frenzy might not yet be at 1999 levels, the odds are inflating in real time.<\/p>\n \u201cWe see a growing risk that the imbalances that built up in the 1990s will become more visible as the\u00a0AI investment boom<\/a>\u00a0extends,\u201d Wilson says. \u201cThere have been echoes of the inflection point in the 1990s boom lately.\u201d<\/p>\n Looking back two-plus decades, Wilson notes, corporate profits reached a peak around 1997, just before the crash. \u201cProfitability peaked well before the boom ended,\u201d he says. \u201cWhile reported profit margins were more robust, declining profitability in the macro data in the later years of the boom came alongside accelerating equity prices.\u201d<\/p>\n The combination of \u201crising investment and falling profitability pushed the corporate sector financial balance \u2014 the difference between savings and investment \u2014 into deficit,\u201d Wilson notes.<\/p>\n Fast-forward to 2025, says Jared Bernstein, former head of the US Council of Economic Advisers, and the gap between AI earnings potential and spending \u201ccertainly looks bubbly.<\/a>\u201d<\/p>\n David Rosenberg, president of Rosenberg Research, notes that \u201call of a sudden, we are starting to see some tech heavyweights fail to live up to investor expectations. Even a company like Palantir has seen its share price hammered, even though its results did top expectations \u2014 nosebleed valuations got in the way.\u201d<\/p>\n This isn\u2019t the conventional wisdom, of course. As Wedbush analyst Dan Ives puts it: \u201cIn my view, [the selloff] is short-lived. I don\u2019t believe this is a start of a more structural selloff. I think it\u2019s just a lot of nervous, sort of white knuckles and the selloff that we saw \u2026 along with the selloff that we\u2019ve seen crypto and others, it was just a massive risk off.\u201d<\/p>\n Alex Karp, CEO of\u00a0Palantir Technologies<\/a>, one of the companies Burry\u2019s Scion Asset Management is shorting along with Nvidia, called the trade \u201csuper weird\u201d and \u201cbatsh*t crazy.\u201d Yet, as Burry puts it: \u201cSometimes, we see\u00a0bubbles<\/a>. Sometimes, there is something to do about it. Sometimes, the only winning move is not to play.\u201d<\/p>\n But Son\u2019s pivot might be the most interesting about-face of all. Son has long been at the forefront of investing in the \u201csingularity,\u201d or the moment when computers outthink humans, a milestone that\u2019s fascinated futurists since the days of\u00a0Isaac Asimov<\/a>\u00a0and\u00a0Arthur C. Clarke<\/a>.<\/p>\n This is a tantalizing move by a man who claims to operate on a\u00a0300-year<\/a>\u00a0time horizon. Son, after all, fancies himself positioning SoftBank for the centuries ahead.<\/p>\n Son has often referred to himself as the \u201ccrazy guy who bet on the future.\u201d In February, Son joined US President Donald Trump, Sam Altman of OpenAI and Larry Ellison of Oracle in announcing a $500 billion investment in a project to develop AI called\u00a0Stargate.<\/a><\/p>\n Another way of looking at SoftBank\u2019s Nvidia sale, of course, is as a contrarian indicator. In recent years, Son has been better known for his investment misses than his successes. Among Son\u2019s biggest misses was being the most extravagant supporter of\u00a0WeWork<\/a>.<\/p>\n Son erred spectacularly by giving founder Adam Neumann a blank check for his fanciful strategy of building an office-sharing ecosystem in the image of Apple Inc. Moreover,\u00a0Son\u2019s penchant for overpaying for stakes in undeserving startups warped venture capital incentives everywhere and loaded SoftBank\u2019s balance sheet with excessive risk.<\/p>\n By late 2021,\u00a0SoftBank\u00a0was suffering its worst decline since 2006. As\u00a0Son\u00a0said in November 2021: \u201cWe\u00a0are in the middle of a blizzard\u201d of red ink. It was quite a contrast from a year earlier when Son said the Vision Fund was the goose laying \u201cgolden eggs.\u201d<\/p>\n SoftBank\u2019s Vision Fund has stabilized since then. Yet Son is still struggling to regain the Warren Buffett vibe on which he\u2019s been dining out for the last decade. In 2000, Son had the remarkable foresight to hand $20 million to an obscure\u00a0English\u00a0teacher\u00a0in Hangzhou.<\/p>\n By the time Jack Ma took Alibaba public in New York in 2014, SoftBank\u2019s stake was worth more than $50 billion. The feat earned Son a reputation as the Buffett of Japan. The launch of his Vision Fund in 2017 was an attempt to replicate that success over and over again.<\/p>\n The current AI boom would seem to be Son\u2019s big moment. His stepping away from Nvidia could indeed herald the next phase of the AI era. Count Paul Golding, head of tech research at Macquarie Bank, among those who are bullish on Son\u2019s Nvidia sale, rating SoftBank \u201coutperform.\u201d<\/p>\n Golding writes that \u201cthis print continues to build on [SoftBank\u2019s] growth pivot, investing in leading AI and automation companies<\/a>, delivering strong underlying fundamentals, and benefiting from thematic\/secular tailwinds for its equity holdings, and the resilience of its balance sheet.\u201d<\/p>\n Yet Son also could be smelling big trouble in AI town. One question is where the next wave of opportunities might be found. It could just as easily be in China<\/a> as in Silicon Valley.<\/p>\n The AI-verse has never been the same since the \u201cDeepSeek shock<\/a>\u201d back in January. The shockwaves the Hangzhou-headquartered company sent through global markets generated the best headlines Xi Jinping\u2019s economy has had in a long while. Then came news that Alibaba is going big on AI.<\/p>\n The sudden arrival of Made-in-China sensation DeepSeek pulled the rug out from under Wall Street\u2019s \u201cTrump trade\u201d party. That excitement was based on Trump\u2019s enthusiasm for AI, one shared with his benefactor Elon Musk. Then came Alibaba, bursting back onto the global scene with an ambition that\u2019s also catching global investors\u00a0off guard.<\/p>\n The DeepSeek\/Alibaba rally may have legs thanks to Xi\u2019s decision to, in the words of economist Stephen Jen, \u201cmake\u00a0Chinese\u00a0equities investible again,\u201d starting with\u00a0tech\u00a0platforms.<\/p>\n Jen, CEO of Eurizon SLJ Asset Management, says that \u201cin many ways, this is a call for a continued bounce-back in a long-depressed and unpopular market. For a change, though, there are now many more reasons to be positive than negative on\u00a0Chinese\u00a0equities and China in general<\/a>.\u201d<\/p>\n Can the same be said of SoftBank\u2019s prospects as Son steps away from a $4.7 trillion company at the center of what\u2019s either a once-in-a-century revolution or an asset bubble for the record books.\u00a0<\/p>\n