Musk’s gamble and the dangerous high-tech arms race

Anyone who had bought SpaceX shares before the initial public offering can now sell them, as the lock-up period – during which shareholders are restricted from trading their shares as they wish – expired yesterday. This is an initial opportunity reserved for insiders, and the number of shares available on the market has doubled to 1.55 billion. The IPO on Wall Street, on 12 June, opened at an official price of $135 per share; by the close of trading that Friday, it had already reached $161, and on Tuesday 16th it hit a record high of $201.80. Yesterday, it opened at €107.16. There were no mass sell-offs yesterday morning, as investors were wary of making a loss; there was a rebound to 115, a dip to 110, a rise back to 111 – in short, a yo-yo effect. The 4,400 employees who hold shares are set to become millionaires, it had been reported; from what we can gather, they prefer to wait, hoping that the world’s richest man will keep his promises. Adding fuel to the fire is a story published just yesterday by the Wall Street Journal, which details a colossal scam perpetrated against an unsuspecting investor who had entrusted his savings to the Late State Management fund, a special-purpose vehicle specialising in indirect purchases of SpaceX shares. Rum Rupireddy had bought the shares in advance and thought he owned 2,500 shares worth at least $30,000, only to discover that they had been sold off in 2024. Musk has nothing to do with it, but it speaks volumes about how the bull run in high-tech companies is becoming dangerous. Many have drawn parallels with the dot-com fever that led to the bursting of the bubble in March 2000. The Wall Street Journal article conjures up the spectre of subprime mortgages.

Cheating aside, it is the economic fundamentals that are faltering. SpaceX has presented its first financial results since going public: between April and June, turnover reached $7.8 billion, a 92 per cent increase compared with the same period in 2025. Taking the first six months of this year as a whole, revenues stand at $12.5 billion. Yet Wall Street took this news badly and the share price fell by 12 per cent. Investors looked not only at revenue but also at profits, and here the balance sheet is still in the red – less so than before, but still $541 million in three months. In any case, SpaceX is expanding, but it is not generating substantial profits and, it seems, not even the spectacular stock market gains of the past. Its growth is driven primarily by Starlink, which accounts for around half of turnover; then there are the space operations and, now, artificial intelligence, which is receiving the bulk of new investment. Musk has allocated $18.4 billion to AI, just $1.4 billion to Starlink and a mere $1.2 billion to rockets. All this has resulted in debts of 36.8 billion dollars, compared with 22 billion three months earlier. Let’s summarise the figures for April, May and June: turnover 7.8 billion; debts 36.8 billion; losses half a billion. Never mind, says Musk, we’ve got plenty in the bank; cash and cash equivalents stand at 94.4 billion. But where does it come from? Not from profits, as there aren’t any, so it must be loans and finance leases, judging by the rise in liabilities. Is Elon opening his well-padded wallet? But this only serves to heighten doubts about the actual soundness of the business model.

Legitimate concerns extend to the entire AI sector. Investment is growing ever more substantial; Goldman Sachs estimates it will reach $750 billion by the end of this year and is set to double. Funding is being provided through debt, both from shareholders and from banks, funds, financial vehicles or the supplier companies themselves. Musk has announced a deal with Nvidia to develop hardware designed to run artificial intelligence models and services directly on solar-powered satellites. He has also revealed that the first Starlink V3 satellites will be launched on the next Starship flight. He does not, therefore, wish to leave space behind, but his goal now is to challenge the AI giants who are ahead of him, starting with Anthropic and OpenAI. Musk is like a cyclist forced to pedal without stopping to avoid falling, hoping not to run out of breath. How much is he willing to spend to inflate his own ego even further? And how much can the many ‘Rum Rupireddys’ who have given him money and placed their trust in him risk? Other disruptive innovations had claimed that the age-old pillars of business were obsolete. That didn’t happen. Artificial intelligence is different; it is neither an internet portal nor a mortgage – it is the new global machine. Fair enough, but who’s footing the bill?